FreeIPA 4.0
FreeIPA: Identity/Policy Management
Managing Identity and Authorization Policies for Linux-Based Infrastructures
Edition git
Legal Notice
Copyright
© 2012 Red Hat This material may only be distributed subject to the terms and conditions set forth in the GNU Free Documentation License (GFDL), V1.2 or later (the latest version is presently available at
http://www.gnu.org/licenses/fdl.txt).
Abstract
Identity and policy management — for both users and machines — is a core function for almost any enterprise environment. IPA provides a way to create an identity domain that allows machines to enroll to a domain and immediately access identity information required for single sign-on and authentication services, as well as policy settings that govern authorization and access. This manual covers all aspects of installing, configuring, and managing IPA domains, including both servers and clients. This guide is intended for IT and systems administrators.
FreeIPA is a Fedora-based way to create a security, identity, and authentication domain. The different security and authentication protocols available to Linux and Unix systems (like Kerberos, NIS, DNS, PAM, and sudo) are complex, unrelated, and difficult to manage coherently, especially when combined with different identity stores.
FreeIPA provides a layer that unifies all of these disparate services and simplifies the administrative tasks for managing users, systems, and security. FreeIPA breaks management down into two categories: identity and policy. It centralizes the functions of managing the users and entities within your IT environment (identity) and then provides a framework to define authentication and authorization for a global security framework and user-friendly tools like single sign-on (policy).
With FreeIPA, a Fedora system can easily become the center of an identity/authentication domain and even provide access to the domain for clients of other operating systems. FreeIPA is an integrated system, that builds on existing and reliable technologies like LDAP and certificate protocols, with a robust yet straightforward set of tools (including a web-based UI). The key to identity/policy management with FreeIPA is simplicity and flexibility:
Centralized identity stores for authentication and single sign-on using both integrated LDAP services (with 389 Directory Server) and, optionally, NIS services
Clear and manageable administrative control over system services like PAM, NTP, and sudo
Simplified DNS domains and maintenance
Scalable Kerberos realms and cross-realms which clients can easily join
This guide is written for systems administrators and IT staff who will manage FreeIPA domains, user systems, and servers. This assumes a moderate knowledge of Linux-based systems administration and familiarity with important concepts like access control, LDAP, and Kerberos.
This guide covers every aspect of using FreeIPA, including preparation and installation processes, administrative tasks, and the FreeIPA tools. This guide also explains the major concepts behind both identity and policy management, generally, and FreeIPA features specifically. Administrative tasks in this guide are categorized as either Identity or Policy in the chapter title to help characterize the administrative functions.
2. Examples and Formatting
Each of the examples used in this guide, such as file locations and commands, have certain defined conventions.
Square brackets ([]) are used to indicate an alternative element in a name. For example, if a tool is available in /usr/lib on 32-bit systems and in /usr/lib64 on 64-bit systems, then the tool location may be represented as /usr/lib[64].
2.2. Client Tool Information
The tools for FreeIPA are located in the /usr/bin and the /usr/sbin directories.
The LDAP tools used to edit the FreeIPA directory services, such as ldapmodify and ldapsearch, are from OpenLDAP. OpenLDAP tools use SASL connections by default. To perform a simple bind using a username and password, use the -x argument to disable SASL.
2.3. Text Formatting and Styles
Certain words are represented in different fonts, styles, and weights. Different character formatting is used to indicate the function or purpose of the phrase being highlighted.
Other formatting styles draw attention to important text.
A note provides additional information that can help illustrate the behavior of the system or provide more detail for a specific issue.
Important information is necessary, but possibly unexpected, such as a configuration change that will not persist after a reboot.
A warning indicates potential data loss, as may happen when tuning hardware for maximum performance.
If there is any error in this book or there is any way to improve the documentation, please let us know. Make the bug report as specific as possible, so we can be more effective in correcting any issues.
We appreciate receiving any feedback — requests for new sections, corrections, improvements, enhancements, even new ways of delivering the documentation or new styles of docs.
4. Document Change History
| Revision History |
|---|
| Revision 2.1.0-2 | July 22, 2011 | Ella Deon Lackey |
| Completing first round of content revisions on the chapters for server installation, client installation, DNS, basic usage, managing clients, and the preface. | | Beginning content revisions on the chapters for users, Kerberos, automount, and managing servers. | | Bare initial draft of a tools appendix. | | Bugzilla work: 646226, 646240, 646257, 646267, 68173, 693843, 701465, 709385, 714603, 715015 | | FreeIPA.org tickets: 1183, 1359, 1449, 1058, 1335, 1107, 1355, 1430, 803, 991, 615, 969, 594, 593 | | freeipa-guide trac tickets: 18, 19, 16, 17 |
|
| Revision 2.1.0-1 | May 10, 2011 | Ella Deon Lackey |
| Beginning draft for the Fedora docs project. |
|
Chapter 1. Introduction to FreeIPA
Fedora FreeIPA is a way to create identity stores, centralized authentication, domain control for Kerberos and DNS services, and authorization policies — all on Linux systems, using native Linux tools. While centralized identity/policy/authorization software is hardly new, FreeIPA is one of the only options that supports Linux/Unix domains.
FreeIPA provides a unifying skin for standards-defined, common network services, including PAM, LDAP, Kerberos, DNS, NTP, and certificate services, and it allows Fedora systems to serve as the domain controllers.
FreeIPA defines a domain, with servers and clients who share centrally-managed services, like Kerberos and DNS. This chapter first explains what FreeIPA is. This chapter also covers how all of these services work together within the domain and how servers and clients work with each other.
1.1. FreeIPA v. LDAP: A More Focused Type of Service
At the most basic level, FreeIPA is a domain controller for Linux and Unix machines. FreeIPA defines the domain, using controlling servers and enrolled client machines. This provides centralized structure that has previously been unavailable to Linux/Unix environments, and it does it using native Linux applications and protocols.
1.1.1. A Working Definition for FreeIPA
Security information frequently relates to identities of users, machines, and services. Once the identity is verified, then access to services and resources can be controlled.
For efficiency, risk management, and ease of administration, IT administrators try to manage identities as centrally as possible and to unite identity management with authentication and authorization policies. Historically, Linux environments have had a very difficult time establishing this centralized management. There are a number of different protocols (such as NIS and Kerberos) which define domains, while other applications store data (such as LDAP) and still others manage access (such as sudo). None of these applications talk to each other or use the same management tools. Every application had to be administered separately and it had to be managed locally. The only way to get a consistent identity policy was to copy configuration files around manually or to try to develop a proprietary application to manage identities and policies.
The goal of FreeIPA is to simplify that administrative overhead. Users, machines, services, and polices are all configured in one place, using the same tools. Because FreeIPA creates a domain, multiple machines can all use the same configuration and the same resources simply by joining the domain. Users only have to sign into services once, and administrators only have to manage a single user account.
FreeIPA does three things:
Create a Linux-based and Linux-controlled domain. Both FreeIPA servers and FreeIPA clients are Linux or Unix machines. While FreeIPA can synchronize data with an Active Directory domain to allow integration with Windows servers, it is not an administrative tools for Windows machines and it does not support Windows clients. FreeIPA is a management tool for Linux domains.
Centralize identity management and identity policies.
Build on existing, native Linux applications and protocols. While FreeIPA has its own processes and configuration, its underlying technologies are familiar and trusted by Linux administrators and are well established on Linux systems.
In a sense, FreeIPA isn't making administrators do something new; it is helping them do it better. There are a few ways to illustrate that.
On one extreme is the low control environment. Little Example Corp. has several Linux and Unix servers, but each one is administered separately. All passwords are kept on the local machine, so there is no central identity or authentication process. Tim the IT Guy just has to manage users on every machine, set authentication and authorization policies separately, and maintain local passwords. With FreeIPA, things come to order. There is a simple way to have central user, password, and policy stores, so Tim the IT Guy only has to maintain the identities on one machine (the FreeIPA server) and users and policies are uniformly applied to all machines. Using host-based access control, delegation, and other rules, he can even set different access levels for laptops and remote users.
In the middle is the medium control environment. Mid-Example Corp. has several Linux and Unix servers, but Bill the IT Guy has tried to maintain a greater degree of control by creating a NIS domain for machines, an LDAP directory for users, and Kerberos for authentication. While his environment is well under control, every application has to be maintained separately, using different tools. He also has to update all of the services manually whenever a new machine is added to his infrastructure or when one is taken offline. In this situation, FreeIPA greatly reduces his administrative overhead because it integrates all of the different applications together seamlessly, using a single and simplified tool set. It also makes it possible for him to implement single sign-on services for all of the machines in his domain.
On the other extreme is the absent control environment. At Big Example Corp., most of the systems are Windows based and are managed in a tightly-knit Active Directory forest. However, development, production, and other teams have many Linux and Unix systems — which are basically excluded from the Windows controlled environment. FreeIPA brings native control to the Linux/Unix servers, using their native tools and applications — something that is not possible in an Active Directory forest. Additionally, because FreeIPA is Windows-aware, data can be synchronized between Active Directory and FreeIPA, preserving a centralized user store.
FreeIPA provides a very simple solution to a very common, very specific problem: identity management.
1.1.2. Contrasting FreeIPA with a Standard LDAP Directory
The closest relative to FreeIPA is a standard LDAP directory like 389 Directory Server, but there are some intrinsic differences between what they do and what they're intended to do.
First, it helps to understand what a directory service is. A directory service is a collection of software, hardware, and processes that stores information. While directory services can be highly specific (for example, DNS is a directory service because it stores information on hostnames), a generic directory service can store and retrieve any kind of information. LDAP directories like 389 Directory Server are generic directories. They have a flexible schema that supports entries for users, machines, network entities, physical equipment, and buildings, and that schema can be customized to define entries of almost anything. Because of its extensibility, LDAP servers like 389 Directory Server are frequently used as backends that store data for other applications. 389 Directory Server not only contains information, it organizes information. LDAP directories uses a hierarchical structure, a directory tree, that organize entries into root entries (suffixes), intermediate or container entries (subtrees or branches), and leaf entries (the actual data). Directory trees can be very complex, with a lot of branch points, or very simple (flat) with few branch points.
The primary feature of an LDAP directory is its generality. It can be made to fit into a variety of applications.
FreeIPA, on the other hand, has a very specific purpose and fits a very specific application. It is not a general LDAP directory, it is not a backend, and it is not a general policy server. It is not generic.
FreeIPA focuses on identities (user and machine) and policies that relate to those identities and their interactions. While it uses an LDAP backend to store its data, FreeIPA has a highly-customized and specific set of schema that defines a particular set of identity-related entries and defines them in detail. It has a relatively flat and simple directory tree because it has only a handful of entry types and relationships that are relevant to its purpose. It has rules and limitations on how the FreeIPA server can be deployed because it can only be deployed for a specific purpose: managing identities.
The restrictions on FreeIPA also give it a great deal of administrative simplicity. It has a simple installation process, a unified set of commands, and a clearly defined role in the overall IT infrastructure. A FreeIPA domain is easy to configure, easy to join, and easy to manage, and the functions that it serves — particularly identity/authentication tasks like enterprise-wide single sign-on — are also easier to do with FreeIPA than with a more general-purpose directory service.
Table 1.1. FreeIPA Compared to 389 Directory Server
|
|
389 Directory Server
|
FreeIPA
|
|---|
|
Use
|
General purpose
|
Single domain, focused on identity management
|
|
Flexibility
|
Highly-customizable
|
Limitations to focus on identity and authentication
|
|
Schema
|
Default LDAP schema
|
Optimized, special schema for identity management
|
|
Directory Tree
|
Standard and flexible hierarchy
|
Flat tree with a fixed hierarchy
|
|
Authentication
|
LDAP
|
Kerberos or Kerberos and LDAP
|
|
Active Directory Synchronization
|
Bi-directional
|
Unidirectional, Active Directory to FreeIPA
|
|
Password Policies
|
LDAP-based
|
Kerberos-based
|
|
User Tools
|
Java Console and standard LDAP utilities
|
Web-based UI and special Python command-line tools
|
LDAP directories like 389 Directory Server have flexibility and adaptability which makes them a perfect backend to any number of applications. Its primary purpose is to store and retrieve data efficiently.
FreeIPA fills a very different niche. It is optimized to perform a single task very effectively. It stores user information and authentication and authorization policies, as well as other information related to access, like host information. Its single purpose is to manage identities.
1.2. Bringing Linux Services Together
FreeIPA unifies disparate yet related Linux services into a single management environment. From there, it establishes a simple, easy way to bring host machines into the domain of those services.
A FreeIPA server is, at its core, an identity and authentication server. The primary FreeIPA server, essentially a domain controller, uses a Kerberos server and KDC for authentication. An LDAP backend contains all of the domain information, including users, client machines, and domain configuration.
Other services are included to provide support for the core identity/authentication functions. DNS is used for machine discovery and for connecting to other clients in the domain. NTP is used to synchronize all domain clocks so that logging, certificates, and operations can occur as expected. A certificate service provides certificates for Kerberos-aware services. All of these additional services work together under the control of the FreeIPA server.
The FreeIPA server also has a set of tools which are used to manage all of the FreeIPA-associated services. Rather than managing the LDAP server, KDC, or DNS settings individually, using different tools on local machines, FreeIPA has a single management toolset (CLI and web UI) that allows centralized and cohesive administration of the domain.
1.2.1. Authentication: Kerberos KDC
Kerberos is an authentication protocol. Kerberos uses symmetric key cryptography to generate tickets to users. Kerberos-aware services check the ticket cache (a keytab) and authenticate users with valid tickets.
Kerberos authentication is significantly safer than normal password-based authentication because passwords are never sent over the network — even when services are accessed on other machines.
In FreeIPA, the Kerberos administration server is set up on the FreeIPA domain controller, and all of the Kerberos data are stored in FreeIPA's backend Directory Server. The Directory Server instance defines and enforces access controls for the Kerberos data.
The FreeIPA Kerberos server is managed through FreeIPA tools instead of Kerberos tools because all of its data are stored in the Directory Server instance. The KDC is unaware of the Directory Server, so managing the KDC with Kerberos tools does not effect the FreeIPA configuration.
1.2.2. Data Storage: 389 Directory Server
FreeIPA contains an internal 389 Directory Server instance. All of the Kerberos information, user accounts, groups, services, policy information, DNS zone and host entries, and all other information in FreeIPA is stored in this 389 Directory Server instance.
When multiple servers are configured, they can talk to each other because 389 Directory Server supports multi-master replication. Agreements are automatically configured between the initial server and any additional replicas which are added to the domain.
1.2.3. Authentication: Dogtag Certificate System
Kerberos can use certificates along with keytabs for authentication, and some services require certificates for secure communication. FreeIPA includes a certificate authority, through Dogtag Certificate System, with the server. This CA issues certificates to the server, replicas, and hosts and services within the FreeIPA domain.
The CA can be a root CA or it can have its policies defined by another, external CA (so that it is subordinate to that CA). Whether the CA is a root or subordinate CA is determined when the FreeIPA server is set up.
1.2.4. Server/Client Discovery: DNS
FreeIPA defines a domain — multiple machines with different users and services, each accessing shared resources and using shared identity, authentication, and policy configuration. The clients need to be able to contact each other, as FreeIPA servers. Additionally, services like Kerberos depend on hostnames to identify their principal identities.
Hostnames are associated with IP address using the Domain Name Service (DNS). DNS maps hostnames to IP addresses and IP addresses to hostnames, providing a resource that clients can use when they need to look up a host. From the time a client is enrolled in the FreeIPA domain, it uses DNS to locate the FreeIPA server and then all of the services and clients within the domain.
Multiple DNS servers are usually configured, each one working as an authoritative resource for machines within a specific domain. Having the FreeIPA server also be a DNS server is optional, but it is strongly recommended. When the FreeIPA server also manages DNS, there is tight integration between the DNS zones and the FreeIPA clients and the DNS configuration can be managed using native FreeIPA tools. Even if a FreeIPA server is a DNS server, other external DNS servers can still be used.
Many services require that servers and clients have the same system time, within a certain variance. For example, Kerberos tickets use time stamps to determine their validity. If the times between the server and client skew outside the allowed range, then any Kerberos tickets are invalidated.
Clocks are synchronized over a network using Network Time Protocol (NTP). A central server acts as an authoritative clock and all of the clients which reference that NTP server sync their times to match.
When the FreeIPA server is the NTP server for the domain, all times and dates are synchronized before any other operations are performed. This allows all of the date-related services — including password expirations, ticket and certificate expirations, account lockout settings, and entry create dates — to function as expected.
The FreeIPA server, by default, works as the NTP server for the domain. Other NTP servers can also be used for the hosts.
1.3. Relationships Between Servers and Clients
FreeIPA itself defines a domain, a group of machines that have shared configuration, policies, and identity stores. This shared configuration allows the machines (and users) within the domain to be aware of each other and operate together. This awareness can be used to enable cross-platform compatibility, like unifying Windows and Linux systems, or to enable infrastructure-wide single sign-on.
1.3.1. About FreeIPA Servers and Replicas
FreeIPA works by having identified servers which are the master stores of information for user and machine identities and domain-wide policies. These servers host domain-related services such as certificate authorities, NTP, Kerberos, SSH, and DNS. The server also acts as a central repository of identity and policy information.
Clients interact indirectly with FreeIPA servers when they attempt to access domain resources, such as fileshares, services, remote machines, or authentication (through SSSD and Kerberos).
As said, a FreeIPA server is a controller for a lot of associated services. While a number of those services are support, most of them are not required. For example, a server may have a CA, a DNS server, or an NTP server — or it can be installed without those services.
Once a FreeIPA server is set up, its configuration can be copied and used as the basis for another FreeIPA server. When a FreeIPA server is copied, that copy is called a replica.
The only real different between a FreeIPA server and a FreeIPA replica is that a server is a new installation and a replica is based on an existing server. Once the instance is configured, servers and replicas are basically identical in functionality and behavior within the FreeIPA domain.
There is a good deal of flexibility in the FreeIPA server (and replica) topology. For example, Server A can be installed with a CA and DNS services, while Replica A can be based on Server A's configuration but not host either DNS or CA services. Server B can be added to the domain, also without CA or DNS services. At any time in the future, a CA or DNS service can be created and configured on Replica A or Server B.
Servers and replicas both use underlying LDAP directories to store user and host entries, configuration data, policy configuration, and keytabs, certificates, and keys. Servers and replicas propagate data among each other through multi-master replication agreements. Both servers and replicas are masters in the replication topology.
The replication topology essentially creates a cloud of FreeIPA servers. One benefit of a server domain is automatic load balancing, using the SRV records in DNS. The SRV record priority sets the order that servers and replicas are contacted, while weight distributed the load between servers/replicas with the same priority. The server and replica DNS entries can be edited to change the load balancing, which is covered in
Example 10.4, “SRV Record” and
Section 10.15, “Changing Load Balancing for FreeIPA Servers and Replicas”.
1.3.2. About FreeIPA Clients
A client is simply any machine which is configured to operate within the FreeIPA domain, using its Kerberos and DNS services, NTP settings, and certificate services. That's an important distinction: a client does not require a daemon or (necessarily) an installed product. It requires only system configurations which direct it to use FreeIPA services.
For Fedora systems, a certain number of platform tools are available for FreeIPA to use, such as SSSD. These are FreeIPA-enabled platform applications, which is simply a way of saying that they are aspects of the underlying platform that work with FreeIPA services. Other tools, like certain PAM and NSS modules and FreeIPA command-line utilities, are provided as FreeIPA-specific packages that must be installed on the machine. These are FreeIPA-related components.
FreeIPA uses the local storage (cache) on a client to improve performance in a few ways:
Store FreeIPA information when the machine is offline.
Keep information active beyond its normal timeout period if the client cannot access the central server. The cache is persistent even after rebooting the machine.
Reduce the round-trip time of requests by checking information locally before looking at the server.
Information is stored either in an LDB database (similar to LDAP) or the local filesystem (as XML files), depending on the type of information.
Identity information (about users, machines, and groups) is stored in the LDB database, which uses the same syntax as an LDAP directory. This identity information is originally stored in the FreeIPA server's 389 Directory Server instance. Because this information changes frequently and is referenced frequently, it is important to be able to call the more current information quickly, which is possible using an LDB database on the client and the Directory Server on the server.
Policy information is more static than identity information, and it can include configuration for SELinux or sudo. These policies are set globally on the server and then are propagated to the clients. On the client, the policy information is stored in the filesystem in XML files which can be downloaded and converted into a native file for whatever service is being managed.
A specific set of services on the FreeIPA server interact with a subset of services and modules on the FreeIPA client. A client is any machine (a host) which can retrieve a keytab or certificates from the FreeIPA domain.
SSSD provides the user authentication for the machine and enforces host-based access control rules.
certmonger monitors and renews the certificates on the client. It can request new certificates for the services on the system, including virtual machines.
When a Fedora client is added to the domain (enrolled), its SSSD and certmonger are configured to connect to the FreeIPA server and the required Kerberos keytab and host certificates are created. (The host certificate is not used directly by FreeIPA; it may be used by other services, such as a web server.)
Chapter 2. Installing a FreeIPA Server
The FreeIPA domain is defined and managed by a FreeIPA server which is essentially a domain controller. There can be multiple domain controllers within a domain for load-balancing and failover tolerance. These additional servers are called replicas of the master FreeIPA server.
Both FreeIPA servers and replicas only run on Fedora systems. For both servers and replicas, the necessary packages must be installed and then the FreeIPA server or replica itself is configured through setup scripts, which configure all of the requisite services.
2.1. Preparing to Install the FreeIPA Server
Before you install FreeIPA, ensure that the installation environment is suitably configured. You also need to provide certain information during the installation and configuration procedures, including realm names and certain usernames and passwords. This section describes the information that you need to provide.
2.1.1. Hardware Recommendations
A basic user entry is about 1 KB in size, as is a simple host entry with a certificate. The most important hardware feature to size properly is RAM. While all deployments are different, depending on the number of users and groups and the type of data stored, there is a rule of thumb to use to help determine how much RAM to use:
For 10,000 users and 100 groups, have at least 2GB of RAM and 1GB swap space.
For 100,000 users and 50,000 groups, have at least 16GB of RAM and 4GB of swap space.
For larger deployments, it is more effective to increase the RAM than to increase disk space because much of the data are stored in cache.
2.1.2. Software Requirements
Most of the packages that a FreeIPA server depends on are installed as dependencies when the FreeIPA packages are installed. There are some packages, however, which are required before installing the FreeIPA packages:
2.1.3. Supported Web Browsers
These browsers are supported for connecting to the web UI:
Firefox 15.x and newer
Firefox 10.x
Firefox 3.6
2.1.4. System Prerequisites
The FreeIPA server is set up using a configuration script, and this script makes certain assumption about the host system. If the system does not meet these prerequisites, then server configuration may fail.
2.1.4.1. Hostname and IP Address Requirements
Regardless of whether the DNS is within the FreeIPA server or external, the server host must have DNS properly configured:
The hostname must be a fully-qualified domain name. For example, ipaserver.example.com.
This must be a valid DNS name, which means only numbers, alphabetic characters, and hyphens (-) are allowed. Other characters, like underscores, in the hostname will cause DNS failures.
The hostname must be all lower-case.
The server's A record must be set and resolve to its public IP address.
The fully-qualified domain name cannot resolve to the loopback address. It must resolve to the machine's public IP address, not to 127.0.0.1. The output of the hostname command cannot be localhost or localhost6.
The server's hostname and IP address must be in its own /etc/hosts file.
It is recommended that a separate DNS domain be allocated for the FreeIPA server. While not required (clients from other domains can still be enrolled in the FreeIPA domain), this is a convenience for overall DNS management.
If the FreeIPA server is configured to host its own DNS server, any previous existing DNS ignored. A records and PTR records do not need to match for the FreeIPA server machine, and the machine can have any configured IP address.
2.1.4.2. Directory Server
There must not be any instances of 389 Directory Server installed on the host machine.
The server script overwrites system files to set up the FreeIPA domain. The system should be clean, without custom configuration for services like DNS and Kerberos, before configuring the FreeIPA server.
FreeIPA uses a number of ports to communicate with its services. These ports, listed in
Table 2.1, “FreeIPA Ports”, must be open and available for FreeIPA to work. They cannot be in use by another service or blocked by a firewall. To make sure that these ports are available, try
iptables to list the available ports or
nc,
telnet, or
nmap to connect to a port or run a port scan.
To open a port:
[root@server ~]# iptables -A INPUT -p tcp --dport 389 -j ACCEPT
The iptables man page has more information on opening and closing ports on a system.
Table 2.1. FreeIPA Ports
|
Service
|
Ports
|
Type
|
|---|
|
HTTP/HTTPS
|
|
TCP
|
|
LDAP/LDAPS
|
|
TCP
|
|
Kerberos
|
|
TCP and UDP
|
|
DNS
|
53
|
TCP and UDP
|
|
NTP
|
123
|
UDP
|
|
Dogtag Certificate System - LDAP
|
7389
|
TCP
|
If a server is being installed on a virtual machine, that server should not run an NTP server. To disable NTP for FreeIPA, use the --no-ntp option.
It is strongly recommended that you avoid or restrict the use of nscd in a FreeIPA deployment. The nscd service is extremely useful for reducing the load on the server, and for making clients more responsive, but there can be problems when a system is also using SSSD, which performs its own caching.
nscd caches authentication and identity information for all services that perform queries through nsswitch, including getent. Because nscd performs both positive and negative caching, if a request determines that a specific FreeIPA user does not exist, it marks this as a negative cache. Values stored in the cache remain until the cache expires, regardless of any changes that may occur on the server. The results of such caching is that new users and memberships may not be visible, and users and memberships that have been removed may still be visible.
Avoid clashes with SSSD caches and to prevent locking out users, avoid using nscd altogether. Alternatively, use a shorter cache time by resetting the time-to-live caching values in the /etc/nscd.conf file:
positive-time-to-live group 3600
negative-time-to-live group 60
positive-time-to-live hosts 3600
negative-time-to-live hosts 20
2.1.5.1. Configuring Networking Services
2.1.5.2. Configuring the /etc/hosts File
You need to ensure that your /etc/hosts file is configured correctly. A misconfigured file can prevent the FreeIPA command-line tools from functioning correctly and can prevent the FreeIPA web interface from connecting to the FreeIPA server.
Configure the /etc/hosts file to list the FQDN for the FreeIPA server before any aliases. Also ensure that the hostname is not part of the localhost entry. The following is an example of a valid hosts file:
127.0.0.1 localhost.localdomain localhost
::1 localhost6.localdomain6 localhost6
192.168.1.1 ipaserver.example.com ipaserver
Do not omit the IPv4 entry in the /etc/hosts file. This entry is required by the FreeIPA web service.
2.2. Installing the FreeIPA Server Packages
Installing only the FreeIPA server requires a single package, freeipa-server. If the FreeIPA server will also manage a DNS server, then it requires two additional packages to set up the DNS.
All of these packages can be installed using the yum command:
[root@server ~]# yum install freeipa-server bind bind-dyndb-ldap
Installing the freeipa-server also installs a large number of dependencies, such as 389-ds-base for the LDAP service and krb5-server for the Kerberos service, along with FreeIPA tools.
2.3. Creating a FreeIPA Server Instance
The FreeIPA setup script creates a server instance, which includes configuring all of the required services for the FreeIPA domain:
The network time daemon (ntpd)
A 389 Directory Server instance
A Kerberos key distribution center (KDC)
Apache (httpd)
An updated SELinux targeted policy
The Active Directory WinSync plug-in
A certificate authority
Optional. A domain name service (DNS) server
The FreeIPA setup process can be minimal, where the administrator only supplies some required information, or it can be very specific, with user-defined settings for many parts of the FreeIPA services. The configuration is passed using arguments with the ipa-server-install script.
2.3.1. About ipa-server-install
A FreeIPA server instance is created by running the ipa-server-install script. This script can accept user-defined settings for services, like DNS and Kerberos, that are used by the FreeIPA instance, or it can supply predefined values for minimal input from the administrator.
While ipa-server-install can be run without any options, so that it prompts for the required information, it has numerous arguments which allow the configuration process to be easily scripted or to supply additional information which is not requested during an interactive installation.
Table 2.2. ipa-server-install Options
|
Argument
|
Description
|
|---|
|
-a ipa_admin_password
|
The password for the FreeIPA administrator. This is used for the admin user to authenticate to the Kerberos realm.
|
|
--hostname=hostname
|
The fully-qualified domain name of the FreeIPA server machine.
This must be a valid DNS name, which means only numbers, alphabetic characters, and hyphens (-) are allowed. Other characters, like underscores, in the hostname will cause DNS failures.
Additionally, the hostname must all be lower-case. No capital letters are allowed.
|
|
-n domain_name
|
The name of the LDAP server domain to use for the FreeIPA domain. This is usually based on the FreeIPA server's hostname.
|
|
-p directory_manager_password
|
The password for the superuser, cn=Directory Manager, for the LDAP service.
|
|
-r realm_name
|
The name of the Kerberos realm to create for the FreeIPA domain.
|
|
--subject=subject_DN
|
Sets the base element for the subject DN of the issued certificates. This defaults to O=realm.
|
|
--forwarder=forwarder
|
Gives a DNS forwarder to use with the DNS service. To specify more than one forwarder, use this option multiple times.
|
|
--no-forwarders
|
Uses root servers with the DNS service instead of forwarders.
|
|
--no-reverse
|
Does not create a reverse DNS zone when the DNS domain is set up. (If a reverse DNS zone is already configured, then that existing reverse DNS zone is used.) If this option is not used, then the default value is true, which assumes that reverse DNS should be configured by the installation script.
|
|
--setup-dns
|
Tells the installation script to set up a DNS service within the FreeIPA domain. Using an integrated DNS service is optional, so if this option is not passed with the installation script, then no DNS is configured.
|
|
--idmax=number
|
Sets the upper bound for IDs which can be assigned by the FreeIPA server. The default value is the ID start value plus 199999.
|
|
--idstart=number
|
Sets the lower bound (starting value) for IDs which can be assigned by the FreeIPA server. The default value is randomly selected.
|
2.3.2. Setting up a FreeIPA Server: Basic Interactive Installation
All that is required to set up a FreeIPA server is to run the ipa-server-install script. This launches the script interactively, which prompts for the required information to set up a server, but without more advanced configuration like DNS and CA options.
Run the ipa-server-install script.
[root@server ~]# ipa-server-install
Do you want to configure integrated DNS (BIND)? [no]:
Enter the hostname. This is determined automatically using reverse DNS.
Server host name [ipaserver.example.com]:
Enter the domain name. This is determined automatically based on the hostname.
Please confirm the domain name [example.com]:
Enter the new Kerberos realm name. This is usually based on the domain name.
Please provide a realm name [EXAMPLE.COM]:
Enter the password for the Directory Server superuser, cn=Directory Manager. There are password strength requirements for this password, including a minimum password length.
Directory Manager password:
Password (confirm):
Enter the password for the FreeIPA system user account, admin. This user is created on the machine.
IPA admin password:
Password (confirm):
The script then reprints the hostname, IP address, domain name and realm name.
The IPA Master Server will be configured with
Hostname: ipaserver.example.com
IP address: 192.168.1.1
Domain name: example.com
Realm name: EXAMPLE.COM
Continue to configure the system with these values? [no]: yes
After that, the script configures all of the associated services for FreeIPA, with task counts and progress bars.
Configuring NTP daemon (ntpd)
[1/4]: stopping ntpd
...
Done configuring NTP daemon (ntpd).
Configuring directory server (dirsrv): Estimated time 1 minute
[1/38]: creating directory server user
...
Configuring certificate server (pki-tomcatd): Estimated time 3 minutes 30 seconds
[1/20]: creating certificate server user
...
Done configuring certificate server (pki-tomcatd).
Configuring Kerberos KDC (krb5kdc): Estimated time 30 seconds
[1/10]: adding sasl mappings to the directory
...
Done configuring Kerberos KDC (krb5kdc).
Configuring kadmin
[1/2]: starting kadmin
[2/2]: configuring kadmin to start on boot
Done configuring kadmin.
Configuring ipa_memcached
[1/2]: starting ipa_memcached
[2/2]: configuring ipa_memcached to start on boot
Done configuring ipa_memcached.
Configuring ipa-otpd
[1/2]: starting ipa-otpd
[2/2]: configuring ipa-otpd to start on boot
Done configuring ipa-otpd.
Configuring the web interface (httpd): Estimated time 1 minute
[1/15]: disabling mod_ssl in httpd
...
Done configuring the web interface (httpd).
Applying LDAP updates
Restarting the directory server
Restarting the KDC
Sample zone file for bind has been created in /tmp/sample.zone.pUfcGp.db
Restarting the web server
========================================================================
Setup complete
Restart the
SSH service to retrieve the Kerberos principal and to refresh the name server switch (NSS) configuration file:
[root@server ~]# service sshd restart
Authenticate to the Kerberos realm using the admin user's credentials to ensure that the user is properly configured and the Kerberos realm is accessible.
[root@server ~]# kinit admin
Password for admin@EXAMPLE.COM:
Test the FreeIPA configuration by running a command like ipa user-find. For example:
[root@server ~]# ipa user-find admin
--------------
1 user matched
--------------
User login: admin
Last name: Administrator
Home directory: /home/admin
Login shell: /bin/bash
UID: 939000000
GID: 939000000
Account disabled: False
Password: True
Kerberos keys available: True
----------------------------
Number of entries returned 1
----------------------------
2.3.3. Examples of Creating the FreeIPA Server
The way that a FreeIPA server is installed can be different depending on the network environment, security requirements within the organization, and the desired topology. These example illustrate some common options when installing the server. These examples are not mutually exclusive; it is entirely possible to use CA options, DNS options, and FreeIPA configuration options in the same server invocation. These are called out separately simply to make it more clear what each configuration area requires.
2.3.3.1. Non-Interactive Basic Installation
This information can be passed with the ipa-server-install, along with the -U to force it to run without requiring user interaction.
Example 2.1. Basic Installation without Interaction
[root@server ~]# ipa-server-install -a secret12 --hostname=ipaserver.example.com -r EXAMPLE.COM -p secret12 -n example.com -U
The script then prints the submitted values:
To accept the default shown in brackets, press the Enter key.
The IPA Master Server will be configured with
Hostname: ipaserver.example.com
IP address: 192.168.1.1
Domain name: example.com
Realm name: EXAMPLE.COM
The server name must be a valid DNS name, which means only numbers, alphabetic characters, and hyphens (-) are allowed. Other characters, like underscores, in the hostname will cause DNS failures. Additionally, the hostname must all be lower-case. No capital letters are allowed.
2.3.3.2. Using Different CA Configurations
FreeIPA uses an integrated certificate authority (CA) to create the certificates and keytabs used by users and hosts within the domain. There are two different ways that FreeIPA incorporates the CA into the FreeIPA server:
The installation script installs a root Dogtag Certificate System CA. The Dogtag Certificate System CA provides the full range of certificate services, based on policies that are defined in the Dogtag Certificate System configuration.
This is the default configuration.
Alternatively, the installation script can set up a Dogtag Certificate System CA that is subordinate to an external CA. A subordinate CA is chained to another CA, and it uses the policies and restrictions defined by that external CA. The root CA can be an external CA like Verisign or a corporate CA.
A Dogtag Certificate System CA is still installed and configured as part of the FreeIPA server installation, but its CA certificates are issued by the external CA rather than by itself.
The FreeIPA server can use a certificate issued by an external CA. This can be a corporate CA or a third-party CA like Verisign or Thawte. As with a normal setup process, using an external CA still uses a Dogtag Certificate System instance for the FreeIPA server for issuing all of its client and replica certificates; the initial CA certificate is simply issued by a different CA.
When using an external CA, there are two additional steps that must be performed: submit the generated certificate request to the external CA and then load the CA certificate and issued server certificate to complete the setup.
Example 2.2. Using an External CA
Run the ipa-server-install script, using the --external-ca option.
[root@server ~]# ipa-server-install -a secret12 -r EXAMPLE.COM -P password -p secret12 -n ipaserver.example.com --external-ca
The script sets up the NTP and Directory Server services as normal.
The script completes the CA setup and returns information about where the certificate signing request (CSR) is located, /root/ipa.csr. This request must be submitted to the external CA.
Configuring certificate server: Estimated time 6 minutes
[1/4]: creating certificate server user
[2/4]: creating pki-ca instance
[3/4]: restarting certificate server
[4/4]: configuring certificate server instance
The next step is to get /root/ipa.csr signed by your CA and re-run ipa-server-install.
Submit the request to the CA. The process differs for every service.
Retrieve the issued certificate and the CA certificate chain for the issuing CA. Again, the process differs for every certificate service, but there is usually a download link on a web page or in the notification email that allows administrators to download all the required certificates. Be sure to get the full certificate chain for the CA, not just the CA certificate.
Rerun ipa-server-install, specifying the locations and names of the certificate and CA chain files. For example:
[root@server ~]# ipa-server-install --external_cert_file=/tmp/servercert20110601.p12 --external_ca_file=/tmp/cacert.p12
FreeIPA can be configured to manage its own DNS, use an existing DNS, or not use DNS services at all (which is the default). Running the setup script alone does not configure DNS; this requires the --setup-dns option.
As with a basic setup, the DNS setup can either prompt for the required information or the DNS information can be passed with the script to allow an automatic or unattended setup process.
Example 2.3. Interactive DNS Setup
Run the ipa-server-install script, using the --setup-dns option.
[root@server ~]# ipa-server-install -a secret12 -r EXAMPLE.COM -P password -p secret12 -n ipaserver.example.com --setup-dns
The script configures the hostname and domain name as normal.
The script then prompts for DNS forwarders. If forwarders will be used, enter yes, and then supply the list of DNS servers. If FreeIPA will manage its own DNS service, then enter no.
Do you want to configure DNS forwarders? [yes]: no
No DNS forwarders configured
The script sets up the NTP, Directory Server, Certificate System, Kerberos, and Apache services.
Before completing the configuration, the script prompts to ask whether it should configure reverse DNS services. If you select yes, then it configures the named service.
Do you want to configure the reverse zone? [yes]: yes
Configuring DNS (named)
[1/11]: adding DNS container
[2/11]: setting up our zone
[3/11]: setting up reverse zone
[4/11]: setting up our own record
[5/11]: setting up records for other masters
[6/11]: setting up CA record
[7/11]: setting up kerberos principal
[8/11]: setting up named.conf
[9/11]: restarting named
[10/11]: configuring named to start on boot
[11/11]: changing resolv.conf to point to ourselves
Done configuring DNS (named).
==============================================================================
Setup complete
If DNS is used with FreeIPA, then two pieces of information are required: any DNS forwarders that will be used and using (or not) reverse DNS. To perform a non-interactive setup, this information can be passed using the --forwarder or --no-forwarders option and --no-reverse option.
Example 2.4. Setting up DNS Non-Interactively
To use DNS always requires the --setup-dns. To user forwarders, use the --forwarder option; for multiple forwarders, use multiple invocations of --forwarder.
[root@server ~]# ipa-server-install ... --setup-dns --forwarder=1.2.3.0 --forwarder=1.2.255.0
Some kind of forwarder information is required. If no external forwarders will be used with the FreeIPA DNS service, then use the --no-forwarders option to indicate that only root servers will be used.
The script always assumes that reverse DNS is configured along with DNS, so it is not necessary to use any options to enable reverse DNS. To disable reverse DNS, use the --no-reverse option; if a reverse DNS zone is already configured, then using the --no-reverse option means that existing reverse DNS zone is used.
[root@server ~]# ipa-server-install ... --setup-dns --no-reverse
2.3.4. Troubleshooting Installation Problems
ipa: ERROR: Kerberos error: ('Unspecified GSS failure. Minor code may provide more information', 851968)/('Decrypt integrity check failed', -1765328353)
There are two potential causes for this:
ipaserver named[6886]: failed to dynamically load driver 'ldap.so': libldap-2.4.so.2: cannot open shared object file: No such file or directory
This usually means that the
bind-chroot package is installed and is preventing the
named service from starting. To resolve this issue, remove the
bind-chroot package and then restart the FreeIPA server.
[root@server ~]# yum remove bind-chroot
[root@server ~]# ipactl restart
2.4. Setting up FreeIPA Replicas
In the FreeIPA domain, there are three types of machines:
Servers, which manage all of the services used by domain members
Replicas, which are essentially copies of servers (and, once copied, are identical to servers)
Clients, which belong to the Kerberos domains, receive certificates and tickets issued by the servers, and use other centralized services for authentication and authorization
A replica is a clone of a specific FreeIPA server. The server and replica share the same internal information about users, machines, certificates, and configured policies. These data are copied from the server to the replica in a process called replication. The two Directory Server instances used by an FreeIPA server — the Directory Server instance used by the FreeIPA server as a data store and the Directory Server instance used by the Dogtag Certificate System to store certificate information — are replicated over to corresponding consumer Directory Server instances used by the FreeIPA replica.
If you are using a Dogtag Certificate System instance as the CA for the FreeIPA domain, then it is possible to make a replica of a replica.
2.4.1. Prepping and Installing the Replica Server
Replicas are functionally the same as FreeIPA servers, so they have the same installation requirements and packages.
The replica must be the same version as the original master server. If the master server is running on Red Hat Enterprise Linux 7.0, FreeIPA version 3.4.x, then the replica must also run on Red Hat Enterprise Linux 7.0 and use the FreeIPA 3.4.x packages. Creating a replica of a different version than the master is not supported.
There is an exception to the rule: A newer version of the replica can be installed as a part of the upgrade process. All other replicas have to be upgraded to the same version in a matter of days or weeks. FreeIPA servers should not be run for a longer time period with different versions.
[root@server ~]# yum install freeipa-server bind bind-dyndb-ldap
Do not run the ipa-server-install script.
The replica and the master server must be running the same version of FreeIPA.
If there is an existing Dogtag Certificate System or Red Hat Certificate System instance on the replica machine, make sure that port 7389 is free. This port is used by the master FreeIPA server to communicate with the replica.
Make sure the appropriate ports are open on both the server and the replica machine during and after the replica configuration. Servers and replicas connect to each other over ports 9443, 9444, 9445, and 7389 during the replica configuration. Once the replica is set up, the server and replica communicate over port 7389.
2.4.2. Creating the Replica
On the master server, create a replica information file. This contains realm and configuration information taken from the master server which will be used to configure the replica server.
Run the ipa-replica-prepare command on the master FreeIPA server. The command requires the fully-qualified domain name of the replica machine. Using the --ip-address option automatically creates DNS entries for the replica, including the A and PTR records for the replica to the DNS.
[root@server ~]# ipa-replica-prepare ipareplica.example.com --ip-address 192.168.1.2
Preparing replica for ipareplica.example.com from ipaserver.example.com
Creating SSL certificate for the Directory Server
Creating SSL certificate for the dogtag Directory Server
Saving dogtag Directory Server port
Creating SSL certificate for the Web Server
Exporting RA certificate
Copying additional files
Finalizing configuration
Packaging replica information into /var/lib/ipa/replica-info-ipareplica.example.com.gpg
Adding DNS records for ipareplica.example.com
Using reverse zone 1.168.192.in-addr.arpa.
The ipa-replica-prepare command was successful
This must be a valid DNS name, which means only numbers, alphabetic characters, and hyphens (-) are allowed. Other characters, like underscores, in the hostname will cause DNS failures.
Additionally, the hostname must all be lower-case. No capital letters are allowed.
For more options with ipa-replica-prepare, see the ipa-replica-prepare manpage.
Each replica information file is created in the /var/lib/ipa/ directory as a GPG-encrypted file. Each file is named specifically for the replica server for which it is intended, such as replica-info-ipareplica.example.com.gpg.
A replica information file cannot be used to create multiple replicas. It can only be used for the specific replica and machine for which it was created.
Replica information files contain sensitive information. Take appropriate steps to ensure that they are properly protected.
Copy the replica information file to the replica server:
[root@server ~]# scp /var/lib/ipa/replica-info-ipareplica.example.com.gpg root@ipareplica:/var/lib/ipa/
On the replica server, run the replica installation script, referencing the replication information file. There are other options for setting up DNS, much like the server installation script. Additionally, there is an option to configure a CA for the replica; while CA's are installed by default for servers, they are optional for replicas.
Some information about DNS forwarders is required. A list can be given of configured DNS forwarders using a --forwarder option for each one, or forwarder configuration can be skipped by specifying the --no-forwarders option.
For example:
[root@ipareplica ~]# ipa-replica-install --setup-ca --setup-dns --no-forwarders /var/lib/ipa/replica-info-ipareplica.example.com.gpg
Directory Manager (existing master) password:
Warning: Hostname (ipareplica.example.com) not found in DNS
Run connection check to master
Check connection from replica to remote master 'ipaserver.example.com':
Directory Service: Unsecure port (389): OK
Directory Service: Secure port (636): OK
Kerberos KDC: TCP (88): OK
Kerberos Kpasswd: TCP (464): OK
HTTP Server: Unsecure port (80): OK
HTTP Server: Secure port (443): OK
The following list of ports use UDP protocol and would need to be
checked manually:
Kerberos KDC: UDP (88): SKIPPED
Kerberos Kpasswd: UDP (464): SKIPPED
Connection from replica to master is OK.
Start listening on required ports for remote master check
Get credentials to log in to remote master
admin@EXAMPLE.COM password:
Execute check on remote master
admin@example.com's password:
Check connection from master to remote replica 'ipareplica.example.com':
Directory Service: Unsecure port (389): OK
Directory Service: Secure port (636): OK
Kerberos KDC: TCP (88): OK
Kerberos KDC: UDP (88): OK
Kerberos Kpasswd: TCP (464): OK
Kerberos Kpasswd: UDP (464): OK
HTTP Server: Unsecure port (80): OK
HTTP Server: Secure port (443): OK
Connection from master to replica is OK.
Connection check OK
Additional options for the replica installation script are listed in the ipa-replica-install manpage.
The replica installation script runs a test to ensure that the replica file being installed matches the current hostname. If they do not match, the script returns a warning message and asks for confirmation. This could occur on a multi-homed machine, for example, where mismatched hostnames may not be an issue.
Enter the Directory Manager password when prompted. The script then configures a Directory Server instance based on information in the replica information file and initiates a replication process to copy over data from the master server to the replica, a process called initialization.
Verify that the proper DNS entries were created so that FreeIPA clients can discover the new server. DNS entries are required for required domain services:
ldap._tcp
_kerberos._tcp
_kerberos._udp
_kerberos-master._tcp
_kerberos-master._udp
_ntp._udp
If the initial FreeIPA server was created with DNS enabled, then the replica is created with the proper DNS entries. For example:
[root@ipareplica ~]# DOMAIN=example.com
[root@ipareplica ~]# NAMESERVER=ipareplica
[root@ipareplica ~]# for i in _ldap._tcp _kerberos._tcp _kerberos._udp _kerberos-master._tcp _kerberos-master._udp _ntp._udp; do echo ""; dig @${NAMESERVER} ${i}.${DOMAIN} srv +nocmd +noquestion +nocomments +nostats +noaa +noadditional +noauthority; done | egrep -v "^;" | egrep _
_ldap._tcp.example.com. 86400 IN SRV 0 100 389 ipaserver1.example.com.
_ldap._tcp.example.com. 86400 IN SRV 0 100 389 ipaserver2.example.com.
_kerberos._tcp.example.com. 86400 IN SRV 0 100 88 ipaserver1.example.com.
...8<...
If the initial FreeIPA server was created without DNS enabled, then each DNS entry, including both TCP and UPD entries for some services, should be added manually. For example:
[root@ipareplica ~]# kinit admin
[root@ipareplica ~]# ipa dnsrecord-add example.com _ldap._tcp --srv-rec="0 100 389 ipareplica.example.com."
Optional. Set up DNS services for the replica. These are not configured by the setup script, even if the master server uses DNS.
Use the
ipa-dns-install command to install the DNS manually, then use the
ipa dnsrecord-add command to add the required DNS records. For example:
[root@ipareplica ~]# ipa-dns-install
[root@ipareplica ~]# ipa dnsrecord-add example.com @ --ns-rec ipareplica.example.com.
Use the fully-qualified domain name of the replica, including the final period (.), otherwise BIND will treat the hostname as relative to the domain.
2.4.3. Troubleshooting Replica Installation
The only resolution is to uninstall the replica:
[root@ipareplica ~]# ipa-server-install --uninstall
After uninstalling the replica, ensure that port 7389 on the replica is available, and retry the replica installation.
slapd_ldap_sasl_interactive_bind - Error: could not perform interactive bind for id [] mech [GSSAPI]: error -2 (Local error) (SASL(-1): generic failure: GSSAPI Error: Unspecified GSS failure. Minor code may provide more information (Credentials cache file '/tmp/krb5cc_496' not found)) ...
The replica is looking for a credentials cache in /tmp/krb5cc_496 (where 496 is the 389 Directory Server user ID) and cannot find it.
There may also be messages that the server could not obtain Kerberos credentials for the host principal:
set_krb5_creds - Could not get initial credentials for principal [ldap/ replica1.example.com] in keytab [WRFILE:/etc/dirsrv/ds.keytab]: -1765328324 (Generic error)
These errors are both related to how and when the 389 Directory Server instance loads its Kerberos credentials cache.
While 389 Directory Server itself supports multiple different authentication mechanisms, FreeIPA only uses GSS-API for Kerberos connections. The 389 Directory Server instance for FreeIPA keeps its Kerberos credentials cache in memory. When the 389 Directory Server process ends — like when the FreeIPA replica is stopped — the credentials cache is destroyed.
Also, the 389 Directory Server is used as the backend storage for the principal information for the KDC.
When the replica then restarts, the 389 Directory Server instance starts first, since it supplies information for the KDC, and then the KDC server starts. This start order is what causes the GSS-API and Kerberos connection errors.
The 389 Directory Server attempts to open a GSS-API connection, but since there is no credentials cache yet and the KDC is not started, the GSS connection fails. Likewise, any attempt to obtain the host credentials also fails.
These errors are transient. The 389 Directory Server re-attempts the GSS-API connection after the KDC starts and it has a credentials cache. The 389 Directory Server logs then record a bind resumed message.
These startup GSS-API connection failures can be ignored as long as that connection is successfully established.
2.5. Uninstalling FreeIPA Servers and Replicas
To uninstall both a FreeIPA server and a FreeIPA replica, pass the
--uninstall option to the
ipa-server-install command:
[root@ipareplica ~]# ipa-server-install --uninstall
FreeIPA is generally updated whenever a system is upgraded to a new release. Upgrades should be transparent and do not require any user or administrative intervention.
2.6.1. Upgrading Packages
The FreeIPA server packages are updated when the system packages are updated:
[root@ipaserver ~]# yum update *
This is the easiest way to upgrade the server because it automatically pulls in updates for related services, like SSSD, which provide FreeIPA functionality.
To upgrade the FreeIPA server packages specifically, run yum on the master server:
[root@ipaserver ~]# yum update freeipa-server
It can take several seconds for the update process to apply all of the changes.
It is not necessary to update all servers and replicas at precisely the same time; the FreeIPA servers will still work with each other and replicate data successfully. The older FreeIPA servers will simply lack the new features.
Chapter 3. Setting up Systems as FreeIPA Clients
A client is any system which is a member of the FreeIPA domain. While this is frequently a Fedora system (and FreeIPA has special tools to make configuring Fedora clients very simple), machines with other operating systems can also be added to the FreeIPA domain.
One important aspect of a FreeIPA client is that only the system configuration determines whether the system is part of the domain. (The configuration includes things like belonging to the Kerberos domain, DNS domain, and having the proper authentication and certificate setup.)
FreeIPA does not require any sort of agent or daemon running on a client for the client to join the domain. However, for the best management options, security, and performance, clients should run the System Security Services Daemon (SSSD).
This chapter explains how to configure a system to join a FreeIPA domain.
Clients can only be configured after at least one FreeIPA server has been installed.
3.1. What Happens in Client Setup
Whether the client configuration is performed automatically on Fedora systems using the client setup script or manually on other systems, the general process of configuring a machine to serve as a FreeIPA client is mostly the same, with slight variation depending on the platform:
Retrieve the CA certificate for the FreeIPA CA.
Create a separate Kerberos configuration to test the provided credentials. This enables a Kerberos connection to the FreeIPA XML-RPC server, necessary to join the FreeIPA client to the FreeIPA domain. This Kerberos configuration is ultimately discarded.
Setting up the Kerberos configuration includes specifying the realm and domain details, and default ticket attributes. Forwardable tickets are configured by default, which facilitates connection to the administration interface from any operating system, and also provides for auditing of administration operations. For example, this is the Kerberos configuration for Fedora systems:
[libdefaults]
default_realm = EXAMPLE.COM
dns_lookup_realm = false
dns_lookup_kdc = false
rdns = false
forwardable = yes
ticket_lifetime = 24h
[realms]
EXAMPLE.COM = {
kdc = ipaserver.example.com:88
admin_server = ipaserver.example.com:749
}
[domain_realm]
.example.com = EXAMPLE.COM
example.com = EXAMPLE.COM
Run the ipa-join command to perform the actual join
Obtain a service principal for the host service and installs it into /etc/krb5.keytab. For example, host/ipa.example.com@EXAMPLE.COM.
Enable certmonger, retrieve an SSL server certificate, and install the certificate in /etc/pki/nssdb.
Disable the nscd daemon.
Configures SSSD or LDAP/KRB5, including NSS and PAM configuration files.
Configures an OpenSSH server and client, as well as enabling the host to create DNS SSHFP records.
Configure NTP.
3.2. Supported Platforms for FreeIPA Clients
These platforms can be configured to be FreeIPA clients:
Fedora 14, 15, 16, 17, 18 and 19
Windows XP and later
Solaris 9 and 10
HP-UX 11i
AIX 5.3 and 6.1
FreeIPA uses a number of ports to communicate with its services. These ports, listed in
Table 3.1, “FreeIPA Ports”, must be open and available for FreeIPA to work. They cannot be in use by another service or blocked by a firewall. To make sure that these ports are available, try
iptables to list the available ports or
nc,
telnet, or
nmap to connect to a port or run a port scan.
To open a port:
# iptables -A INPUT -p tcp --dport 389 -j ACCEPT
The iptables man page has more information on opening and closing ports on a system.
Table 3.1. FreeIPA Ports
|
Service
|
Ports
|
Type
|
|---|
|
HTTP/HTTPS
|
|
TCP
|
|
LDAP/LDAPS
|
|
TCP
|
|
Kerberos
|
|
TCP and UDP
|
|
DNS
|
53
|
TCP and UDP
|
|
NTP
|
123
|
UDP
|
3.4. Configuring a Fedora System as a FreeIPA Client
There are two elements to prepare before beginning the client setup process for the Fedora client:
There must be a way to connect the client machine to the Kerberos domain, either by having an available Kerberos identity (such as the admin user) or by manually adding the client machine to the KDC on the server with a one-time password before beginning the enrollment process for the client machine.
If there is an Active Directory server on the same network that serves DNS records, the Active Directory DNS records could prevent the client from automatically detecting the FreeIPA server address. The ipa-client-install script retrieves the Active Directory DNS records instead of any records that were added for FreeIPA.
In this case, it is necessary to pass the FreeIPA server address directly to the ipa-client-install script.
To configure the client:
Install the client packages. These packages provide a simple way to configure the system as a client; they also install and configure SSSD.
For a regular user system, this requires only the ipa-client package:
[root@client ~]# yum install freeipa-client
An administrator machine requires the freeipa-admintools package, as well:
[root@client ~]# yum install freeipa-client freeipa-admintools
If the FreeIPA server is configured as the DNS server and is in the same domain as the client, add the server's IP address as the first entry in the client's /etc/resolv.conf file.
If every machine in the domain will be a FreeIPA client, then add the FreeIPA server address to the DHCP configuration.
Run the client setup command.
[root@client ~]# ipa-client-install --enable-dns-updates
The --enable-dns-updates option updates DNS with the client machine's IP address. This option should only be used if the FreeIPA server was installed with integrated DNS or if the DNS server on the network accepts DNS entry updates with the GSS-TSIG protocol.
When using the --server option to specify the FreeIPA server to register with, the server name must be a fully-qualified domain name.
This must be a valid DNS name, which means only numbers, alphabetic characters, and hyphens (-) are allowed. Other characters, like underscores, in the hostname will cause DNS failures.
Other options for ipa-client-install are listed in the ipa-client-install manpage.
There is an --on-master option that is used as part of configuring an FreeIPA server (which also is an FreeIPA client, since it is within the domain). This option should never be used when configuring a regular FreeIPA client, because it results in slightly different client configuration which may not work on a non-FreeIPA server machine.
If prompted, enter the domain name for the FreeIPA DNS domain.
DNS discovery failed to determine your DNS domain
Please provide the domain name of your IPA server (ex: example.com): example.com
If prompted, enter the fully-qualified domain name of the FreeIPA server. Alternatively, use the --server option with the client installation script to supply the fully-qualified domain name of the FreeIPA server.
DNS discovery failed to find the IPA Server
Please provide your IPA server name (ex: ipa.example.com): ipaserver.example.com
This must be a valid DNS name, which means only numbers, alphabetic characters, and hyphens (-) are allowed. Other characters, like underscores, in the hostname will cause DNS failures.
The client script then prompts for a Kerberos identity to use to contact and then join the Kerberos realm. When these credentials are supplied, then the client is able to join the FreeIPA Kerberos domain and then complete the configuration:
Continue to configure the system with these values? [no]: yes
User authorized to enroll computers: admin
Synchronizing time with KDC...
Password for admin@EXAMPLE.COM:
Successfully retrieved CA cert
Subject: CN=Certificate Authority,O=EXAMPLE.COM
Issuer: CN=Certificate Authority,O=EXAMPLE.COM
Valid From: Tue Aug 13 09:29:07 2013 UTC
Valid Until: Sat Aug 13 09:29:07 2033 UTC
Enrolled in IPA realm EXAMPLE.COM
Created /etc/ipa/default.conf
New SSSD config will be created
Configured /etc/sssd/sssd.conf
Configured /etc/krb5.conf for IPA realm EXAMPLE.COM
Failed to update DNS records.
Adding SSH public key from /etc/ssh/ssh_host_rsa_key.pub
Adding SSH public key from /etc/ssh/ssh_host_dsa_key.pub
Could not update DNS SSHFP records.
SSSD enabled
Configured /etc/openldap/ldap.conf
NTP enabled
Configured /etc/ssh/ssh_config
Configured /etc/ssh/sshd_config
Client configuration complete.
Test that the client can connect successfully to the FreeIPA domain and can perform basic tasks. For example, check that the FreeIPA tools can be used to get user and group information:
$ id
$ getent passwd admin
$ getent group admins
3.4.1. Set up NFS to work with Kerberos.
To help troubleshoot potential NFS setup errors, enable debug information in the /etc/sysconfig/nfs file.
RPCGSSDARGS="-vvv"
RPCSVCGSSDARGS="-vvv"
Kerberos authentication requires that both client and server participating in an exchange have own Kerberos principals. NFS servers are configured by default to use a Kerberos principal named nfs/server's host name. NFS client software constructs NFS server's Kerberos principal by prepending nfs/ to the server's hostname. The client's Kerberos principal is not that important; as long as KDC permits issuing a ticket towards nfs/server.example.com, the client's Kerberos principal can be used to perform the authentication. For example, an NFS client can connect using host/client.example.com principal or as a specific user.
When NFS server machine nfs.example.com is enrolled to a FreeIPA domain as a FreeIPA client, a service named nfs/nfs.example.com has to be created by a FreeIPA administrator and the service key has to be extracted and stored on the NFS server.
Get credentials from Kerberos.
[root@server ~]#kinit admin
On a FreeIPA server, add an NFS service principal for the NFS client.
[root@server ~]# ipa service-add nfs/nfs.example.com@EXAMPLE
This command must be run on a machine with the ipa-admintools package installed so that the ipa command is available.
Next we need to obtain a keytab for the NFS service principal. The keytab contains long term Kerberos key which uniquely authenticates nfs/nfs.example.com to the KDC. This key must be protected or otherwise anyone who has access to the key may impersonate the NFS server. Ideally, NFS server is a FreeIPA client, thus it should already have ipa-getkeytab installed.
If NFS server has
ipa-getkeytab installed, it is safer to fetch the key directly from the NFS server. Given that FreeIPA-enrolled machine owns all services associated with it,
host/nfs.example.com principal may retrieve the NFS service keytab directly on the NFS server:
[root@nfs ~]# kinit -k host/`hostname`
[root@nfs ~]# ipa-getkeytab -s ipaserver.example.com -p nfs/`hostname`@EXAMPLE -k /etc/krb5.keytab
[root@nfs ~]# kdestroy
ipa-getkeytab utility makes sure to not overwrite other Kerberos keys in the keytab.
If
ipa-getkeytab is not available on the NFS server, the keys can be fetched up on the FreeIPA server and copied over to the NFS server:
[root@server ~]# ipa-getkeytab -s ipaserver.example.com -p nfs/nfs.example.com@EXAMPLE -k /tmp/krb5.keytab
[root@server ~]# chmod 0600 /tmp/krb5.keytab
[root@server ~]# scp -p /tmp/krb5.keytab root@nfs.example.com:/tmp/krb5.keytab
... switch over to the NFS server
[root@nfs ~]# ( echo rkt /tmp/krb5.keytab ; echo wkt /etc/krb5.keytab ) | ktutil
[root@nfs ~]# rm -f /tmp/krb5.keytab
It is recommended to use ktutil utility to transfer Kerberos keys between keytabs to avoid overwriting keys that are already placed in the keytab by other applications. For example, all FreeIPA clients have their host/client.example.com keys stored in /etc/krb5.keytab file. Simple overwrite of the file will render SSH authentication to the client impossible as SSH server will not be able to authenticate to the KDC using host/client.example.com key. The same can be said about users and groups as SSSD on FreeIPA clients is configured to authenticate with the help of the host key.
Some versions of the Linux NFS implementation have limited encryption type support. If the NFS server is hosted on a version older than Fedora 16, use the -e des-cbc-crc option to the ipa-getkeytab command for any nfs/<FQDN> service keytabs to set up, both on the server and on all clients. This instructs the KDC to generate only DES keys.
When using DES keys, all clients and servers that rely on this encryption type need to have the allow_weak_crypto option enabled in the [libdefaults] section of the /etc/krb5.conf file. Without these configuration changes, NFS clients and servers are unable to authenticate to each other, and attempts to mount NFS filesystems may fail. The client's rpc.gssd and the server's rpc.svcgssd daemons may log errors indicating that DES encryption types are not permitted.
Configure the /etc/exports file on the NFS server.
/ipashare gss/krb5p(rw,no_root_squash,subtree_check,fsid=0)
On the client, mount the NFS share.
Always specify the share as nfs.example.com:/share /mountpoint, using fully-qualified domain name of the NFS server.
Use the same -o sec setting as is used in the /etc/exports file for the NFS server.
[root@ipaclient ~]# mount -v -t nfs4 -o sec=krb5p nfs.example.com:/ipashare /mnt/ipashare
3.5. Manually Configuring a Linux Client
The ipa-client-install command automatically configures services like Kerberos, SSSD, PAM, and NSS. However, if the ipa-client-install command cannot be used on a system for some reason, then the FreeIPA client entries and the services can be configured manually.
Install SSSD 1.5.x or later, if it is not already installed.
Optional. Install the FreeIPA tools so that administrative tasks can be performed from the host.
[root@ipaclient ~]# yum install freeipa-admintools
Log in as FreeIPA administrator.
[user@server ~]$ kinit admin
On a FreeIPA server. Create a host entry for the client.
[user@server ~]$ ipa host-add --force --ip-address=192.168.166.31 ipaclient.example.com
On a FreeIPA server. Create keytabs for the client.
Set the client host to be managed by the server.
[user@server ~]$ ipa host-add-managedby --hosts=ipaserver.example.com ipaclient.example.com
Generate the keytab for the client.
[user@server ~]$ ipa-getkeytab -s ipaserver.example.com -p host/ipaclient.example.com -k /tmp/ipaclient.keytab
Copy the keytab to the client machine and rename it /etc/krb5.keytab.
If there is an existing /etc/krb5.keytab that should be preserved, the two files can be combined using ktutil.
Set the correct user permissions and, if necessary, SELinux contexts for the /etc/krb5.keytab file.
To verify permissions with SELinux context use ls -Z /etc/krb5.keytab.
Change permissions:
[root@ipaclient ~]# chown root:root /etc/krb5.keytab
[root@ipaclient ~]# chmod 0600 /etc/krb5.keytab
Change SELinux context (should be system_u:object_r:krb5_keytab_t:s0):
[root@ipaclient ~]# restorecon /etc/krb5.keytab
Configure SSSD by editing the /etc/sssd/sssd.conf file to point to the FreeIPA domain.
[sssd]
config_file_version = 2
services = nss, pam
domains = example.com
[nss]
[pam]
[domain/example.com]
cache_credentials = True
krb5_store_password_if_offline = True
ipa_domain = example.com
id_provider = ipa
auth_provider = ipa
access_provider = ipa
ipa_hostname = ipaclient.example.com
chpass_provider = ipa
ipa_server = ipaserver.example.com
ldap_tls_cacert = /etc/ipa/ca.crt
Configure NSS to use SSSD for passwords, groups, users, and netgroups.
vim /etc/nsswitch.conf
...
passwd: files sss
shadow: files sss
group: files sss
...
netgroup: files sss
...
Configure the /etc/krb5.conf file to point to the FreeIPA KDC.
[logging]
default = FILE:/var/log/krb5libs.log
kdc = FILE:/var/log/krb5kdc.log
admin_server = FILE:/var/log/kadmind.log
[libdefaults]
default_realm = EXAMPLE.COM
dns_lookup_realm = false
dns_lookup_kdc = false
rdns = false
ticket_lifetime = 24h
forwardable = yes
allow_weak_crypto = true
[realms]
EXAMPLE.COM = {
kdc = ipaserver.example.com:88
admin_server = ipaserver.example.com:749
default_domain = example.com
}
[domain_realm]
.example.com = EXAMPLE.COM
example.com = EXAMPLE.COM
Update the /etc/pam.d configuration to use the pam_sss.so modules.
For /etc/pam.d/fingerprint-auth:
...
account [default=bad success=ok user_unknown=ignore] pam_sss.so
...
session optional pam_sss.so
For /etc/pam.d/system-auth:
...
auth sufficient pam_sss.so use_first_pass
...
account [default=bad success=ok user_unknown=ignore] pam_sss.so
...
password sufficient pam_sss.so use_authtok
...
session optional pam_sss.so
For /etc/pam.d/password-auth:
...
auth sufficient pam_sss.so use_first_pass
...
account [default=bad success=ok user_unknown=ignore] pam_sss.so
...
password sufficient pam_sss.so use_authtok
...
session optional pam_sss.so
For /etc/pam.d/smartcard-auth:
...
account [default=bad success=ok user_unknown=ignore] pam_sss.so
...
session optional pam_sss.so
Install the FreeIPA server's CA certificate.
Obtain the certificate from the server.
[root@ipaclient ~]# wget -O /etc/ipa/ca.crt http://ipa.example.com/ipa/config/ca.crt
Install the certificate in the system's NSS database.
[root@ipaclient ~]# certutil -A -d /etc/pki/nssdb -n "IPA CA" -t CT,C,C -a -i /etc/ipa/ca.crt
Set up a host certificate for the host in FreeIPA.
Make sure certmonger is running.
[root@ipaclient ~]# service certmonger start
Configure chkconfig so that the certmonger service starts by default.
[root@ipaclient ~]# chkconfig certmonger on
$ ipa-getcert request -d /etc/pki/nssdb -n Server-Cert -K HOST/ipaclient.example.com -N 'CN=ipaclient.example.com,O=EXAMPLE.COM'
If administrative tools were not installed on the client, then the certificate can be generated on a FreeIPA server, copied over to the host, and installed using certutil.
3.6. Setting up a Linux Client Through Kickstart
A kickstart enrollment automatically adds a new system to the FreeIPA domain at the time it is provisioned.
This requires pre-creating the hosts on the FreeIPA server, with a predefined password that can be used to authenticate to complete the enrollment operation.
Create the host entry on the FreeIPA server and set a temporary Kerberos password for the entry.
When the ipa-client-install script is run normally (interactively), it prompts for authentication credentials to access the FreeIPA domain. However, when the script is run automatically, the system has to have some way to access the FreeIPA domain without using an existing FreeIPA user; this is done by setting the host principal in the script and using a Kerberos password (configured for the host account) to access the FreeIPA domain.
For example:
[jsmith@ipaserver ~]$ ipa host-add kickstart-server.example.com --password=secret
The password expires after the first authentication attempt. After enrollment completes, the host is authenticated using its keytab.
Include the ipa-client package with the other install packages.
%packages
@ X Window System
@ Desktop
@ Sound and Video
ipa-client
...
Create a post-install instruction that runs the ipa-client-install script, passes all the required information to access and configure the FreeIPA domain services, and specifies the pre-set password. Use the --unattended option to instruct the script to run non-interactively.
%post --log=/root/ks-post.log
# Get the hostname to set as the host principal
/bin/hostname > /tmp/hostname.txt
# Run the client install script
/usr/sbin/ipa-client-install --domain=EXAMPLEDOMAIN --enable-dns-updates --mkhomedir -w secret --realm=EXAMPLEREALM --server=ipaserver.example.com --unattended
Run the kickstart script.
3.7. Configuring a Microsoft Windows System to Join the FreeIPA Realm
Download the MIT Kerberos 3.x package for Windows.
http://web.mit.edu/kerberos/dist/index.html
Run the kfw-3.x-exe file to launch the MIT Kerberos Installation Wizard.
Read and accept the license agreement.
Install the KfW client. All other components are optional.
Accept the default destination path.
Select Download from web path, and enter the URL to the FreeIPA server. For example:
http://ipaserver.example.com/ipa/config/
Include the trailing backslash, or the configuration will fail.
Select Autostart the Network Identity Manager each time you login to Windows.
Click Install to begin the installation. When the installation is complete, click Finish to exit the Wizard.
Edit the hosts file and add the FreeIPA server. For example:
1.2.3.4 ipaserver.example.com ipaserver
Depending on the version of Windows, the HOSTS file could be located in different directories. For Windows XP and later systems, this is in C:\WINDOWS\system32\drivers\etc\.
One potential problem is that a ticket is not generated by Kerberos on Windows. Windows can use multiple ticket caches with MIT Kerberos. This can create odd scenarios, where it is possible to authenticate against FreeIPA's domain in the command line, but not to open the web UI.
3.8. Configuring a Solaris System as a FreeIPA Client
3.8.1. Configuring Solaris 10
FreeIPA provides an example profile for configuring Solaris 10 as a FreeIPA client. This can be loaded using
ldapclient and the
init command:
[root@solaris ~]# ldapclient init ipa.example.com
The ldapclient can also be run to enter the information for the FreeIPA domain manually:
[root@solaris ~]# ldapclient manual
-a credentialLevel=proxy
-a authenticationMethod=tls:simple
-a defaultSearchBase=dc=example,dc=com
-a domainName=example.com
-a defaultServerList=192.168.0.1
-a proxyDN=cn=proxyagent,ou=profile,dc=example,dc=com
-a proxyPassword={NS1}fbc123a92116812
-a attributeMap=group:memberuid=memberUid
-a attributeMap=group:gidnumber=gidNumber
-a attributeMap=passwd:gidnumber=gidNumber
-a attributeMap=passwd:uidnumber=uidNumber
-a attributeMap=passwd:homedirectory=homeDirectory
-a attributeMap=passwd:loginshell=loginShell
-a attributeMap=shadow:userpassword=userPassword
-a objectClassMap=group:posixGroup=posixgroup
-a objectClassMap=passwd:posixAccount=posixaccount
-a objectClassMap=shadow:shadowAccount=posixaccount
-a serviceSearchDescriptor=passwd:cn=users,cn=accounts,dc=example,dc=com
-a serviceSearchDescriptor=group:cn=groups,cn=accounts,dc=example,dc=com
-a serviceSearchDescriptor=netgroup:cn=sysaccounts,cn=etc,dc=example,dc=com
-a serviceSearchDescriptor=shadow:cn=sysaccounts,cn=etc,dc=example,dc=com
-a serviceSearchDescriptor=sudoers:cn=sysaccounts,cn=etc,dc=example,dc=com
Create a Solaris profile in the FreeIPA Directory Server instance for the Solaris domain clients to use. The LDAP entry should reflect the configuration that was passed to the Solaris machine in the ldapclient command.
[root@ipaserver ~]# ldapadd -h 192.168.0.1 -p 389 -D "cn=directory manager" -w secret
dn: cn=solaris,ou=profile,dc=example,dc=com
objectClass: top
objectClass: DUAConfigProfile
cn: solaris
credentialLevel: proxy
authenticationMethod: tls:simple
defaultSearchBase: dc=example,dc=com
defaultServerList: 192.168.0.1
objectclassMap: group:posixGroup=posixgroup
objectclassMap: passwd:posixAccount=posixaccount
objectclassMap: shadow:shadowAccount=posixAccount
serviceSearchDescriptor: passwd:cn=users,cn=accounts,dc=example,dc=com
serviceSearchDescriptor: group:cn=groups,cn=accounts,dc=example,dc=com
serviceSearchDescriptor: shadow:cn=sysaccounts,cn=etc,dc=example,dc=com
serviceSearchDescriptor: netgroup:cn=sysaccounts,cn=etc,dc=example,dc=com
serviceSearchDescriptor: sudoers:cn=sysaccounts,cn=etc,dc=example,dc=com
bindTimeLimit: 10
profileTTL: 43200
searchTimeLimit: 30
defaultSearchScope: one
followReferrals: FALSE
Create the cn=proxyagent account in the FreeIPA Directory Server instance.
[root@ipaserver ~]# ldapadd -h 192.168.0.1 -p 389 -D "cn=directory manager" -w secret
dn: cn=proxyagent,ou=profile,dc=example,dc=com
objectClass: top
objectClass: person
sn: proxyagent
cn: proxyagent
userPassword:: e1NTSEF9Mm53KytGeU81Z1dka1FLNUZlaDdXOHJkK093TEppY2NjRmt6Wnc9PQ=
On the FreeIPA server, use the certutil command to create cert8.db and key3.db databases.
[root@ipaserver ~]# certutil -N -d .
Then, copy the database over to the Solaris machine in the /var/ldap directory. For example:
[root@ipaserver ~]# scp cert8.db solaris.example.com:/var/ldap
[root@ipaserver ~]# scp key3.db solaris.example.com:/var/ldap
Remove the ldap option from all entries in /etc/nsswitch.conf except for the passwd, group, shadow, netgroup, and sudoers entries.
Configure and enable NTP and synchronize the time between the client and the FreeIPA server.
[root@solaris ~]# ntpdate ipaserver.example.com
Configure the Kerberos client. The Kerberos configuration includes specifying the realm and domain details and default ticket attributes.
[root@solaris ~]# vim /etc/krb5/krb5.conf
[libdefaults]
default_realm = EXAMPLE.COM
verify_ap_req_nofail = false
[realms]
EXAMPLE.COM = {
kdc = ipaserver.example.com
admin_server = ipaserver.example.com
}
[domain_realm]
example.com = EXAMPLE.COM
.example.com = EXAMPLE.COM
[logging]
default = FILE:/var/krb5/kdc.log
kdc = FILE:/var/krb5/kdc.log
[appdefaults]
kinit = {
renewable = true
forwardable= true
}
The default file created by ldapclient configures forwardable tickets by default, which makes it possible to connect to the UI from any system and provides a way to audit administration operations.
Configure PAM to use Kerberos authentication. For example:
[root@solaris ~]# vim /etc/pam.conf
login auth requisite pam_authtok_get.so.1
login auth required pam_dhkeys.so.1
login auth sufficient pam_krb5.so.1 try_first_pass
login auth required pam_unix_auth.so.1
login auth required pam_dial_auth.so.1
other auth requisite pam_authtok_get.so.1
other auth required pam_dhkeys.so.1
other auth required pam_unix_cred.so.1
other auth sufficient pam_krb5.so.1
other auth required pam_unix_auth.so.1
other account requisite pam_roles.so.1
other account required pam_unix_account.so.1
other account required pam_krb5.so.1
other password requisite pam_authtok_check.so.1 force_check
other password sufficient pam_krb5.so.1
other password required pam_authtok_store.so.1
Configure NFS to work with the Kerberos domain.
Add an NFS service principal for the client.
[root@ipaserver ~]# ipa service-add nfs/client.example.com
Create the NFS keytab file.
[root@ipaserver ~]# ipa-getkeytab -s ipaserver.example.com -p nfs/client.example.com -k /tmp/krb5.keytab -e des-cbc-crc
Copy the keytab from the server to the client.
[root@ipaserver ~]# scp /tmp/krb5.keytab root@client.example.com:/tmp/krb5.keytab
On the FreeIPA client, use the
ktutil command to import the contents into the main host keytab.
ktutil: read_kt /tmp/krb5.keytab
ktutil: write_kt /etc/krb5/krb5.keytab
ktutil: q
Verify that the NFS service keytab was created:
[root@solaris ~]# klist -ket /etc/krb5/krb5.keytab
Verify that the NFS server is accessible:
[root@solaris ~]# showmount -e ipaserver.example.com
Make sure that this line is uncommented in the
/etc/nfssec.conf file.
krb5 390003 kerberos_v5 default -
Mount the NFS share.
[root@solaris ~]# mount -t nfs4 ipaserver.example.com:/ /mnt/ -o sec=krb5
Configure sudo on the Solaris machine to work with the FreeIPA server.
If necessary, install the required packages for SASL, OpenSSL, sudo and LDAP, and BerkeleyDB:
CSWbdb4 (BerkeleyDB 4)
CSWcommon
CSWlibnet
CSWoldaprt
CSWossl
CSWossldevel
CSWosslrt
CSWosslutils
CSWsasl
CSWsudo-common
CSWsudoldap
Edit the OpenLDAP ldap.conf file to use the secure URL for the FreeIPA Directory Server instance and to use the FreeIPA CA certificate.
[root@solaris ~]# vim /opt/csw/etc/openldap/ldap.conf
base dc=example,dc=com
timelimit 120
bind_timelimit 120
idle_timelimit 3600
uri ldaps://ipaserver.example.com
ssl start_tls
sudoers_base ou=SUDOers,dc=example,dc=com
ssl on
TLS_REQCERT allow
TLS_CACERT /etc/openldap/cacerts/ca.crt
TLS_CACERTFILE /etc/openldap/cacerts/ca.crt
TLS_CACERTDIR /etc/openldap/cacerts
...
Download the FreeIPA CA certificate:
http://ipaserver.example.com/ipa/config.ca.cert
Copy the FreeIPA CA certificate to the /etc/openldap/cacerts directory.
3.8.2. Configuring Solaris 9
Configure the NFS client.
Configure the /etc/exports file on the NFS server.
/nfs client.example.com(sec=krb5p,rw,sync,fsid=0,no_subtree_check)
Add an NFS service principal for the client.
[root@ipaserver ~]# ipa service-add nfs/client.example.com
Create the NFS keytab file.
[root@ipaserver ~]# ipa-getkeytab -s ipaserver.example.com -p nfs/client.example.com -k /tmp/krb5.keytab -e des-cbc-crc
Copy the keytab from the server to the client.
[root@ipaserver ~]# scp /tmp/krb5.keytab root@client.example.com:/tmp/krb5.keytab
Make sure that this line is uncommented in the
/etc/nfssec.conf file.
krb5 390005 kerberos_v5 default -
Obtain a ticket for the NFS client.
[root@solaris ~]# kinit -k nfs/client.example.com
Mount the NFS share.
[root@solaris ~]# mount -F nfs -o sec=krb5p ipaserver.example.com:/nfs /mnt/
On the FreeIPA client, use the
ktutil command to import the contents into the main host keytab.
ktutil: read_kt /tmp/krb5.keytab
ktutil: write_kt /etc/krb5/krb5.keytab
ktutil: q
3.9. Configuring an HP-UX System as a FreeIPA Client
The FreeIPA client installation process requires that a FreeIPA server already exist.
Configure and enable NTP and make sure that time is synchronized between the client and the FreeIPA server.
3.9.2. Configuring LDAP Authentication
Install the ldapux client.
[root@hp-server ~]# swinstall -s /path/to/J4269AA_B.04.15.01_HP-UX_B.11.23_IA_PA.depot
Change to the configuration directory, and run the setup script.
[root@hp-server ~]# cd /opt/ldapux/config/
Running the setup script is only necessary for the first HP-UX client. Every subsequent HP-UX client only needs to know where the LDAP profile is stored. All clients will then use the same configuration.
The setup script prompts for information about the FreeIPA LDAP service, such as its port and host, Directory Manager credentials, and schema and directory suffixes.
Would you like to continue with the setup? [Yes]
Select which Directory Server you want to connect to ? [RedHat Directory]
Directory server host ? [ipaserver.example.com]
Directory Server port number [389]
Would you like to extend the printer schema in this directory server? [No]
Would you like to install PublicKey schema in this directory server? [No]
Would you like to install the new automount schema ? [No]
Profile Entry DN: [cn=ldapuxprofile,cn=etc,dc=example,dc=com]
User DN [cn=Directory Manager]
Password ? [Directory Manager's Password]
Authentication method ? [ SIMPLE ]
Enter the number of the hosts you want to specify [1]
Default Base DN ? [dc=example,dc=com]
Accept remaining defaults ? [n]
Client binding [Anonymous]
Bind time limit [5 seconds]
Search time limit [no limit]
Do you want client searches of the directory to follow referrals? [Yes]
Profile TTL [0 = infinite]
Do you want to remap any of the standard RFC 2307 attribute? [Yes]
Specify the service you want to map? [ 3 ]
[ group ]
Specify the attribute you want to map [3 for memberuid ]
Type the name of the attribute memberuid should be mapped to [member]
Specify the service you want to map? [ 0 = exit ]
Do you want to remap any of the standard RFC 2307 attribute? [ no this time ]
Do you want to create custom search descriptors? [ No ]
Ensure that the LDAP client daemon is running.
If necessary, start the daemon:
Check that the user and group entries in the LDAP client are correct and available:
Create a new group on the FreeIPA server.
Add a test user to the new group.
Validate the new user and group:
To ensure that the LDAP client daemon starts when the system boots, add the following lines to the
/etc/opt/ldapux/ldapclientd.conf file:
[StartOnBoot]
enable=yes
3.9.3. Configuring Kerberos
Edit the /etc/krb5.conf file to reflect the Kerberos domain used by the FreeIPA server. Setting up the Kerberos configuration includes specifying the realm and domain details, and default ticket attributes. Forwardable tickets are configured by default, which facilitates connection to the administration interface from any operating system, and also provides for auditing of administration operations. For example:
[libdefaults]
default_realm = EXAMPLE.COM
default_keytab_name = FILE:/etc/krb5.keytab
default_tkt_enctypes = des3-cbc-sha1 arcfour-hmac aes256-cts des-cbc-md5 des-cbc-crc
default_tgs_enctypes = des3-cbc-sha1 arcfour-hmac aes256-cts des-cbc-md5 des-cbc-crc
ccache_type = 2
[realms]
EXAMPLE.COM = {
kpasswd_server = ipaserver.example.com
kdc = ipaserver.example.com:88
admin_server = ipaserver.example.com:749
default_domain = example.com
}
[domain_realm]
.example.com = EXAMPLE.COM
example.com = EXAMPLE.COM
[appdefaults]
kinit = {
forwardable = true
}
The PAM configuration differs slightly between different versions of HP-UX.
Edit the /etc/pam.conf file so that all of the required modules are loaded for authentication. For example:
#
# PAM configuration
#
# This pam.conf file is intended as an example only.
# see pam.conf(4) for more details
# Authentication management
#
login auth required libpam_hpsec.so.1
login auth sufficient libpam_krb5.so.1
login auth required libpam_unix.so.1 try_first_pass
su auth required libpam_hpsec.so.1
su auth sufficient libpam_krb5.so.1
su auth required libpam_unix.so.1 try_first_pass
dtlogin auth required libpam_hpsec.so.1
dtlogin auth sufficient libpam_krb5.so.1
dtlogin auth required libpam_unix.so.1 try_first_pass
dtaction auth required libpam_hpsec.so.1
dtaction auth sufficient libpam_krb5.so.1
dtaction auth required libpam_unix.so.1 try_first_pass
ftp auth required libpam_hpsec.so.1
ftp auth sufficient libpam_krb5.so.1
ftp auth required libpam_unix.so.1 try_first_pass
sshd auth required libpam_hpsec.so.1
sshd auth sufficient libpam_krb5.so.1
sshd auth required libpam_unix.so.1 try_first_pass
OTHER auth required libpam_unix.so.1
#
# Account management
#
login account required libpam_hpsec.so.1
login account sufficient libpam_krb5.so.1
login account required libpam_unix.so.1
su account required libpam_hpsec.so.1
su account sufficient libpam_krb5.so.1
su account required libpam_unix.so.1
dtlogin account required libpam_hpsec.so.1
dtlogin account sufficient libpam_krb5.so.1
dtlogin account required libpam_unix.so.1
dtaction account required libpam_hpsec.so.1
dtaction account sufficient libpam_krb5.so.1
dtaction account required libpam_unix.so.1
ftp account required libpam_hpsec.so.1
ftp account sufficient libpam_krb5.so.1
ftp account required libpam_unix.so.1
sshd account required libpam_hpsec.so.1
sshd account sufficient libpam_krb5.so.1
sshd account required libpam_unix.so.1
OTHER account required libpam_unix.so.1
#
# Session management
#
login session required libpam_hpsec.so.1
login session sufficient libpam_krb5.so.1
login session required libpam_unix.so.1
dtlogin session required libpam_hpsec.so.1
dtlogin session sufficient libpam_krb5.so.1
dtlogin session required libpam_unix.so.1
dtaction session required libpam_hpsec.so.1
dtaction session sufficient libpam_krb5.so.1
dtaction session required libpam_unix.so.1
sshd session required libpam_hpsec.so.1
sshd session sufficient libpam_krb5.so.1
sshd session required libpam_unix.so.1
OTHER session required libpam_unix.so.1
#
# Password management
#
login password required libpam_hpsec.so.1
login password sufficient libpam_krb5.so.1
login password required libpam_unix.so.1
passwd password required libpam_hpsec.so.1
passwd password sufficient libpam_krb5.so.1
passwd password required libpam_unix.so.1
dtlogin password required libpam_hpsec.so.1
dtlogin password sufficient libpam_krb5.so.1
dtlogin password required libpam_unix.so.1
dtaction password required libpam_hpsec.so.1
dtaction password sufficient libpam_krb5.so.1
dtaction password required libpam_unix.so.1
OTHER password required libpam_unix.so.1
Edit the /etc/pam.conf file to reflect the following example:
#
# PAM configuration
#
# This pam.conf file is intended as an example only.
# see pam.conf(4) for more details
#
# Authentication management
#
login auth sufficient /usr/lib/security/libpam_krb5.1
login auth required /usr/lib/security/libpam_unix.1 try_first_pass
su auth sufficient /usr/lib/security/libpam_krb5.1
su auth required /usr/lib/security/libpam_unix.1 try_first_pass
dtlogin auth sufficient /usr/lib/security/libpam_krb5.1
dtlogin auth required /usr/lib/security/libpam_unix.1 try_first_pass
dtaction auth sufficient /usr/lib/security/libpam_krb5.1
dtaction auth required /usr/lib/security/libpam_unix.1 try_first_pass
ftp auth sufficient /usr/lib/security/libpam_krb5.1
ftp auth required /usr/lib/security/libpam_unix.1 try_first_pass
OTHER auth required /usr/lib/security/libpam_unix.1
#
# Account management
#
login account sufficient /usr/lib/security/libpam_krb5.1
login account required /usr/lib/security/libpam_unix.1
su account sufficient /usr/lib/security/libpam_krb5.1
su account required /usr/lib/security/libpam_unix.1
dtlogin account sufficient /usr/lib/security/libpam_krb5.1
dtlogin account required /usr/lib/security/libpam_unix.1
dtaction account sufficient /usr/lib/security/libpam_krb5.1
dtaction account required /usr/lib/security/libpam_unix.1
ftp account sufficient /usr/lib/security/libpam_krb5.1
ftp account required /usr/lib/security/libpam_unix.1
OTHER account required /usr/lib/security/libpam_unix.1
#
# Session management
#
login session sufficient /usr/lib/security/libpam_krb5.1
login session required /usr/lib/security/libpam_unix.1
dtlogin session sufficient /usr/lib/security/libpam_krb5.1
dtlogin session required /usr/lib/security/libpam_unix.1
dtaction session sufficient /usr/lib/security/libpam_krb5.1
dtaction session required /usr/lib/security/libpam_unix.1
OTHER session required /usr/lib/security/libpam_unix.1
#
# Password management
#
login password sufficient /usr/lib/security/libpam_krb5.1
login password required /usr/lib/security/libpam_unix.1
passwd password sufficient /usr/lib/security/libpam_krb5.1
passwd password required /usr/lib/security/libpam_unix.1
dtlogin password sufficient /usr/lib/security/libpam_krb5.1
dtlogin password required /usr/lib/security/libpam_unix.1
dtaction password sufficient /usr/lib/security/libpam_krb5.1
dtaction password required /usr/lib/security/libpam_unix.1
OTHER password required /usr/lib/security/libpam_unix.1
Edit the /etc/opt/ssh/ssh_config file:
Remove the /etc/krb5.keytab file.
Set up the NFS/Kerberos mapping for the Solaris client on the FreeIPA server.
Add a host service principal for the HP-UX client.
Create the host keytab file.
Copy this keytab to the HP-UX machine, and save it as /etc/krb5/krb5.keytab.
3.9.6. Configuring Access Control
Example 3.1. pam_authz.policy File: Allow User Access, Deny Admin Access
This configuration in /etc/opt/ldapux/pam_authz.policy prevents the admin user from logging in while still allowing regular users to log in.
# pam_authz.policy.template:
#
# An example file that could be copied over to /etc/opt/ldapux/pam_authz.policy.
# pam_authz.policy is a local policy file that PAM_AUTHZ would use to help
# determine which users would be allowed to login to the local host.
#
# In this template file, by default, the only active access rule is
# "allow:unix_local_user"
# All the local users are authorized to login.
#
# The policy file contains one or more access rule. The format of an access
# rule is <action>:<type>:<object>
#
# where <action> could be "deny", "allow", "status"
# "PAM_SUCCESS", "PAM_PERM_DENIED", "PAM_MAXTRIES"
# "PAM_AUTH_ERR", "PAM_NEW_AUTHTOK_REQD",
# "PAM_AUTHTOKEN_REQD, "PAM_CRED_INSUFFICIENT",
# "PAM_AUTHINFO_UNAVAIL", "PAM_USER_UNKNOWN"
# "PAM_ACCT_EXPIRED", "PAM_AUTHOK_EXPIRED"
#
# Note: "status" must use along with "rhds" or
# "ads" <type>.
# <type> could be "unix_user", "unix_local_user", "unix_group",
# "netgroup", ldap_filter", "ldap_group"
# "rhds" or "ads"
#
# Note: When <type> is set to "rhds" or "ads",
# the <action> filed must set to "status".
# <object> contains search information. For example,
#
deny:unix_group:admins
allow:unix_local_user
3.9.7. Testing the Configuration
By default, the admin user is given /bin/bash as the shell to use and /home/admin as the home directory. It may be necessary to install bash to be able to log in.
There are two quick ways to check the Kerberos and PAM configuration for the HP client:
Authenticate as an administrator on a Linux box that is a client in the FreeIPA domain, and then attempt to SSH into the HP machine. The admin user should be able to log in using SSH without being asked for a password.
Log into the FreeIPA web UI using the administrator credentials on the HP machine.
3.10. Configuring an AIX System as a FreeIPA Client
Make sure that all of these packages are installed on the AIX machine before beginning the client configuration:
v5.3 OS
v5.3 Updates
krb5 client packages
openssh
wget
bash
krb5 server
ldap.client
openssl
modcrypt.base (for gssd)
Configure and enable NTP and make sure that time is synchronized between the client and the FreeIPA server.
3.10.2. Configuring the AIX Client
Setting up an AIX client requires setting up the client to work in the FreeIPA Kerberos domain and, optionally, to enable SSH authentication to the AIX client using FreeIPA credentials.
Kerberos configuration includes specifying the realm and domain details, and default ticket attributes. Forwardable tickets are configured by default, which facilitates connection to the administration interface from any operating system, and also provides for auditing of administration operations. For example:
Configure the krb5 client settings to use the FreeIPA Kerberos domain:
Get a Kerberos ticket:
On the FreeIPA server, add a user that is only used for authentication. (This can be substituted with krb5 authentication if that works from the LDAP client). Otherwise go to the FreeIPA server and use ldapmodify, bind as Directory Manager and create this user. The user should be assigned a shared password.
# ldapmodify -D "cn=directory manager" -w secret -p 389 -h ipaserver.example.com -x -a
dn: uid=nss,cn=sysaccounts,cn=etc,dc=example,dc=com
objectClass: account
objectClass: simplesecurityobject
objectClass: top
uid: nss
userPassword: secretpassword
On the AIX system, configure the LDAP client settings to use the FreeIPA directory services:
In the /etc/security/ldap directory, create user and group map files:
For example, for the FreeIPAuser.map file:
#FreeIPAuser.map file
keyobjectclass SEC_CHAR posixaccount s
# The following attributes are required by AIX to be functional
username SEC_CHAR uid s
id SEC_INT uidnumber s
pgrp SEC_CHAR gidnumber s
home SEC_CHAR homedirectory s
shell SEC_CHAR loginshell s
gecos SEC_CHAR gecos s
spassword SEC_CHAR userpassword s
lastupdate SEC_INT shadowlastchange s
For example, for the FreeIPAgroup.map file:
#FreeIPAgroup.map file
groupname SEC_CHAR cn s
id SEC_INT gidNumber s
users SEC_LIST member m
Modify the /etc/security/ldap/ldap.cfg file to set the REALM and base DN values for the FreeIPA domain.
userbasedn:cn=users,cn=accounts,dc=example,dc=com
groupbasedn:cn=groups,cn=accounts,dc=example,dc=com
userattrmappath:/etc/security/ldap/FreeIPAuser.map
groupattrmappath:/etc/security/ldap/FreeIPAgroup.map
userclasses:posixaccount
Start the LDAP client daemon:
Test the LDAP client connection to the FreeIPA server:
Add the following sections to the
/usr/lib/security/methods.cfg file to configure the system login to use Kerberos and LDAP:
KRB5A:
program = /usr/lib/security/KRB5A
program_64 = /usr/lib/security/KRB5A_64
options = authonly
LDAP:
program = /usr/lib/security/LDAP
program_64 =/usr/lib/security/LDAP64
KRB5ALDAP:
options = auth=KRB5A,db=LDAP
Edit the
/etc/security/user file, and modify the
default section to use the Kerberos/LDAP system and the LDAP user registry.
SYSTEM = "KRB5ALDAP"
registry = LDAP
To test the Kerberos configuration, log in as a FreeIPA user and verify that the user and group information is correct:
$ id
Optionally, configure the FreeIPA client to accept incoming SSH requests and authenticate with the user's Kerberos credentials.
Set the SSH syslog configuration:
auth.info /var/log/sshd.log
auth.info /var/log/sshd.log
auth.crit /var/log/sshd.log
auth.warn /var/log/sshd.log
auth.notice /var/log/sshd.log
auth.err /var/log/sshd.log
Set the SSH logging configuration:
SyslogFacility AUTH
LogLevel INFO
Configure sshd to use GSS-API, including disabling DNS for GSS-API:
vim /etc/ssh/sshd_config
# GSSAPI options
GSSAPIAuthentication yes
#GSSAPICleanupCredentials yes
Restart the sshd daemon:
Restart the syslogd daemon:
Add the client to the FreeIPA server's Kerberos configuration.
Add a host service principal for the client.
Retrieve the host keytab.
Copy the keytab from the server to the client.
On the FreeIPA client, use the ktutil command to import the contents into the main host keytab.
# ktutil
ktutil: read_kt /tmp/krb5.keytab
ktutil: write_kt /etc/krb5/krb5.keytab
ktutil: q
On the FreeIPA server, get a ticket for the admin user.
To test the SSH configuration, try to log in as the admin user using SSH without providing a password.
By default, the admin user is given /bin/bash as the shell to use and /home/admin as the home directory. It may be necessary to install bash to be able to log in.
3.11. Troubleshooting Client Installations
For clients configured using
ipa-client-install, the client installation log is located in
/var/log/ipaclient-install.log. The FreeIPA logs, both for the server and client and for FreeIPA-associated services, are covered in
Section 19.1.3, “Checking FreeIPA Server Logs”.
These are some issues and workarounds for client installation problems.
3.11.1. The client can't resolve reverse hostnames when using an external DNS.
While FreeIPA can host its own DNS server as part of the domain services, it can also use external DNS name server. However, because of some of the limitations of reverse DNS, there can be problems with resolving reverse lookups if the external DNS is listed in the client's /etc/resolv.conf file or if there are other resources on the network with SRV records, like Active Directory.
The problem is that the external DNS name server returns the wrong hostname for the FreeIPA server.
One way this exhibits is errors with finding the FreeIPA server in the Kerberos database:
Jun 30 11:11:48 server1 krb5kdc[1279](info): AS_REQ (4 etypes {18 17 16 23}) 192.168.60.135: NEEDED_PREAUTH: admin EXAMPLE COM for krbtgt/EXAMPLE COM EXAMPLE COM, Additional pre-authentication required
Jun 30 11:11:48 server1 krb5kdc[1279](info): AS_REQ (4 etypes {18 17 16 23}) 192.168.60.135: ISSUE: authtime 1309425108, etypes {rep=18 tkt=18 ses=18}, admin EXAMPLE COM for krbtgt/EXAMPLE COM EXAMPLE COM
Jun 30 11:11:49 server1 krb5kdc[1279](info): TGS_REQ (4 etypes {18 17 16 23}) 192.168.60.135: UNKNOWN_SERVER: authtime 0, admin EXAMPLE COM for HTTP/server1.wrong.example.com@EXAMPLE.COM, Server not found in Kerberos database
There are several ways to work around this issue:
Edit the /etc/resolv.conf file to remove the external DNS name server references.
Add reverse lookup records for each FreeIPA server.
Give the FreeIPA client or domain a subnet and forward all requests for that subnet.
3.11.2. The client is not added to the DNS zone.
If a client is in a subnet not controlled by a FreeIPA DNS server, then the nsupdate command may fail to add the client to the DNS zone when ipa-client-install runs.
[jsmith@ipaserver ~]$ kinit admin
[jsmith@ipaserver ~]$ ipa dnsrecord-add ipaclient.example.com www --a-rec 1.2.3.4
If the DNS domain is managed outside of FreeIPA, the resource record can be added manually to the zone configuration. For information on DNS in Fedora, see
the DNS chapter in the Deployment Guide.
3.12. Uninstalling a FreeIPA Client
For Fedora clients, the ipa-client-install utility can be used to uninstall the client and remove it from the FreeIPA domain. To remove the client, use the --uninstall option.
There is an uninstall option with the ipa-join command. This is called by ipa-client-install --uninstall as part of the uninstallation process. However, while the ipa-join option removes the client from the domain, it does not actually uninstall the client or properly remove all of the FreeIPA-related configuration. Do not run ipa-join -u to attempt to uninstall the FreeIPA client. The only way to uninstall a client completely is to use ipa-client-install --uninstall.
All of the access to FreeIPA, both through the web UI and through the command line, is done by a user authenticating to the FreeIPA domain. This chapter covers the basics of setting up browsers to handle Kerberos authentication, logging into FreeIPA, and troubleshooting some common connection issues.
4.1. About the FreeIPA Client Tools
FreeIPA creates a domain of recognized services, host machines, and users with universally-applied authentication sources and common policies. From the perspective of a client machine and a FreeIPA user, the domain itself is fairly transparent after the initial configuration. All users need to do is log into the domain using Kerberos, and that's it.
However, an administrator has two ongoing tasks: add principals to the FreeIPA Kerberos domain and set the domain policies and server configuration that govern domain interactions. FreeIPA has both command-line and web-based interfaces for administrators to use to manage the domain, services, and FreeIPA entries.
4.1.1. About the FreeIPA Command-Line Tools
The most common method to maintain the domain is using the command-line tools. FreeIPA has an incredibly broad set of scripts and commands that are available to administrators. The entry management functions of the domain are carried out with a single script: ipa. This script is a parent or control script for associated subcommands; each subcommand relates to a specific entry type.
The command-line scripts offer a number of benefits:
The scripts allow management tasks to be automated and performed repeatedly in a consistent way without manual intervention.
Entries can be added with all possible attributes configured (or a desired subset of attributes) in a single step. The web UI frequently requires two steps to fully configure an entry: the first to create the entry and the next to add optional attributes.
The command-line scripts support adding additional attributes which may not be available in the UI or even custom attributes to entries, if the schema is configured.
4.1.1.1. The Structure of the ipa Command
The ipa command is essentially a big plug-in container. It supports dozens of subcommands; these subcommands are actually plug-ins which manage specific types of objects in FreeIPA.
The first type of a subcommand identifies the object type (such as user, sudo, group, host, or dns), and the second part identifies the operation being performed on that object.
ipa objectType-operation objectName --option=value
For example, adding a user is done using the user-add subcommand:
ipa user-add entryName options
Related subcommands are grouped together into plug-in modules. Commands for managing DNS entries like dnszone-add and dnsrecord-add all belong to the dns module or topic. All of the information for managing a specific area, with all of the supported commands and examples for each, are available by viewing the help for that topic:
ipa help topic
To get a list of all available topics, run the help command with topics param:
ipa help topics
To get a list of all available commands, run the help command with commands param:
ipa help commands
To get more information about help run command without a topic name:
ipa help
All topic or command areas follow a consistent pattern for how entries are managed.
4.1.1.1.1. Logging into the FreeIPA Domain Before Running
Before running any FreeIPA commands (with the exception of the installation scripts, such as ipa-server-install), the user must first authenticate to the FreeIPA domain by obtaining a Kerberos ticket. This is done using kinit:
[user@ipaserver ~]$ kinit admin
4.1.1.1.2. Adding, Editing, and Deleting Entries with ipa
New entries are added using an *-add command. For example:
[user@ipaserver ~]$ ipa user-add jsmith
[user@ipaserver ~]$ ipa user-add
First name: John
Last name: Smith
User login [jsmith]: jsmith
--------------------
Added user "jsmith"
--------------------
...
Likewise, entries are usually edited through a *-mod commands, and then any new or edited attributes are listed as options after it.
$ ipa user-mod jsmith --title="Editor III"
Last, entries can be deleted using the *-del command and the entry's name.
[user@ipaserver ~]$ ipa user-del jsmith
4.1.1.1.3. Finding and Displaying Entries with ipa
Entries for an entire type are searched for using the *-find command and an optional search criterion. The criterion is a string.
[user@ipaserver ~]$ ipa user-find jsmith
With no search criterion, every entry of that type is displayed.
Searches (any
*-find command) have certain limits imposed as part of the server configuration, specifically how many entries are returned (size limits) and how long a search will run (time limits). This is covered in
Section 4.4.2, “Setting FreeIPA Search Limits”. Part of the server configuration is setting global defaults for size and time limits on searches. While these limits are always enforced in the web UI, they can be overridden with any
*-find command with the
--sizelimit and
--timelimit options. For example, if the default time limit is 60 seconds and a search is going to take longer, the time limit can be increased to 120 seconds:
[user@ipaserver ~]$ ipa user-find jsmith --timelimit=120
Not every possible attribute in an entry type can be searched for. A certain subset of attributes are predefined and indexed for searches. (This list is configurable for users and groups, but not for other types of entries.)
When entries are returned, only certain default attributes are displayed with the entry; to return all attributes currently set for entries, use the --all option.
To display a specific entry, use the *-show command and the entry name. As with searches, only a subset of attributes are displayed with the entry unless the --all option is used.
[user@ipaserver ~]$ ipa user-show jsmith
User login: jsmith
First name: John
Last name: Smith
Home directory: /home/jsmith
Login shell: /bin/sh
Email address: jsmith@example.com
UID: 1035400001
GID: 1035400001
Account disabled: False
Password: True
Member of groups: ipausers
Kerberos keys available: True
4.1.1.1.4. Adding Members to Groups and Containers with ipa
Group members are added and removed with separate commands, apart from simply modifying an entry. Member commands essentially create a relationship between different FreeIPA entries. While this is obvious in traditional group-member roles, it is also true for some policy entries (like SELinux and sudo policies) where entries are associated with another entry.
Most commonly, the command format for adding a member entry is *-add-member, although the command may specify an entry type, such as *-add-user.
Likewise, entries are removed as members (not deleted) using a *-remove-member or *-remove-type command.
4.1.1.2. Positional Elements in ipa Commands
Usually, ipa subcommands have only two elements: the name of the entry being modified (the object) and then any options available for the subcommand:
ipa command entryName --options=values
With a few types of entries, however, not only the entry name itself needs to be specified; the entry's parent must also be specified. This is the case with automount commands, for example. With automount, the location must be included whenever a new key or map is created.
The parent entry name is given first, and then the child entry name. For example, for automount, the location is given first, and then the map or key entry name.
ipa command parentEntryName chidlEntryName --childOptions=childValues
4.1.1.3. Managing Entry Attributes with --setattr, --addattr, and --delattr
All identities and configuration in FreeIPA are stored as LDAP entries, with standard attribute-value assertions (AVAs). Whether an entry is created through the UI or the CLI, there are certain attributes which are required and others which are available, depending on the default and custom object classes for that entry type.
For the most common attributes, the ipa use specified command-line arguments to set values. For example, adding a mail attribute to a user can be done with the --mail argument; enabling dynamic updates for a DNS zone can be done with the --allow-dynupdate option with zone commands; and a map key for an automount map is given in the --key option.
However, entries can also allow attributes that may not have command-line (or UI) options for setting them. Partially, this is because the underlying LDAP schema is very rich, particularly for user entries, with many possible allowed attributes. Additionally, FreeIPA allows schema extensions for users and groups, and those custom schema elements are not necessarily reflected in the UI or command-line tools.
Any supported attribute can be added or edited to an entry using the --setattr and --addattr options.
Both options have this format:
--setattr=attribute=value
The --setattr option sets one value for the given attribute; any existing values are overwritten, even for multi-valued attributes.
The --addattr option adds a new value for an attribute; for a multi-valued attribute, it adds the new value while preserving any existing values.
Both --setattr option and --addattr can be used multiple times in the same command invocation. For example:
[user@ipaserver ~]$ ipa user-mod jsmith --addattr=mail=johnnys@me.com --addattr=mail=jsmith@example.com --setattr=description="backup IT manager for the east coast branch"
Likewise, an attribute or specific attribute value can be removed from an entry using the --delattr option. For a single-valued attribute, this removes the attribute; for a multi-valued attribute, it removes only the specified value. For example:
[user@ipaserver ~]$ ipa user-mod jsmith --delattr=mail=johnnys@me.com
Options are evaluated in order: --setattr, --addattr and --delattr. If the same attribute is added and deleted in the same modify operation, it is a no-op.
[user@ipaserver ~]$ ipa user-mod jsmith --addattr=mail=johnnys@me.com --delattr=mail=johnnys@me.com
ipa: ERROR: no modifications to be performed
4.1.1.4. Using Special Characters with FreeIPA Tools
The FreeIPA command-line tools are run as any other utilities in a shell. If there are special characters in the command — such as angle brackets (> and <), ampersands (&), asterisks (*), and pipes (|) — the characters must be escaped. Otherwise, the command fails because the shell cannot properly parse the unescaped characters.
4.1.2. Looking at the FreeIPA UI
The FreeIPA web UI is designed for simplicity. This was the primary design goal, and this means that the web UI offers benefits that make using FreeIPA simpler and clearer:
It shows instant, visual relationships between entries (such as a user and all the groups, sudo rules, netgroups, and policies which are associated with that user).
All entries are listed immediately without having to run a search. This makes it possible to browse entries. The UI also has a simple search box which quickly filters the list of entries.
The interface is intuitive to use, without having to learn the command-line tools.
The web UI can be accessed from machines outside the FreeIPA domain, so the domain can be managed from anywhere.
Using the web UI requires a valid Kerberos ticket for the FreeIPA domain (by default), meaning that it can only be accessed from a machine within the FreeIPA domain. Alternatively, the web UI can be configured to allow password authentication along with Kerberos authentication, or a machine outside the FreeIPA can be configured to work with Kerberos (
Section 4.3.4, “Using a Browser on Another System”).
The web UI has three major functional areas which correspond to each of the major functions of FreeIPA: identity management, policy management, and domain configuration.
Table 4.1. Configuration Areas Per Tab
|
Main Menu Tab
|
Configuration Areas
|
|---|
|
Identity
|
User entries
User groups entries
Host/client entries
Host group entries
Netgroups entries
Domain services entries
DNS (if configured)
Certificates entries
Realm domain entries
|
|
Policy
|
|
|
IPA Server (access controls within FreeIPA)
|
Role-based access control (permissions based on group membership)
Self service permissions
Delegations (user access control over other users)
ID ranges
Trusts (shown if ipa-adtrust-install was run)
Configuration
|
The
main menu at the top of every page has three tabs which correspond to the functional areas listed in
Table 4.1, “Configuration Areas Per Tab”. When a tab is selected, there is a submenu of the different configuration areas. Some configuration areas may have multiple possible entries; for example, role-based access controls define user roles/groups, the areas that access can be granted or denied (privileges), and then the permissions granted to those areas. Each separate configuration entry has its own task area beneath the primary configuration area.
All entries for a configuration area are listed together on the main page for that area. This page provides direct links to individual entry pages, as well as basic information (the attributes) about the entry. (This is usually just the description, but user entries show a lot more information.)
The page also has some tasks that can be performed on it. For a list page that shows entries, this can be creating or deleting an entry. For a list page for groups, the tasks are for establishing relationships between entities, either by adding (enrolling) or removing an entity from that group. Both individual entries and groups can be searched for through the list page.
Each entry page is a form which allows that entry to be edited. This is done by editing text fields or by selecting items from drop-down menus.
The web UI uses common elements on all pages.
The most basic is that all blue text is a link to an entry or to an action.
When a task like adding an entry or saving a change is possible, the task link is blue. When it is not possible (such as no items have been selected to be deleted) then the task is grayed out.
All list pages display direct links to entry pages. However, some entries are essentially nested. For example, in automount configuration, the primary entry is the location, and then keys, mount points, and maps are associated with that location as children entries. This hierarchy is reflected in breadcrumb navigation near the top of the page, so it is easy to identify where you are in the UI and how this entry relates to any other related entries.
Most entries have a variety of different configuration areas. A simple user entry has account activity settings, personal information, address information, organizational information, and other contact information. Related attributes are grouped together logically in the UI. These entry form areas can be collapsed or expanded using the arrows to control the amount of information displayed on the page.
When entries are created, they are added with only the required attributes. Additional attributes can be added manually. Some attributes have default values added to the entry and simply need to be edited; other attributes may not exist at all in the new entry and need to be added.
Any changes to any attribute can be undone. A single attribute change can be undone by clicking the dynamic undo button; all changes can be undone by clicking the Reset link at the top of the entry details page.
4.1.2.3. Showing and Changing Group Members
Members can be added to a group through the group configuration. There are tabs for all the member types which can belong to the group, and an administrator picks all of the marching entries and adds them as members.
However, it is also possible for an entity to be added to a group through its own configuration. Each entry has a list of tabs that displays group types that the entry can join. The list of all groups of that type are displayed, and the entity can be added to multiple groups at the same time.
4.2. Logging into FreeIPA
Users are authenticated to FreeIPA services, including the command-line tools and the web UI, using Kerberos authentication. This means that logging into FreeIPA requires running kinit.
Running kinit issues the user a Kerberos ticket. This ticket is checked by any FreeIPA or Kerberos-aware service, so that a user only needs to log in once to access all domain services. Domain services include the FreeIPA web UI, mounted file shares, wikis, or any other application which uses FreeIPA as its identity/authentication store.
4.2.1. Logging into FreeIPA
Logging into FreeIPA requires running kinit on a client within the FreeIPA domain.
[user@ipaserver ~]$ kinit
The kinit command must be run from a machine which has been configured as a client within the FreeIPA domain, so that the client retrieves authenticates with the FreeIPA KDC.
Simply running kinit logs into FreeIPA as the currently logged-in user account. This user account must also be an FreeIPA user for them to authenticate to the FreeIPA Kerberos domain successfully. For example, if you are logged into the machine as jsmith:
[jsmith@ipaserver ~]$ kinit
Password for jsmith@EXAMPLE.COM:
If SSSD or pam_krb5 is configured on the FreeIPA client machine, then when a user logs into the machine, a ticket is created which can be used for machine services which require authentication, such as sudo.
4.2.2. Logging in When an FreeIPA User Is Different Than the System User
To specify an FreeIPA username — because a person's system username is different then the FreeIPA username or to switch FreeIPA user accounts — simply rerun the kinit command, specifying the new user. For example:
[user@ipaserver ~]$ kinit userName
Password for userName@EXAMPLE.COM:
When the server was first set up, an administrative user, admin, is created to perform normal administrative activities. To authenticate as the admin user, use the name admin when running kinit:
[user@ipaserver ~]$ kinit admin
Only one set of tickets can be stored per logged-in user. The current stored credentials are the ones that will be used when accessing FreeIPA services.
If you were already connected to the FreeIPA web UI as another user, refresh the browser to display the updated details for the new user.
4.2.3. Checking the Current Logged in User
Use the klist command to verify the identity and the ticket granting ticket (TGT) from the server:
[user@ipaserver ~]$ klist
Ticket cache: FILE:/tmp/krb5cc_500
Default principal: ipaUser@EXAMPLE.COM
Valid starting Expires Service principal
11/10/08 15:35:45 11/11/08 15:35:45 krbtgt/EXAMPLE.COM@EXAMPLE.COM
Kerberos 4 ticket cache: /tmp/tkt500
klist: You have no tickets cached
It's important to know who the authenticated user is because the currently-authenticated user is the only one who can access the FreeIPA services. The Kerberos client libraries for kinit have some limitation, one of them being that the current ticket is overwritten with any new invocation of kinit. Authenticating as User A and then authenticating as User B overwrites User A's ticket.
To allow there to be multiple authenticated users on a machine, set the KRB5CCNAME environment variable. This variable keeps credential caches separate in different shells.
4.2.4. Caching User Kerberos Tickets
Only one set of tickets can be stored per logged-in user. The current stored credentials are the ones that will be used when accessing FreeIPA services.
For example, if you authenticated as admin, added a new user, set the password, and then tried to authenticate as that user, the administrator's ticket is lost.
To keep separate credential caches in different shells, a special environment variable, KRB5CCNAME, can be used.
4.3. Using the FreeIPA Web UI
In order to use the web UI, the user must be authenticated with the FreeIPA Kerberos domain and have an active Kerberos ticket (
Section 4.2, “Logging into FreeIPA”). Generally, the web UI can only be accessed from a FreeIPA server or client machine and the user must be locally authenticated. There are a couple of ways to work around this, either by configuring Kerberos on a non-domain machine to connect to the Kerberos domain (
Section 4.3.4, “Using a Browser on Another System”) or by password authentication to the UI.
4.3.1. Supported Web Browsers
These browsers are supported for connecting to the web UI:
Firefox 15.x and newer
Firefox 10.x
Firefox 3.6
4.3.2. Opening the FreeIPA Web UI
To open the web UI:
Open the FreeIPA URL. The full URL is https://IPAserver-FQDN/ipa/ui, but this service is also accessed simply by opening https://IPAserver-FQDN. For example:
https://server.example.com
https://server.example.com/ipa/ui
4.3.3. Configuring the Browser
Firefox can use Kerberos credentials to authenticate to the FreeIPA UI, but Kerberos negotiation needs to be configured to use the FreeIPA domain. At the first log-in attempt, if Firefox has not been configured to support Kerberos authentication, then an error message appears with form based authentication, as described in
Section 4.3.5, “Form based authentication”.
If you want to configure Firefox to use Kerberos credentials, then the FreeIPA web UI can perform the required configuration:
Click the configured link.
Click the Firefox configuration page link.
Click the button to import the CA certificate for the FreeIPA server.
Set the web site and software developer (first and last) trust bits for the CA certificate.
Click the button to install the Kerberos Configuration Firefox extension.
Click the Configure Firefox button. This automatically fills out all the negotiate settings in the Firefox configuration to use the FreeIPA domain settings.
When the process is complete, a text shows saying that Firefox has been successfully configured.
Click the Return to Web UI button to log in.
This can also be done manually:
Open Firefox.
Type about:config in the address bar.
In the Search field, type negotiate to filter out the Kerberos-related parameters.
On Fedora, enter the domain name for the URI parameters, including the preceding period (.) and set the
gsslib parameter to true:
network.negotiate-auth.trusted-uris .example.com
network.negotiate-auth.using-native-gsslib true
On Windows, set the trusted URIs and library path, and disable the built-in Microsoft Kerberos for authentication:
network.negotiate-auth.trusted-uris .example.com
network.auth.use-sspi false
network.negotiate-auth.gsslib: C:\Program Files\MIT\Kerberos\bin\gssapi32.dll
On a 64-bit system, the library location is in C:\Program Files(x86)\MIT\Kerberos\bin\gssapi32.dll.
Open the web UI by going to the fully-qualified domain name of the FreeIPA server such as http://ipaserver.example.com. Make sure that you can open the web UI and that there are no Kerberos authentication errors.
Next, download the FreeIPA server's CA certificate from http://ipa.example.com/ipa/config/ca.crt.
Select the first (Trust this CA to identify web sites) and third (Trust this CA to identify software developers) check boxes.
4.3.4. Using a Browser on Another System
It is possible to connect to the FreeIPA web UI from a system which is not a member of the FreeIPA domain. In this case, it is possible to specify a FreeIPA-specific Kerberos configuration file on the external (non-FreeIPA) machine before running kinit, and then the user can authenticate against the FreeIPA server domain.
This is especially useful there are multiple realms or overlapping domains across your infrastructure.
Copy the /etc/krb5.conf file from the FreeIPA server.
# scp /etc/krb5.conf root@externalmachine.example.com:/etc/krb5_ipa.conf
Do not overwrite the existing krb5.conf file.
On the external machine, set the terminal session to use the copied FreeIPA Kerberos configuration file:
$ export KRB5_CONFIG=/etc/krb5_ipa.conf
4.3.5. Form based authentication
Form based authentication for the UI allows users to log in even if there are problems with the Kerberos service or if the system is outside the FreeIPA domain.
When the FreeIPA server cannot find a valid Kerberos ticket for the user attempting to log into the web UI, it shows a login form. Preferred method of connecting to FreeIPA domain services (including the UI) is using Kerberos authentication.
Simply supply the UID and password for a configured FreeIPA user to log into the web UI.
4.3.6. Logging in the FreeIPA Web UI as Another User
To log into the FreeIPA web UI as another user:
Remove any Kerberos credentials.
kdestroy
Log out of any existing FreeIPA browser session.
Run the kinit command specifying the new user. For example:
kinit anotherUserName
Log back into the FreeIPA web UI.
4.3.7. Troubleshooting UI Connection Problems
If negotiate authentication is not working, turn on verbose logging for the authentication process to help diagnose the issue:
Close all browser windows.
In a terminal, set the new log levels for Firefox:
export NSPR_LOG_MODULES=negotiateauth:5
export NSPR_LOG_FILE=/tmp/moz.log
This enables verbose logging and logs all information to /tmp/moz.log.
Restart the browser from the same terminal window and attempt t .
Table 4.2. UI Error Log Messages
|
Error Log Message
|
Description and Fix
|
|---|
-1208550944[90039d0]: entering nsNegotiateAuth::GetNextToken()
-1208550944[90039d0]: gss_init_sec_context() failed: Miscellaneous failure
No credentials cache found
|
There are no Kerberos tickets. Run kinit.
|
-1208994096[8d683d8]: entering nsAuthGSSAPI::GetNextToken()
-1208994096[8d683d8]: gss_init_sec_context() failed: Miscellaneous failure
Server not found in Kerberos database
|
This can occur when you have successfully obtained Kerberos tickets but are still unable to authenticate to the UI. This indicates that there is a problem with the Kerberos configuration. The first place to check is the [domain_realm] section in the /etc/krb5.conf file. Make sure that the FreeIPA Kerberos domain entry is correct and matches the configuration in the Firefox negotiation parameters. For example:
.example.com = EXAMPLE.COM
example.com = EXAMPLE.COM
|
|
Nothing is in the log file.
|
It is possible that you are behind a proxy which is removing the HTTP headers required for negotiate authentication. Try to connect to the server using HTTPS instead, which allows the request to pass through unmodified. Then check the log file again.
|
4.4. Understanding Search Limits and Settings
Some searches can result in a large number of entries being returned, possibly even all entries. Search limits improve overall server performance by limiting how long the server spends in a search and how many entries are returned.
4.4.1. Types of Search Limits and Where They Apply
Search limits have a dual purpose to improve server performance by reducing the search load and to improve usability by returning a smaller — and therefore easier to browse — set of entries.
The FreeIPA server has several different limits imposed on searches:
The search limit configuration for the FreeIPA server. This is a setting for the FreeIPA server itself, which is applied to all requests sent to the server from all FreeIPA clients, the FreeIPA CLI tools, and the FreeIPA web UI for normal page display.
By default, this limit is 100 entries.
The time limit configuration for the FreeIPA server. Much like the search size limit, the time limit sets a maximum amount of time that the FreeIPA server, itself, waits for searches to run. Once it reaches that limit, the server stops the search and returns whatever entries were returned in that time.
By default, this limit is two seconds.
The page size limit. Although not strictly a search limit, the page size limit does limit how many entries are returned per page. The server returns the set of entries, up to the search limit, and then selects up to 20 entries per page for display. Paging results makes the results more understandable and more viewable.
This is hard-coded to 20 for all searches.
The LDAP search limit (--pkey-only option). All searches performed in the UI, and CLI searches which use the --pkey-only option, override the search limit set in the FreeIPA server configuration and use the search limit set in the underlying LDAP directory.
By default, this limit is 2000 entries. It can be edited by editing the 389 Directory Server configuration.
4.4.2. Setting FreeIPA Search Limits
Search limits set caps on the number of records returned or the time spent searching when querying the database for user or group entries. There are two types of search limits: time limits and size (number) limits.
With the default settings, users are limited to two-second searches and no more than 100 records returned per search.
Setting search size or time limits too high can negatively affect FreeIPA server performance.
Open the IPA Server tab.
Select the Configuration subtab.
Scroll to the Search Options area.
Change the search limit settings.
Search size limit, the maximum number of records to return in a search.
Search time limit, the maximum amount of time, in seconds, to spend on a search before the server returns results.
Setting the time limit or size limit value to -1 means that there are no limits on searches.
When the changes are complete, click the Update link at the top of the Configuration page.
4.4.2.2. With the Command Line
The search limits can be changed using the config-mod command.
[user@ipaserver ~]$ ipa config-mod --searchtimelimit=5 --searchrecordslimit=500
Max. username length: 32
Home directory base: /home
Default shell: /bin/sh
Default users group: ipausers
Default e-mail domain for new users: example.com
Search time limit: 5
Search size limit: 50
User search fields: uid,givenname,sn,telephonenumber,ou,title
Group search fields: cn,description
Enable migration mode: FALSE
Certificate Subject base: O=EXAMPLE.COM
Password Expiration Notification (days): 4
Setting the time limit or size limit value to -1 means that there are no limits on searches.
4.4.3. Overriding the Search Defaults
Part of the server configuration is setting global defaults for size and time limits on searches. While these limits are always enforced in the web UI, they can be overridden with any *-find command run through the command line.
The --sizelimit and --timelimit options set alternative size and time limits, respectively, for that specific command run. The limits can be higher or lower, depending on the kinds of results you need.
For example, if the default time limit is 60 seconds and a search is going to take longer, the time limit can be increased to 120 seconds:
[user@ipaserver ~]$ ipa user-find jsmith --timelimit=120
4.4.4. Setting Search Attributes
A search for users or groups does not automatically search every possible attribute for that attribute. Rather, it searches a specific subset of attributes, and that list is configurable.
When adding attributes to the user or group search fields, make sure that there is a corresponding index within the LDAP directory for that attribute. Searches are performed based on indexes. Most standard LDAP attributes have indexes, but any custom attributes must have indexes created for them. Creating indexes is described in the
indexes chapter in the Directory Server Administrator's Guide.
4.4.4.1. Setting User Search Attributes
4.4.4.1.1. From the Web UI
Open the IPA Server tab.
Select the Configuration subtab.
Scroll to the User Options area.
Add any additional search attributes, in a comma-separated list, in the User search fields field.
When the changes are complete, click the Update link at the top of the Configuration page.
4.4.4.1.2. From the Web UI
To change the search attributes, use the --usersearch option to set the attributes for user searches.
[user@ipaserver ~]$ ipa config-mod --usersearch=uid,givenname,sn,telephonenumber,ou,title
Always give the complete list of search attributes. Whatever values are passed with the configuration argument overwrite the previous settings.
4.4.4.2. Setting Group Search Attributes
A search for users or groups does not automatically search every possible attribute for that attribute. Rather, it searches a specific subset of attributes, and that list is configurable.
When adding attributes to the user or group search fields, make sure that there is a corresponding index within the LDAP directory for that attribute. Searches are performed based on indexes. Most standard LDAP attributes have indexes, but any custom attributes must have indexes created for them. Creating indexes is described in the
indexes chapter in the Directory Server Administrator's Guide.
4.4.4.2.1. From the Web UI
Open the IPA Server tab.
Select the Configuration subtab.
Scroll to the Group Options area.
Add any additional search attributes, in a comma-separated list, in the Group search fields field.
When the changes are complete, click the Update link at the top of the Configuration page.
4.4.4.2.2. From the Command Line
To change the search attributes, use the --groupsearch options to set the attributes for group searches.
[user@ipaserver ~]$ ipa config-mod --groupsearch=cn,description
Always give the complete list of search attributes. Whatever values are passed with the configuration argument overwrite the previous settings.
4.4.5. Attributes Returned in Search Results
Searches can be performed on attributes that are not displayed in the UI. This means that entries can be returned in a search that do not appear to match the given filter. This is especially common if the search information is very short, which increases the likelihood of a match.
Chapter 5. Identity: Managing Users and User Groups
Users in FreeIPA are able to access services and servers within the domain through Kerberos authentication. This chapter covers general management tasks for users, groups, password policies, and other configuration for users.
5.1. Setting up User Home Directories
A home directory is required for any FreeIPA user. Without a home directory in the expected location, a user may be unable to log into the domain. While systems administrators can manage home directories outside of FreeIPA, it is also possible to use a PAM module to create home directories automatically on both FreeIPA servers and clients.
5.1.1. About Home Directories
FreeIPA, as part of managing users, can manage user home directories. However, FreeIPA has certain defined parameters for any managed home directories:
The default prefix for users' home directories is /home.
FreeIPA does not automatically create home directories when users log in. Automatically creating home directories requires either the
pam_oddjob_mkhomedir module or the
pam_mkhomedir module. This module can be configured as part of client installation or after installation, as described in
Section 5.1.2, “Enabling the PAM Home Directory Module”.
The home directory process for FreeIPA first attempts to use the pam_oddjob_mkhomedir module because this requires fewer user privileges and access to create the home directories, as well as integrating smoothly with SELinux. If this module is not available, then the process falls back to the pam_mkhomedir module.
On Red Hat Enterprise Linux 5 clients, the client installation script uses the pam_mkhomedir module even if the pam_oddjob_mkhomedir module is available. To use the pam_oddjob_mkhomedir module on Red Hat Enterprise Linux 5, edit the PAM configuration manually.
It is possible to use an NFS file server that provides /home that can be made available to all machines in the domain and then automounted on the FreeIPA server.
There are potential issues when using NFS, such as security issues related to granting root access to the NFS user, performance issues with loading the entire /home tree, and network performance issues for using remote servers for home directories. There are some general guidelines for using NFS with FreeIPA:
Use automount to mount only the user's home directory and only when the user logs in, rather than loading the entire /home tree.
Use a remote user who has limited permissions to create home directories and mount the share on the FreeIPA server as that user. Since the FreeIPA server runs as an httpd process, it is possible to use sudo or a similar program to grant limited access to the FreeIPA server to create home directories on the NFS server.
Use a mechanism, such as the pam_oddjob_mkhomedir module, to create the home directory as that user.
If a suitable directory and mechanism are not available for to create home directories, users may not be able to log in.
5.1.2. Enabling the PAM Home Directory Module
For a home directory to be created automatically when a user logs in, FreeIPA can use either the pam_oddjob_mkhomedir module or the pam_mkhomedir module. Because it requires fewer permissions and works well with SELinux, FreeIPA preferentially uses the pam_oddjob_mkhomedir module. If that module is not installed, then it falls back to the pam_mkhomedir module.
FreeIPA does not require the pam_oddjob_mkhomedir module or pam_mkhomedir module. This is because the *_mkhomedir module may try to create home directories even when the shared storage is not available. If the module is unable to create the home directory, then users can be blocked from logging into the FreeIPA domain.
The system administrator must activate this module on each client or server as needed.
There are two ways to enable the pam_oddjob_mkhomedir (or pam_mkhomedir) module:
The --mkhomedir option can be used with the ipa-client-install command. While this is possible for clients, this option is not available to servers when they are set up.
The pam_oddjob_mkhomedir module can be enabled using the system's authconfig command. For example:
authconfig --enablemkhomedir --update
This option can be used for both server and client machines post-installation.
On Red Hat Enterprise Linux 5 clients, the client installation script uses the pam_mkhomedir module even if the pam_oddjob_mkhomedir module is available. To use the pam_oddjob_mkhomedir module on Red Hat Enterprise Linux 5, edit the PAM configuration manually.
5.1.3. Manually Mounting Home Directories
While PAM modules can be used to create home directories for users automatically, this may not be desirable behavior in every environment. In that case, home directories can be manually added to the FreeIPA server from separate locations using NFS shares and automount.
Create a new location for the user directory maps:
[user@ipaserver ~]$ ipa automountlocation-add userdirs
Location: userdirs
Add a direct map to the new location's auto.direct file. In this example, the mount point is /share:
[user@ipaserver ~]$ ipa automountkey-add userdirs auto.direct --key=/share --info="-ro,soft, ipaserver.example.com:/home/share"
Key: /share
Mount information: -ro,soft, ipaserver.example.com:/home/share
5.2. Managing User Entries
5.2.1. About Username Formats
The default length for usernames is 32 characters.
FreeIPA supports a wide range of username formats, based on this regular expression:
[a-zA-Z0-9_.][a-zA-Z0-9_.-]{0,252}[a-zA-Z0-9_.$-]?
The trailing $ symbol is permitted for Samba 3.x machine support.
Any system limits — such as starting a username with a number on Unix systems — apply to the usernames in FreeIPA.
Open the Identity tab, and select the Users subtab.
Click the Add link at the top of the users list.
Fill in the user's first and last names. The user login (UID) is automatically generated based on the user's full name, but this can be set manually by clicking the Optional field link.
Click the
Add and Edit button to go directly to the expanded entry page and fill in more attribute information, as in
Section 5.2.3.1, “From the Web UI”. The user entry is created with some basic information already filled in, based on the given user information and the user entry template.
5.2.2.2. From the Command Line
ipa user-add [username] [attributes]
When no arguments are used, the command prompts for the required user account information and uses the defaults for the other attributes, with the defaults printed below. For example:
[user@ipaserver ~]$ ipa user-add
First name: John
Last name: Smith
User login [jsmith]: jsmith
--------------------
Added user "jsmith"
--------------------
User login: jsmith
First name: John
Last name: Smith
Full name: John Smith
Display name: John Smith
Initials: JS
Home directory: /home/jsmith
GECOS: John Smith
Login shell: /bin/sh
Kerberos principal: jsmith@EXAMPLE.COM
Email address: jsmith@example.com
UID: 882600007
GID: 882600007
Password: False
Member of groups: ipausers
Kerberos keys available: False
Any of the user attributes can be passed with the command. This will either set values for optional attributes or override the default values for default attributes.
[user@ipaserver ~]$ ipa user-add jsmith --first=John --last=Smith --manager=bjensen --email=johnls@example.com --homedir=/home/work/johns --password
When a user is created without specifying a UID or GID number, then the user account is automatically assigned an ID number that is next available in the server or replica range. (Number ranges are described more in
Section 5.8, “Managing Unique UID and GID Number Assignments”.) This means that a user always has a unique number for its UID number and, if configured, for its private group.
If a number is manually assigned to a user entry, the server does not validate that the uidNumber is unique. It will allow duplicate IDs; this is expected (though discouraged) behavior for POSIX entries.
If two entries are assigned the same ID number, only the first entry is returned in a search for that ID number. However, both entries will be returned in searches for other attributes or with ipa user-find --all.
Open the Identity tab, and select the Users subtab.
Click the name of the user to edit.
There are a number of different types of attributes that can be edited for the user. All of the default attributes are listed in
Table 5.2, “Default FreeIPA User Attributes”. Most of the attributes in the
Identity Settings and
Account Settings areas have default values filled in for them, based on the user information or on the user entry template.
Edit the fields or, if necessary, click the Add link by an attribute to create the attribute on the entry.
When the edits are done, click the Update link at the top of the page.
5.2.3.2. From the Command Line
The user-mod command edits user accounts by adding or changing attributes. At its most basic, the user-mod specifies the user account by login ID, the attribute to edit, and the new value:
ipa user-mod loginID --attributeName=newValue
For example, to change a user's work title from Editor II to Editor III:
[user@ipaserver ~]$ ipa user-mod jsmith --title="Editor III"
FreeIPA allows multi-valued attributes, based on attributes in LDAP that are allowed to have multiple values. For example, a person may have two email addresses, one for work and one for personal, that are both stored in the mail attribute. Managing multi-valued attributes can be done using the --addattr option.
If an attribute allows multiple values — like mail — simply using the command-line argument will overwrite the value with the new value. This is also true for using --setattr. However, using --addattr will add a new attribute; for a multi-valued attribute, it adds the new value in addition to any existing values.
Example 5.1. Multiple Mail Attributes
A user is created first using his work email account.
[user@ipaserver ~]$ ipa user-add jsmith --first=John --last=Smith --email=johnls@example.com
Then, his personal email account is added.
[user@ipaserver ~]$ ipa user-mod jsmith --addattr=mail=johnnys@me.com
Both email addresses are listed for the user.
[user@ipaserver ~]$ ipa user-find jsmith --all
--------------
1 user matched
--------------
dn: uid=jsmith,cn=users,cn=accounts,dc=example,dc=com
User login: jsmith
.....
Email address: jsmith@example.com, jsmith@new.com
To set two values at the same time, use the --addattr option twice:
[user@ipaserver ~]$ ipa user-add jsmith --first=John --last=Smith --email=johnls@example.com --addattr=mail=johnnys@me.com --addattr=mail=admin@example.com
5.2.4. Activating and Deactivating User Accounts
User accounts can be deactivated. A deactivated user cannot log into FreeIPA or its related services (like Kerberos) and he cannot perform any tasks. However, the user account still exists within FreeIPA and all of the associated information remains unchanged.
Any existing connections remain valid until the Kerberos TGT and other tickets expire. Once the ticket expires, the user cannot renew the ticket.
Open the Identity tab, and select the Users subtab.
Activation or deactivation user(s):
Directly from the users list:
Select the checkboxes by the names of the users to activate or deactivate.
Click the Disable/Enable link at the top of the task area.
Confirm operation.
From the user view:
Click the name of the user for whom to deactivate or activate.
Choose the Disable/Enable action.
Click the Apply button at the top of the page.
5.2.4.2. From the Command Line
Users are activated and disabled using user-enable and user-disable commands. All that is required is the user login. For example:
[user@ipaserver ~]$ ipa user-disable jsmith
Deleting a user account permanently removes the user entry and all its information from FreeIPA, including group memberships and passwords. External configuration — like a system account and home directory — will still exist on any server or local machine where they were created, but they cannot be accessed through FreeIPA.
Deleting a user account is permanent. The information cannot be recovered; a new account must be created.
The ipa user-del and ipa group-remove-member commands prevent the accidential deletion of the last user in the admins group.
However, if all users from the admins group are removed in some way, you can use the Directory Manager account to add another user to the group:
ldapmodify -x -D 'cn=directory manager' -W
dn: cn=admins,cn=groups,cn=accounts,dc=example,dc=com
changetype: modify
add: member
member: uid=youruser,cn=users,cn=accouns,dc=example,dc=com
Once you have done this, you may use this account to re-create the admin user.
Open the Identity tab, and select the Users subtab.
Select the checkboxes by the names of the users to delete.
Click the Delete link at the top of the task area.
When prompted, confirm the delete action.
5.2.5.2. From the Command Line
Users are deleted using the user-del command and then the user login. For example, a single user:
[user@ipaserver ~]$ ipa user-del jsmith
To delete multiple users, simply list the users, separated by spaces.
[user@ipaserver ~]$ ipa user-del jsmith bjensen mreynolds cdickens
When deleting multiple users, use the --continue option to force the command to continue regardless of errors. A summary of the successful and failed operations is printed to stdout when the command completes. If --continue is not used, then the command proceeds with deleting users until it encounters an error, and then it exits.
5.3. Managing Public SSH Keys for Users
OpenSSH uses public-private key pairs to authenticate users. A user attempts to access some network resource and presents its key pair. The first time the user authenticates, the administrator on the target machine has to approve the request manually. The machine then stores the user's public key in an authorized_keys file. Any time that the user attempts to access the resource again, the machine simply checks its authorized_keys file and then grants access automatically to approved users.
There are a couple of problems with this system:
SSH keys have to be distributed manually and separately to all machines in an environment.
Administrators have to approve user keys to add them to the configuration, but it is difficult to verify either the user or key issuer properly, which can create security problems.
On Fedora, the System Security Services Daemon (SSSD) can be configured to cache and retrieve user SSH keys so that applications and services only have to look in one location for user keys. Because SSSD can use FreeIPA as one of its identity information providers, FreeIPA provides a universal and centralized repository of keys. Administrators do not need to worry about distributing, updating, or verifying user SSH keys.
5.3.1. About the SSH Key Format
When keys are uploaded to the FreeIPA entry, the key format can be either an
OpenSSH-style key or a raw
RFC 4253-style blob. Any RFC 4253-style key is automatically converted into an OpenSSH-style key before it is imported and saved into the FreeIPA LDAP server.
The FreeIPA server can identify the type of key, such as an RSA or DSA key, from the uploaded key blob. However, in a key file such as id_rsa.pub, a key entry is identified by its type, then the key itself, and then an additional comment or identifier. For example, for an RSA key associated with a specific hostname:
"ssh-rsa ABCD1234...== ipaclient.example.com"
All three parts from the key file can be uploaded to and viewed for the user entry, or only the key itself can be uploaded.
5.3.2. Uploading User SSH Keys Through the Web UI
Generate a user key. For example, using the OpenSSH tools:
[user@ipaserver ~]$ ssh-keygen -t rsa -C jsmith@example.com
Generating public/private rsa key pair.
Enter file in which to save the key (/home/jsmith/.ssh/id_rsa):
Created directory '/home/jsmith/.ssh'.
Enter passphrase (empty for no passphrase):
Enter same passphrase again:
Your identification has been saved in /home/jsmith/.ssh/id_rsa.
Your public key has been saved in /home/jsmith/.ssh/id_rsa.pub.
The key fingerprint is:
a5:fd:ac:d3:9b:39:29:d0:ab:0e:9a:44:d1:78:9c:f2 jsmith@example.com
The key's randomart image is:
+--[ RSA 2048]----+
| |
| + . |
| + = . |
| = + |
| . E S.. |
| . . .o |
| . . . oo. |
| . o . +.+o |
| o .o..o+o |
+-----------------+
Copy the public key from the key file. The full key entry has the form type key== comment. Only the key== is required, but the entire entry can be stored.
[user@ipaserver ~]$ cat /home/jsmith/.ssh/id_rsa.pub
ssh-rsa AAAAB3NzaC1yc2E...tJG1PK2Mq++wQ== jsmith@example.com
Open the Identity tab, and select the Users subtab.
Click the name of the user to edit.
In the Account Settings area of the Settings tab, click the SSH public keys: Add link.
The UI opens a new link, New: key not set Show/Set key. Click the Show/Set key link.
Paste in the public key for the user, and click the Set button.
The SSH public keys field now shows New: key set. Clicking the Show/Set key link opens the submitted key.
To upload multiple keys, click the Add link below the list of public keys, and upload the other keys.
When all the keys have been submitted, click the Update link at the top of the user's page to save the changes.
When the public key is saved, the entry is displayed as the key fingerprint, the comment (if one was included), and the key type
.
After uploading the user keys, configure SSSD to use FreeIPA as one of its identity domains and set up OpenSSH to use the SSSD tooling for managing user keys. This is covered in the
SSSD and OpenSSH Integration.
5.3.3. Uploading User SSH Keys Through the Command Line
The --sshpubkey option uploads the 64 bit-encoded public key to the user entry. For example:
[user@server ~]$ ipa user-mod jsmith --sshpubkey="ssh-rsa 12345abcde= ipaclient.example.com"
With a real key, the key is longer and usually ends with an equals sign (=).
To upload multiple keys:
[user@server ~]$ ipa user-mod jsmith --sshpubkey="12345abcde==" --sshpubkey="key2==" --sshpubkey="key3=="
After uploading the user keys, configure SSSD to use FreeIPA as one of its identity domains and set up OpenSSH to use the SSSD tooling for managing user keys. This is covered in the
SSSD and OpenSSH Integration.
5.3.4. Deleting User Keys
Open the Identity tab, and select the Users subtab.
Click the name of the user to edit.
Open the Account Settings area of the Settings tab.
Click the Delete link by the fingerprint of the key to remove.
Click the Update link at the top of the user's page to save the changes.
The command-line tools can be used to remove all keys. This is done by running ipa user-mod with the --sshpubkey= set to a blank value; this removes all public keys for the user. For example:
[user@server ~]$ kinit admin
[user@server ~]$ ipa user-mod --sshpubkey= jsmith
Regular, non-administrative users can change only their personal passwords, and all passwords are constrained by the FreeIPA password policies.
This allows administrators to create intro passwords or to reset passwords easily, while still keeping the final password confidential. Since any password sent by an administrator to the user is temporary, there is little security risk.
Changing a password as the FreeIPA admin user overrides any FreeIPA password policies, but the password expires immediately. This requires the user to change the password at the next login. Similarly, any user who has password change rights can change a password and no password policies are applied, but the other user must reset the password at the next log in.
Changing a password as the LDAP Directory Manager user, using LDAP tools, overrides any FreeIPA password policies.
Open the Identity tab, and select the Users subtab.
Click the name of the user for whom to reset the password. All users can change their own password; only administrators or users with delegated permissions can change other user's passwords.
Scroll to the Account Settings area.
Click the Reset Password link.
In the pop-up box, enter and confirm the new password.
5.4.2. From the Command Line
Changing a password — your own or another user's — is done using the user-mod command, as with other user account changes.
[user@ipaserver ~]$ kinit admin
[user@ipaserver ~]$ ipa user-mod jsmith --password
5.5. Unlocking User Accounts After Password Failures
If a user attempts to log in and uses the wrong password a certain number of times, then that user account is locked. The exact number of failed attempts that locks and account and the duration of the lockout is defined as part of the password policy (
Section 12.6, “Setting Account Lockout Policies”).
A password policy can implicitly define a reset period, where the account unlocks naturally after a certain amount of time lapses. However, if the duration is fairly long or if the deployment requires stronger security checks before unlocking an account, then an administrator can unlock an account manually.
An account is unlocked using the user-unlock command. For example:
[user@ipaserver ~]$ kinit admin
[user@ipaserver ~]$ ipa user-unlock jsmith
5.6. Managing User Private Groups
On Fedora systems, every time a user is created, a corresponding, secret user group is automatically created with that new user as its only member. This is a user private group. Using user private groups makes it simpler and safer to manage file and directory permissions because umask defaults only have to restrict user access, not group access.
When a new user is created in the FreeIPA domain, it is also created with a corresponding private group, following the Fedora convention. For most environments, this is an acceptable default behavior, but there may be certain users or types of users which do not require a private group or the environment may already have those GIDs
assigned to NIS groups or other system groups.
Command ipa group-find with option --private shows user private groups:
[user@ipaserver ~]$ ipa group-find --private jsmith
----------------
1 group matched
----------------
Group name: jsmith
Description: User private group for jsmith
GID: 1035400001
----------------------------
Number of entries returned 1
----------------------------
5.6.1. Disabling Private Groups for a Specific User
Private group creation can be disabled when the user is created by using the --noprivate option.
[user@ipaserver ~]$ ipa user-add jsmith --first=John --last=Smith --noprivate
5.6.2. Disabling Private Groups Globally
User private groups are managed through the Managed Entries Plug-in in 389 Directory Server. This plug-in can be disabled, which effectively disables private group creation for all new users.
This is done using the ipa-managed-entries command.
Use the ipa-managed-entries command to list possible Managed Entries Plug-in definitions. By default, there are two, one for new users (UPG) and one for netgroups (NGP).
[root@ipaserver ~]# ipa-managed-entries --list -p DMpassword
Available Managed Entry Definitions:
UPG Definition
NGP Definition
Disable the desired Managed Entries Plug-in instance. For example:
[root@ipaserver ~]# ipa-managed-entries -e "UPG Definition" -p DMpassword disable
Disabling Plugin
Restart the 389 Directory Server to load the new plug-in configuration.
[root@ipaserver ~]# service dirsrv restart
Managed Entries Plug-in instances can be re-enabled with the enable option.
5.7. Repairing Changed UID and GID Numbers
When a user is created, the user is automatically assigned a user ID number and a group ID number.
When the user logs into a FreeIPA system or service, SSSD on that system caches that username with the associated UID/GID numbers. The UID number is then used as the identifying key for the user. If a user with the same name but a different UID attempts to log into the system, then SSSD treats it as two different users with a name collision.
What this means is that SSSD does not recognize UID number changes. It interprets it as a different and new user, not an existing user with a different UID number. If an existing user changes the UID number, that user is prevented from logging into SSSD and associated services and domains. This also has an impact on any client applications which use SSSD for identity information; the user with the conflict will not be found or accessible to those applications.
UID/GID changes are not supported in FreeIPA or in SSSD.
If a user for some reason has a changed UID/GID number, then the SSSD cache must be cleared for that user before that user can log in again. For example:
[root@server ~]# sss_cache -u jsmith
5.8. Managing Unique UID and GID Number Assignments
A FreeIPA server must generate random UID and GID values and simultaneously ensure that replicas never generate the same UID or GID value. The need for unique UID and GID numbers might even cross FreeIPA domains, if a single organization has multiple disparate domains.
The UID and GID numbers are divided into ranges. By keeping separate numeric ranges for individual servers and replicas, the chances are minimal that any numbers issued by one server or replica will duplicate those from another. Ranges are updated and shared intelligently between servers and replicas through the Dynamic Numeric Assignment (DNA) Plug-in, as part of the backend 389 Directory Server instance for the domain. The same range is used for user IDs (uidNumber) and group IDs (gidNumber). A user and a group may have the same ID, but since the ID is set in different attributes, there is no conflict. Using the same ID number for both a user and a group also allows an administrator to configure user private groups, where a unique system group is created for each user and the ID number is the same for both the user and the group.
When a user is created interactively or without specifying a UID or GID number, then the user account is created with an ID number that is next available in the server or replica range. This means that a user always has a unique number for its UID number and, if configured, for its private group.
If a number is manually assigned to a user entry, the server does not validate that the uidNumber is unique. It will allow duplicate IDs; this is expected (though discouraged) behavior for POSIX entries. The same is true for group entries: a duplicate gidNumber can be manually assigned to the entry.
If two entries are assigned the same ID number, only the first entry is returned in a search for that ID number. However, both entries will be returned in searches for other attributes or with ipa user-find --all.
5.8.1. About ID Range Assignments During Installation
The FreeIPA administrator can initially define a range during server installation, using the --idstart and --idmax options with ipa-server-install. These options are not required, so the setup script can assign random ranges during installation.
If no range is set manually when the first FreeIPA server is installed, a range of 200,000 IDs is randomly selected. There are 10,000 possible ranges. Selecting a random range from that number provides a high probability of non-conflicting IDs if two separate FreeIPA domains are ever merged in the future.
With a single FreeIPA server, IDs are assigned to entries in order through the range. With replicas, the initial server ID range is split and distributed.
When a replica is installed, it is configured with an invalid range. It also has a directory entry (that is shared among replicas) that instructs the replica where it can request a valid range. When the replica starts, or as its current range is depleted so that less than 100 IDs are available, it can contact one of the available servers for a new range allotment. A special extended operation splits the range in two, so that the original server and the replica each have half of the available range.
If the range for the entire domain is close to depletion, a new range can be manually selected and assigned to one of the master servers. All replicas then request ID ranges from the master as necessary.
The changes to the range are done by editing the 389 Directory Server configuration to change the DNA Plug-in instance. The range is defined in the dnaNextRange parameter. For example:
[user@ipaserver ~]$ ldapmodify -x -D "cn=Directory Manager" -W -h server.example.com -p 389
Enter LDAP Password: *******
dn: cn=Posix IDs,cn=Distributed Numeric Assignment Plugin,cn=plugins,cn=config
changetype: modify
add: dnaNextRange
dnaNextRange: 123400000-123500000
This command only adds the specified range of values; it does not check that the values in that range are actually available. This check is performed when an attempt is made to allocate those values. If a range is added that contains mostly values that were already allocated, the system will cycle through the entire range searching for unallocated values, and then the operation ultimately fails if none are available.
5.9. Managing User and Group Schema
When a user entry is created, it is automatically assigned certain LDAP object classes which, in turn, make available certain attributes. LDAP attributes are the way that information is stored in the directory. (This is discussed in detail in the Directory Server Deployment Guide and the Directory Server Schema Reference.)
Table 5.1. Default FreeIPA User Object Classes
|
Description
|
Object Classes
|
|---|
|
FreeIPA object classes
|
ipaobject
|
|
Person object classes
|
| person | | organizationalperson | | inetorgperson | | inetuser | | posixaccount | | ipasshuser |
|
|
Kerberos object classes
|
| krbprincipalaux | | krbticketpolicyaux |
|
|
Managed entries (template) object classes
|
mepOriginEntry
|
Table 5.2. Default FreeIPA User Attributes
|
UI Field
|
Command-Line Option
|
Required, Optional, or Default
|
|---|
|
User login
|
username
|
Required
|
|
First name
|
--first
|
Required
|
|
Last name
|
--last
|
Required
|
|
Full name
|
--cn
|
Optional
|
|
Display name
|
--displayname
|
Optional
|
|
Initials
|
--initials
|
Default
|
|
Home directory
|
--homedir
|
Default
|
|
GECOS field
|
--gecos
|
Default
|
|
Shell
|
--shell
|
Default
|
|
Kerberos principal
|
--principal
|
Default
|
|
Email address
|
--email
|
Optional
|
|
Password
|
--password
Unlike the other options, this accepts no value. The script prompts for the new password.
|
Optional
|
|
User ID number
|
--uid
|
Default
|
|
Group ID number
|
--gidnumber
|
Default
|
|
Street address
|
--street
|
Optional
|
|
City
|
--city
|
Optional
|
|
State/Province
|
--state
|
Optional
|
|
Zip code
|
--postalcode
|
Optional
|
|
Telephone number
|
--phone
|
Optional
|
|
Mobile telephone number
|
--mobile
|
Optional
|
|
Pager number
|
--pager
|
Optional
|
|
Fax number
|
--fax
|
Optional
|
|
Organizational unit
|
--orgunit
|
Optional
|
|
Job title
|
--title
|
Optional
|
|
Manager
|
--manager
|
Optional
|
|
Car license
|
--carlicense
|
Optional
|
|
|
--noprivate
|
Optional
|
|
SSH Keys
|
--sshpubkey
|
Optional
|
|
Additional attributes
|
--addattr
|
Optional
|
5.9.1. About Changing the Default User and Group Schema
The FreeIPA configuration provides some validation when object classes are changed:
There are limits to the FreeIPA schema validation, however. Most important, the FreeIPA server does not check that the defined user or group object classes contain all of the required object classes for FreeIPA entries. For example, all FreeIPA entries require the ipaobject object class. However, when the user or group schema is changed, the server does not check to make sure that this object class is included; if the object class is accidentally deleted, then future entry add operations will fail.
Also, all object class changes are atomic, not incremental. The entire list of default object classes has to be defined every time there is a change. For example, a company may create a custom object class to store employee information like birthdays and employment start dates. The administrator cannot simply add the custom object class to the list; he must set the entire list of current default object classes plus the new object class. The existing default object classes must always be included when the configuration is updated. Otherwise, the current settings will be overwritten, which causes serious performance problems.
5.9.2. Applying Custom Object Classes to New User Entries
User and group accounts are created with a pre-defined set of LDAP object classes applied to the entry. Any attributes which belong to the object class can be added to the user entry.
While the standard and FreeIPA-specific LDAP object classes will cover most deployment scenarios, administrators may have custom object classes with custom attributes which should be applied to user entries.
Open the IPA Server tab.
Select the Configuration subtab.
Scroll to the User Options area.
At the bottom of the users area, click the Add link to add a new field for another object class.
Always include the existing default object classes when the configuration is updated. Otherwise, the current settings will be overwritten. If any object classes required by FreeIPA are not included, then subsequent attempts to add an entry will fail with object class violations.
When the changes are complete, click the Update link at the top of the Configuration page.
5.9.2.2. From the Command Line
Add the new object class to the list of object classes added to entries. The option for user object classes is --userobjectclasses.
Always include the existing default object classes when the configuration is updated. Otherwise, the current settings will be overwritten. If any object classes required by FreeIPA are not included, then subsequent attempts to add an entry will fail with object class violations.
For example:
[user@ipaserver ~]$ ipa config-mod --userobjectclasses=top --userobjectclasses=person --userobjectclasses=organizationalperson --userobjectclasses=inetorgperson --userobjectclasses=inetuser --userobjectclasses=posixaccount --userobjectclasses=krbprincipalaux --userobjectclasses=krbticketpolicyaux --userobjectclasses=ipaobject --userobjectclasses=employeeinfo
5.9.3. Applying Custom Object Classes to New Group Entries
As with user entries, administrators may have custom object classes with custom attributes which should be applied to group entries. These can be added automatically by adding the object classes to the FreeIPA server configuration.
Open the IPA Server tab.
Select the Configuration subtab.
Scroll to the Group Options area.
Click the Add link to add a new field for another object class.
Always include the existing default object classes when the configuration is updated. Otherwise, the current settings will be overwritten. If any object classes required by FreeIPA are not included, then subsequent attempts to add an entry will fail with object class violations.
When the changes are complete, click the Update link at the top of the Configuration page.
5.9.3.2. From the Command Line
Add the new object class to the list of object classes added to entries. The option for group object classes is --groupobjectclasses.
Always include the existing default object classes when the configuration is updated. Otherwise, the current settings will be overwritten. If any object classes required by FreeIPA are not included, then subsequent attempts to add an entry will fail with object class violations.
For example:
[user@ipaserver ~]$ ipa config-mod --userobjectclasses=top --userobjectclasses=groupofnames --userobjectclasses=nestedgroup --userobjectclasses=ipausergroup --userobjectclasses=ipaobject --userobjectclasses=ipasshuser --userobjectclasses=employeegroup
5.10. Managing User Groups
User groups are a way of centralizing control over important management tasks, particularly access control and password policies. Four groups are created during the installation, specifically for use by FreeIPA operations:
ipausers, which contains all users.
admins, which contains administrative users. The initial admin user belongs to this group.
editors, which is a special group for users working through the web UI. This group allows users to edit other users' entries, though without all of the rights of the admin user.
trust admins, which is a special group for users working with AD trust resources.
All groups in FreeIPA are essentially static groups, meaning that the members of the group are manually and explicitly added to the group. Tangentially, FreeIPA allows nested groups, where a group is a member of another group. In that case, all of the group members of the member group automatically belong to the parent group, as well.
Because groups are easy to create, it is possible to be very flexible in what groups to create and how they are organized. Groups can be defined around organizational divisions like departments, physical locations, or FreeIPA or infrastructure usage guidelines for access controls.
Some operating systems limit the number of groups that can be assigned to system users. For example, Solaris and AIX systems both limit users to 16 groups per user. This can be an issue when using nested groups, when a user may be automatically added to multiple groups.
When a group entry is created, it is automatically assigned certain LDAP object classes. (LDAP object classes and attributes are discussed in detail in the Directory Server Deployment Guide and the Directory Server Schema Reference.) For groups, only two attributes truly matter: the name and the description.
Table 5.3. Default FreeIPA Group Object Classes
|
Description
|
Object Classes
|
|---|
|
FreeIPA object classes
|
| ipaobject | | ipausergroup | | nestedgroup |
|
|
Group object classes
|
|
5.10.1. Creating User Groups
5.10.1.1. With the Web UI
Open the Identity tab, and select the User Groups subtab.
Click the Add link at the top of the groups list.
Enter all of the information for the group.
A unique name. This is the identifier used for the group in the FreeIPA domain, and it cannot be changed after it is created. The name cannot contain spaces, but other separators like an underscore (_) are allowed.
A text description of the group.
Whether the group is a Posix group, which adds Linux-specific information to the entry. By default, all groups are Posix groups unless they are explicitly configured not to be. Non-Posix groups can be created for interoperability with Windows or Samba.
Optionally, the GID number for the group. All Posix groups require a GID number, but FreeIPA automatically assigns the GID number.
Setting a GID number is not necessary because of the risk of collisions. If a GID number is given manually, FreeIPA will not override the specified GID number, even if it is not unique.
Click the Add and Edit button to go immediately to the member selection page.
5.10.1.2. With the Command Line
New groups are created using the group-add command. (This adds only the group; members are added separately.)
Two attributes are always required: the group name and the group description. If those attributes are not given as arguments, then the script prompts for them.
ipa group-add groupName --desc="description" [--nonposix]
Additionally, there is one other configuration option, --nonposix. (By default, all groups are created as POSIX groups.) To enable interoperability with Windows users and groups and programs like Samba, it is possible to create non-POSIX groups by using the --nonposix option. This option tells the script not to add the posixGroup object class to the entry.
For example:
[user@ipaserver ~]$ ipa group-add examplegroup --desc="for examples" --nonposix
----------------------
Added group "examplegroup"
----------------------
Group name: examplegroup
Description: for examples
When no arguments are used, the command prompts for the required group account information:
[user@ipaserver ~]$ ipa group-add
Group name: engineering
Description: for engineers
-------------------------
Added group "engineering"
-------------------------
Group name: engineering
Description: for engineers
GID: 387115842
When a group is created without specifying a GID number, then the group entry is assigned the ID number that is next available in the server or replica range. (Number ranges are described more in
Section 5.8, “Managing Unique UID and GID Number Assignments”.) This means that a group always has a unique number for its GID number.
If a number is manually assigned to a group entry, the server does not validate that the gidNumber is unique. It will allow duplicate IDs; this is expected (though discouraged) behavior for POSIX entries.
If two entries are assigned the same ID number, only the first entry is returned in a search for that ID number. However, both entries will be returned in searches for other attributes or with ipa group-find --all.
You cannot edit the group name. The group name is the primary key, so changing it is the equivalent of deleting the group and creating a new one.
5.10.2. Adding Group Members
5.10.2.1. With the Web UI (Group Page)
This procedure adds a user to a group. User groups can contain other user groups as their members. These are nested groups.
It can take up to several minutes for the members of the child group to show up as members of the parent group. This is especially true on virtual machines where the nested groups have more than 500 members.
When creating nested groups, be careful not to create recursive groups. For example, if GroupA is a member of GroupB, do not add GroupB as a member of GroupA. Recursive groups are not supported and can cause unpredictable behavior.
Open the Identity tab, and select the User Groups subtab.
Click the name of the group to which to add members.
Click the Add link at the top of the task area.
Click the checkbox by the names of the users to add, and click the right arrows button, >>, to move the names to the selection box.
Click the Add button.
Group members can be users or other user groups. It can take up to several minutes for the members of the child group to show up as members of the parent group. This is especially true on virtual machines where the nested groups have more than 500 members.
5.10.2.2. With the Web UI (User's Page)
Users can also be added to a group through the user's page.
Open the Identity tab, and select the Users subtab.
Click the name of the user to edit.
Open the User Groups tab on the user entry page.
Click the Add link at the top of the task area.
Click the checkbox by the names of the groups for the user to join, and click the right arrows button, >>, to move the groups to the selection box.
Click the Add button.
5.10.2.3. With the Command Line
Members are added to a group using the group-add-member command. This command can add both users as group members and other groups as group members.
The syntax of the group-add-member command requires only the group name and --users/--groups option(s) of user/group to add:
[user@ipaserver ~]$ ipa group-add-member groupName [--users=user [--users=...]] [--groups=group [--groups=...]]
For example, this adds three users to the engineering group:
[user@ipaserver ~]$ ipa group-add-member engineering --users=jsmith --users=bjensen --users=mreynolds
Group name: engineering
Description: for engineers
GID: 387115842
Member users: jsmith,bjensen,mreynolds
-------------------------
Number of members added 3
-------------------------
Likewise, other groups can be added as members, which creates nested groups:
[user@ipaserver ~]$ ipa group-add-member engineering --groups=dev --groups=qe1 --groups=dev2
Group name: engineering
Description: for engineers
GID: 387115842
Member groups: dev,qe1,dev2
-------------------------
Number of members added 3
-------------------------
When displaying nested groups, members are listed as members and the members of any member groups are listed as indirect members. For example:
[user@ipaserver ~]$ ipa group-show examplegroup
Group name: examplegroup
Description: for examples
GID: 93200002
Member users: jsmith,bjensen,mreynolds
Member groups: californiausers
Indirect Member users: sbeckett,acalavicci
It can take up to several minutes for the members of the child group to show up as members of the parent group. This is especially true on virtual machines where the nested groups have more than 500 members.
When creating nested groups, be careful not to create recursive groups. For example, if GroupA is a member of GroupB, do not add GroupB as a member of GroupA. Recursive groups are not supported and can cause unpredictable behavior.
A group member is removed using the group-remove-member command.
[user@ipaserver ~]$ ipa group-remove-member engineering --users=jsmith
Group name: engineering
Description: for engineers
GID: 855800009
Member users: bjensen,mreynolds
---------------------------
Number of members removed 1
---------------------------
5.10.2.4. Viewing Direct and Indirect Members of a Group
User groups can contain other user groups as members. This is called a nested group. This also means that a group has two types of members:
Direct members, which are added explicitly to the group
Indirect members, which are members of the group because they are members of another user group which is a member of the group
The FreeIPA web UI has an easy way to view direct and indirect members of a group. The members list is filtered by member type, and this can be toggled by selecting the Direct and Indirect radio buttons at the top right corner of the members list.
Being able to track indirect members makes it easier to assign group membership properly, without duplicating membership.
5.10.3. Deleting User Groups
When a user group is deleted, only the group is removed. The user accounts of group members (including nested groups) are not affected. Additionally, any access control delegations that apply to that group are removed.
Deleting a group is immediate and permanent. If any group configuration (such as delegations) is required, it must be assigned to another group or a new group created.
5.10.3.1. With the Web UI
Open the Identity tab, and select the User Groups subtab.
Select the checkbox by the name of the group to delete.
Click the Delete link at the top of the task area.
When prompted, confirm the delete action.
5.10.3.2. With the Command Line
The group-del command to deletes the specified group. For example:
[user@ipaserver ~]$ ipa group-del examplegroup
5.11. Searching for Users and Groups
By default, there are six attributes that are indexed for user searches and two that are indexed for group searches. These are listed in
Table 5.4, “Default Search Attributes”. All search attributes are searched in a user/group search.
Table 5.4. Default Search Attributes
|
User Search Attributes
|
|
First name
|
Last name
|
|
Login ID
|
Job title
|
|
Organizational unit
|
Phone number
|
|
Group Search Attributes
|
|
Name
|
Description
|
Both user and group main pages have a search bar in the upper right corner of the task area. This search box searches against all of the fields listed in
Table 5.4, “Default Search Attributes”. Type in any string, even a single letter, and click the magnifying glass icon. The UI filters the displayed list to match the search string.
5.11.2. With the Command Line
Searches are simple:
ipa user-find|group-find string options
There are a few general rules with searches:
If there is no string, then the search returns every entry in FreeIPA, up to the search limit.
With the command-line tools, only a single search string can be used for user and group searches. With the UI, multiple strings can be used.
Searches are case insensitive.
Search results are displayed alphabetically, with exact matches listed first, followed by partial matches.
Wildcards cannot be used in searches. The search string must include at least one character that appears in one of the indexed search fields.
Following additional option variations of group-find command can be executed to filter group searching:
ipa group-find --external
ipa group-find --posix
ipa group-find --nonposix
Example 5.2. User Search for John
The basic search looks for the string john, which can appear in any of the search indexes.
[user@ipaserver ~]$ ipa user-find john
---------------
2 users matched
---------------
User login: jpeterson
First name: john
Last name: peterson
Home directory: /home/jpeterson
Login shell: /bin/sh
UID: 855800007
GID: 855800007
Account disabled: False
User login: jsmith
First name: john
Last name: smith
Home directory: /home/jsmith
Login shell: /bin/sh
UID: 855800004
GID: 855800004
Account disabled: False
----------------------------
Number of entries returned 2
----------------------------
A search can also accept options like --raw. --raw prints the LDAP attributes for the user account rather than the reading-friendly field names.
[user@ipaserver ~]$ ipa user-find john --raw
---------------
2 users matched
---------------
uid: jpeterson
givenname: john
sn: peterson
homedirectory: /home/jpeterson
loginshell: /bin/sh
uidnumber: 855800007
gidnumber: 855800007
nsaccountlock: False
uid: jsmith
givenname: john
sn: smith
homedirectory: /home/jsmith
loginshell: /bin/sh
uidnumber: 855800004
gidnumber: 855800004
nsaccountlock: False
----------------------------
Number of entries returned 2
----------------------------
If the desired entry is not listed, it is possible that the search hit the preset search size limit before the entry was found. Change the search record or time limits, as in
Section 4.4.2, “Setting FreeIPA Search Limits”, to allow more entries to be returned.
5.12. Specifying Default User and Group Settings
FreeIPA uses a template when it creates new entries.
For users, the template is very specific. FreeIPA uses default values for several core attributes for FreeIPA user accounts. These defaults can define actual values for user account attributes (such as the home directory location) or it can define the format of attribute values, such as the username length. These settings also define the object classes assigned to users.
For groups, the template only defines the assigned object classes.
These default definitions are all contained in a single configuration entry for the FreeIPA server, cn=ipaconfig,cn=etc,dc=example,dc=com.
The configuration can be changed using the ipa config-mod command.
Table 5.5. Default User Parameters
|
Field
|
Command-Line Option
|
Descriptions
|
|---|
|
Maximum username length
|
--maxusername
|
Sets the maximum number of characters for usernames. The default value is eight.
|
|
Root for home directories
|
--homedirectory
|
Sets the default directory to use for user home directories. The default value is /home.
|
|
Default shell
|
--defaultshell
|
Sets the default shell to use for users. The default value is /bin/sh.
|
|
Default user group
|
--defaultgroup
|
Sets the default group to which all newly created accounts are added. The default value is ipausers, which is automatically created during the FreeIPA server installation process.
|
|
Default e-mail domain
|
--emaildomain
|
Sets the email domain to use to create email addressed based on the new accounts. The default is the FreeIPA server domain.
|
|
Search time limit
|
--searchtimelimit
|
Sets the maximum amount of time, in seconds, to spend on a search before the server returns results.
|
|
Search size limit
|
--searchrecordslimit
|
Sets the maximum number of records to return in a search.
|
|
User search fields
|
--usersearch
|
Sets the fields in a user entry that can be used as a search string. Any attribute listed has an index kept for that attribute, so setting too many attributes could affect server performance.
|
|
Group search fields
|
--groupsearch
|
Sets the fields in a group entry that can be used as a search string.
|
|
Certificate subject base
|
|
Sets the base DN to use when creating subject DNs for client certificates. This is configured when the server is set up.
|
|
Default user object classes
|
--userobjectclasses
|
Sets a list of object classes that are used to create FreeIPA user accounts.
|
|
Default group object classes
|
--groupobjectclasses
|
Sets a list of object classes that are used to create FreeIPA group accounts.
|
|
Password expiration notification
|
--pwdexpnotify
|
Sets how long, in days, before a password expires for the server to send a notification.
|
|
Password plug-in features
|
|
Sets the format of passwords that are allowed for users.
|
5.12.1. Viewing Settings from the Web UI
Open the IPA Server tab.
Select the Configuration subtab.
The complete configuration entry is shown in three sections, one for all search limits, one for user templates, and one for group templates.
5.12.2. Viewing Settings from the Command Line
The config-show command shows the current configuration which applies to all new user accounts. By default, only the most common attributes are displayed; use the --all option to show the complete configuration.
[user@ipaserver ~]$ ipa config-show --all
dn: cn=ipaConfig,cn=etc,dc=example,dc=com
Maximum username length: 32
Home directory base: /home
Default shell: /bin/sh
Default users group: ipausers
Default e-mail domain: example.com
Search time limit: 2
Search size limit: 100
User search fields: uid,givenname,sn,telephonenumber,ou,title
Group search fields: cn,description
Enable migration mode: FALSE
Certificate Subject base: O=EXAMPLE.COM
Default group objectclasses: top, groupofnames, nestedgroup, ipausergroup, ipaobject
Default user objectclasses: top, person, organizationalperson, inetorgperson, inetuser, posixaccount, krbprincipalaux, krbticketpolicyaux, ipaobject, ipasshuser
Password Expiration Notification (days): 4
Password plugin features: AllowNThash
SELinux user map order: guest_u:s0$xguest_u:s0$user_u:s0$staff_u:s0-s0:c0.c1023$unconfined_u:s0-s0:c0.c1023
Default SELinux user: unconfined_u:s0-s0:c0.c1023
Default PAC types: MS-PAC, nfs:NONE
cn: ipaConfig
objectclass: nsContainer, top, ipaGuiConfig, ipaConfigObject
Chapter 6. Identity: Managing Hosts and Services
Both DNS and Kerberos are configured as part of the initial client configuration. This is required because these are the two services that bring the machine within the FreeIPA domain and allow it to identify the FreeIPA server it will connect with. After the initial configuration, FreeIPA has tools to manage both of these services in response to changes in the domain services, changes to the IT environment, or changes on the machines themselves which affect Kerberos, certificate, and DNS services, like changing the client hostname.
This chapter describes how to manage identity services that relate directly to the client machine:
6.1. About Hosts, Services, and Machine Identity and Authentication
The basic function of an enrollment process is to create a host entry for the client machine in the FreeIPA directory. This host entry is used to establish relationships between other hosts and even services within the domain. These relationships are part of delegating authorization and control to hosts within the domain.
A host entry contains all of the information about the client within FreeIPA:
Service entries associated with the host
The host and service principal
Access control rules
Machine information, such as its physical location and operating system
Some services that run on a host can also belong to the FreeIPA domain. Any service that can store a Kerberos principal or an SSL certificate (or both) can be configured as an FreeIPA service. Adding a service to the FreeIPA domain allows the service to request an SSL certificate or keytab from the domain. (Only the public key for the certificate is stored in the service record. The private key is local to the service.)
A FreeIPA domain establishes a commonality between machines, with common identity information, common policies, and shared services. Any machine which belongs to a domain functions as a client of the domain, which means it uses the services that the domain provides. A FreeIPA domain provides three main services specifically for machines:
DNS
Kerberos
Certificate management
Machines are treated as another identity that is managed by FreeIPA. Clients use DNS to identify FreeIPA servers, services, and domain members — which, like user identities are stored in the 389 Directory Server instance for the FreeIPA server. Like users, machines can be authenticated to the domain using Kerberos or certificates to verify the machine's identity.
From the machine perspective, there are several tasks that can be performed that access these domain services:
Joining the DNS domain (machine enrollment)
Managing DNS entries and zones
Managing machine authentication
Authentication in FreeIPA includes machines as well as users. Machine authentication is required for the FreeIPA server to trust the machine and to accept FreeIPA connections from the client software installed on that machine. After authenticating the client, the FreeIPA server can respond to its requests. FreeIPA supports three different approaches to machine authentication:
SSH keys. The SSH public key for the host is created and uploaded to the host entry. From there, the System Security Services Daemon (SSSD) uses FreeIPA as an identity provider and can work in conjunction with OpenSSH and other services to reference the public keys located centrally in FreeIPA. This is described in
Section 6.8, “Managing Public SSH Keys for Hosts” and the
Fedora Deployment Guide.
Key tables (or
keytabs, a symmetric key resembling to some extent a user password) and machine certificates. Kerberos tickets are generated as part of the Kerberos services and policies defined by the server. Initially granting a Kerberos ticket, renewing the Kerberos credentials, and even destroying the Kerberos session are all handled by the FreeIPA services. Managing Kerberos is covered in
Chapter 13, Policy: Managing the Kerberos Domain.
Machine certificates. In this case, the machine uses an SSL certificate that is issued by the FreeIPA server's certificate authority and then stored in FreeIPA's Directory Server. The certificate is then sent to the machine to present when it authenticates to the server. On the client, certificates are managed by a service called
certmonger, which is described in
Appendix B, Working with certmonger.
A host entry is always created when a client is configured. On Fedora systems, this is done automatically with the
ipa-client-install script. On other platforms — and in alternative enrollment scenarios, as in
Section 6.3, “Enrolling Clients Manually” — the host entry is created manually.
6.2.1. Adding Host Entries from the Web UI
Open the Identity tab, and select the Hosts subtab.
Click the Add link at the top of the hosts list.
Fill in the machine name and select the domain from the configured zones in the drop-down list. If the host has already been assigned a static IP address, then include that with the host entry so that the DNS entry is fully created.
DNS zones can be created in FreeIPA, which is described in
Section 10.4.1, “Adding DNS Zones”. If the FreeIPA server does not manage the DNS server, the zone can be entered manually in the menu area, like a regular text field.
Select the Force checkbox to add the host DNS record, even if the hostname cannot be resolved.
This is useful for hosts which use DHCP and do not have a static IP address. This essentially creates a placeholder entry in the FreeIPA DNS service. When the DNS service dynamically updates its records, the host's current IP address is detected and its DNS record is updated.
Click the Add and Edit button to go directly to the expanded entry page and fill in more attribute information. Information about the host hardware and physical location can be included with the host entry.
6.2.2. Adding Host Entries from the Command Line
Host entries are created using the host-add command. This commands adds the host entry to the FreeIPA Directory Server. The full list of options with host-add are listed in the ipa host manpage. At its most basic, an add operation only requires the client hostname to add the client to the Kerberos realm and to create an entry in the FreeIPA LDAP server:
$ ipa host-add client1.example.com
If the FreeIPA server is configured to manage DNS, then the host can also be added to the DNS resource records using the --ip-address and --force options.
Example 6.1. Creating Host Entries with Static IP Addresses
$ ipa host-add --force --ip-address=192.168.166.31 client1.example.com
Commonly, hosts may not have a static IP address or the IP address may not be known at the time the client is configured. For example, laptops may be preconfigured as FreeIPA clients, but they do not have IP addresses at the time they're configured. Hosts which use DHCP can still be configured with a DNS entry by using --force. This essentially creates a placeholder entry in the FreeIPA DNS service. When the DNS service dynamically updates its records, the host's current IP address is detected and its DNS record is updated.
Example 6.2. Creating Host Entries with DHCP
$ ipa host-add --force client1.example.com
Host records are deleted using the
host-del command. If the FreeIPA domain uses DNS, then the
--updatedns option also removes the associated records of any kind for the host from the DNS.
$ ipa host-del --updatedns client1.example.com
6.3. Enrolling Clients Manually
Enrolling machines as clients in the FreeIPA domain is a two-part process. A host entry is created for the client (and stored in the 389 Directory Server instance), and then a keytab is created to provision the client.
Both parts are performed automatically by the ipa-client-install command. It is also possible to perform those steps separately; this allows for administrators to prepare machines and FreeIPA in advance of actually configuring the clients. This allows more flexible setup scenarios, including bulk deployments.
When performing a manual enrollment, the host entry is created separately, and then enrollment is completed when the client script is run, which creates the requisite keytab.
There are two ways to set the password. You can either supply your own or have FreeIPA generate a random one.
6.3.1. Performing a Split Enrollment
There may be a situation where an administrator in one group is prohibited from creating a host entry and, therefore, from simply running the ipa-client-install command and allowing it to create the host. However, that administrator may have the right to run the command after a host entry exists. In that case, one administrator can create the host entry manually, then the second administrator can complete the enrollment by running the ipa-client-install command.
When the second administrator runs the setup script, he must pass his Kerberos password and username (principal) with the ipa-client-install command. For example:
$ ipa-client-install -w secret -p admin2
The keytab is generated on the server and provisioned to the client machine, so that the client machine is not able to connect to the FreeIPA domain. The keytab is saved with root:root ownership and 0600 permissions.
6.4. Manually Unconfiguring Client Machines
A machine may need to be removed from one FreeIPA domain and moved to another domain or a virtual machine may be copied. There are a number of different situations where a FreeIPA client needs to be reconfigured. The easiest solution is to uninstall the client and then configure it afresh.
ipa-client-install --uninstall
If it is not possible to uninstall the client directly, then the FreeIPA configuration can be manually removed from the virtual machine.
When a machine is unenrolled, the procedure cannot be undone. The machine can only be enrolled again.
Remove the old hostname from the main keytab. This can be done by removing every principal in the realm or by removing specific principals. For example, to remove all principals:
$ ipa-rmkeytab -k /etc/krb5.keytab -r EXAMPLE.COM
To remove specific principals:
$ ipa-rmkeytab -k /etc/krb5.keytab -p host/server.example.com@EXAMPLE.COM
Disable tracking in
certmonger for every certificate. Each certificate must be removed from tracking individually.
$ ipa-getcert stop-tracking -n Server-Cert -d /etc/pki/nssdb
$ ipa-getcert stop-tracking -n Server2-Cert -d /etc/pki/nssdb
Remove the old host from the FreeIPA DNS domain. While this is optional, it cleans up the old FreeIPA entries associated with the system and allows it to be re-enrolled cleanly at a later time.
$ ipa host-del server.example.com
If the system should be re-added to a new FreeIPA domain — such as a virtual machine which was moved from one location to another — then the system can be rejoined to FreeIPA using the ipa-join command.
$ ipa-join
6.5.1. Adding and Editing Service Entries and Keytabs
As with host entries, service entries for the host (and any other services on that host which will belong to the domain) must be added manually to the FreeIPA domain. This is a two step process. First, the service entry must be created, and then a keytab must be created for that service which it will use to access the domain.
By default, FreeIPA saves its HTTP keytab to /etc/httpd/conf/ipa.keytab.
This keytab is used for the web UI. If a key were stored in ipa.keytab and that keytab file is deleted, the FreeIPA web UI will stop working, because the original key would also be deleted.
Similar locations can be specified for each service that needs to be made Kerberos aware. There is no specific location that must be used, but, when using ipa-getkeytab, you should avoid using /etc/krb5.keytab. This file should not contain service-specific keytabs; each service should have its keytab saved in a specific location and the access privileges (and possibly SELinux rules) should be configured so that only this service has access to the keytab.
6.5.1.1. Adding Services and Keytabs from the Web UI
Open the Identity tab, and select the Services subtab.
Click the Add link at the top of the services list.
Select the service type from the drop-down menu, and give it a name.
Select the hostname of the FreeIPA host on which the service is running. The hostname is used to construct the full service principal name.
Click the Add button to save the new service principal.
Use the ipa-getkeytab command to generate and assign the new keytab for the service principal.
# ipa-getkeytab -s server.example.com -p HTTP/server.example.com -k /etc/httpd/conf/krb5.keytab -e des-cbc-crc
The realm name is optional. The FreeIPA server automatically appends the Kerberos realm for which it is configured. You cannot specify a different realm.
The hostname must resolve to a DNS A record for it to work with Kerberos. You can use the --force flag to force the creation of a principal should this prove necessary.
The -e argument can include a comma-separated list of encryption types to include in the keytab. This supersedes any default encryption type.
Creating a new key resets the secret for the specified principal. This means that all other keytabs for that principal are rendered invalid.
6.5.1.2. Adding Services and Keytabs from the Command Line
Create the service principal. The service is recognized through a name like service/FQDN:
# ipa service-add serviceName/hostname
For example:
$ ipa service-add HTTP/server.example.com
-------------------------------------------------------
Added service "HTTP/server.example.com@EXAMPLE.COM"
-------------------------------------------------------
Principal: HTTP/server.example.com@EXAMPLE.COM
Managed by: ipaserver.example.com
Create the service keytab file using the ipa-getkeytab command. This command is run on the client in the FreeIPA domain. (Actually, it can be run on any FreeIPA server or client, and then the keys copied to the appropriate machine. However, it is simplest to run the command on the machine with the service being created.)
The command requires the Kerberos service principal (-p), the FreeIPA server name (-s), the file to write (-k), and the encryption method (-e). Be sure to copy the keytab to the appropriate directory for the service.
For example:
# ipa-getkeytab -s server.example.com -p HTTP/server.example.com -k /etc/httpd/conf/krb5.keytab -e des-cbc-crc
The realm name is optional. The FreeIPA server automatically appends the Kerberos realm for which it is configured. You cannot specify a different realm.
The hostname must resolve to a DNS A record for it to work with Kerberos. You can use the --force flag to force the creation of a principal should this prove necessary.
The -e argument can include a comma-separated list of encryption types to include in the keytab. This supersedes any default encryption type.
The ipa-getkeytab command resets the secret for the specified principal. This means that all other keytabs for that principal are rendered invalid.
6.5.2. Adding Services and Certificates for Services
While services can use keytabs, some services require certificates for access. In that case, a service can be added (or modified) to include a certificate with its service entry.
6.5.2.1. Adding Services and Certificates from the Web UI
Open the Identity tab, and select the Services subtab.
Click the Add link at the top of the services list.
Select the service type from the drop-down menu, and give it a name.
Select the hostname of the FreeIPA host on which the service is running. The hostname is used to construct the full service principal name.
Click the Add and Edit button to go directly to the service entry page.
Scroll to the bottom of the page, to the Service Certificate section.
Click the New Certificate button to create the service certificate.
6.5.2.2. Adding Services and Certificates from the Command Line
Create the service principal. The service is recognized through a name like service/FQDN:
[jsmith@ipaserver ~]$ kinit admin
[jsmith@ipaserver ~]$ ipa service-add serviceName/hostname
For example:
$ ipa service-add HTTP/server.example.com
-------------------------------------------------------
Added service "HTTP/server.example.com@EXAMPLE.COM"
-------------------------------------------------------
Principal: HTTP/server.example.com@EXAMPLE.COM
Managed by: ipaserver.example.com
Create a certificate for the service. Be sure to copy the keytab to the appropriate directory for the service.
For example:
$ ipa cert-request --principal=HTTP/web.example.com example.csr
Use the --add option to create the service automatically when requesting the certificate.
$ ipa-getcert request -d /etc/httpd/alias -n Server-Cert -K HTTP/client1.example.com -N 'CN=client1.example.com,O=EXAMPLE.COM'
6.5.3. Storing Certificates in NSS Databases
When services use certificates, the certificates and keys can be stored in NSS databases (which may also be used by the services themselves, as well as FreeIPA).
Create the NSS databases.
$ certutil -N -d /path/to/database/dir
Request the certificate using certutil, an NSS tool.
$ certutil -R -s "CN=client1.example.com,O=EXAMPLE.COM" -d /path/to/database/dir -a > example.csr
If the FreeIPA domain is using Certificate System for its CA, only the CN of the subject name is used. With a self-signed CA, the subject must match the configured certificate subject base. The FreeIPA server rejects requests with a subject base that differs from this value.
6.5.4. Configuring Clustered Services
The FreeIPA server is not cluster aware. However, it is possible to configure a clustered service to be part of FreeIPA by synchronizing Kerberos keys across all of the participating hosts and configuring services running on the hosts to respond to whatever names the clients use.
Enroll all of the hosts in the cluster into the FreeIPA domain.
Create any service principals and generate the required keytabs.
Collect any keytabs that have been set up for services on the host, including the host keytab at /etc/krb5.keytab.
Use the ktutil command to produce a single keytab file that contains the contents of all of the keytab files.
For each file, use the rkt command to read the keys from that file.
Use the wkt command to write all of the keys which have been read to a new keytab file.
Replace the keytab files on each host with the newly-created combined keytab file.
At this point, each host in this cluster can now impersonate any other host.
Some services require additional configuration to accommodate cluster members which do not reset hostnames when taking over a failed service.
For sshd, set GSSAPIStrictAcceptorCheck no in /etc/ssh/sshd_config.
For mod_auth_kerb, set KrbServiceName Any in /etc/httpd/conf.d/auth_kerb.conf.
For SSL servers, the subject name or a subject alternative name for the server's certificate must appear correct when a client connects to the clustered host. If possible, share the private key among all of the hosts. If each cluster member contains a subject alternative name which includes the names of all the other cluster members will satisfy any client connection requirements.
6.5.5. Using the Same Service Principal for Multiple Services
Within a cluster, the same service principal can be used for multiple services, spread across different machines.
Retrieve a service principal using the ipa-getkeytab command.
# ipa-getkeytab -s kdc.example.com -p HTTP/server.example.com -k /etc/httpd/conf/krb5.keytab -e des-cbc-crc
Either direct multiple servers or services to use the same file, or copy the file to individual servers as required.
6.6. Disabling and Re-enabling Host and Service Entries
Active services and hosts can be accessed by other services, hosts, and users within the domain. There can be situations when it is necessary to remove a host or a service from activity. However, deleting a service or a host removes the entry and all the associated configuration, and it removes it permanently.
6.6.1. Disabling Host and Service Entries
Disabling a host or service prevents domain users from access it without permanently removing it from the domain. This can be done by using the host-disable and service-disable commands.
For example, for a host:
[jsmith@ipaserver ~]$ kinit admin
[jsmith@ipaserver ~]$ ipa host-disable server.example.com
For a service, specify the principal rather than the hostname:
$ ipa service-disable http/server.example.com
Disabling a host entry not only disables that host. It disables every configured service on that host as well.
6.6.2. Re-enabling Hosts and Services
Disabling a service or host essentially kills its current, active keytabs. Removing the keytabs effectively removes the host or service from the FreeIPA domain without otherwise touching its configuration entry.
To re-enable a host or service, simply use the ipa-getkeytab command. The -s option sets which FreeIPA server to request the keytab, -p gives the principal name, and -k gives the file to which to save the keytab.
For example, requesting a new host keytab:
[jsmith@ipaserver ~]$ ipa-getkeytab -s ipaserver.example.com -p host/server.example.com -k /etc/krb5.keytab -D fqdn=server.example.com,cn=computers,cn=accounts,dc=example,dc=com -w password
If the ipa-getkeytab command is run on an active FreeIPA client or server, then it can be run without any LDAP credentials (-D and -w). The FreeIPA user uses Kerberos credentials to authenticate to the domain. To run the command directly on the disabled host, then supply LDAP credentials to authenticate to the FreeIPA server. The credentials should correspond to the host or service which is being re-enabled.
6.7. Extending Access Permissions over Other Hosts and Services
As discussed in
Section 1.3, “Relationships Between Servers and Clients”, within the FreeIPA domain,
manage means being able to retrieve a keytab and certificates for another host or service. Every host and service has a
managedby entry which lists what hosts or services can manage it. By default, a host can manage itself and all of its services. It is also possible to allow a host to manage other hosts, or services on other hosts, by updating the appropriate delegations or providing a suitable
managedby entry.
A FreeIPA service can be managed from any FreeIPA host, as long as that host has been granted, or delegated, permission to access the service. Likewise, hosts can be delegated permissions to other hosts within the domain.
If a host is delegated authority to another host through a managedBy entry, it does not mean that the host has also been delegated management for all services on that host. Each delegation has to be performed independently.
6.7.1. Delegating Service Management
A host is delegated control over a service using the service-add-host command. There are two parts to delegating the service: specifying the principal and identifying the hosts (in a comma-separated list) with control:
# ipa service-add-host principal --hosts=hostnames
For example:
# ipa service-add-host http/web.example.com --hosts=client1.example.com
Once the host is delegated authority, the host principal can be used to manage the service:
# kinit -kt /etc/krb5.keytab host/`hostname`
# ipa-getkeytab -s `hostname` -k /tmp/test.keytab -p http/web.example.com
Keytab successfully retrieved and stored in: /tmp/test.keytab
To create a ticket for this service, create a certificate request on the host with the delegated authority and use the cert-request command to create a service entry and load the certification information:
# ipa cert-request --add --principal=http/web.example.com web.csr
Certificate: MIICETCCAXqgA...[snip]
Subject: CN=web.example.com,O=EXAMPLE.COM
Issuer: CN=EXAMPLE.COM Certificate Authority
Not Before: Tue Feb 08 18:51:51 2011 UTC
Not After: Mon Feb 08 18:51:51 2016 UTC
Fingerprint (MD5): c1:46:8b:29:51:a6:4c:11:cd:81:cb:9d:7c:5e:84:d5
Fingerprint (SHA1):
01:43:bc:fa:b9:d8:30:35:ee:b6:54:dd:a4:e7:d2:11:b1:9d:bc:38
Serial number: 1005
6.7.2. Delegating Host Management
Hosts are delegated authority over other hosts through the host-add-managedby command. This creates a managedby entry. Once the managedby entry is created, then the host can retrieve a keytab for the host it has delegated authority over.
Log in as the admin user.
# kinit admin
Add the
managedby entry. For example, this delegates authority
over client2
to client1.
# ipa host-add-managedby client2.example.com --hosts=client1.example.com
Obtain a ticket as the host
client1 and then retrieve a keytab for
client2:
# kinit -kt /etc/krb5.keytab host/`hostname`
# ipa-getkeytab -s `hostname` -k /tmp/client2.keytab -p host/client2.example.com
Keytab successfully retrieved and stored in: /tmp/client2.keytab
6.7.3. Delegating Host or Service Management in the Web UI
Each host and service entry has a configuration tab that indicates what hosts have been delegated management control over that host or service.
Open the Identity tab, and select the Hosts or Services subtab.
Click the name of the host or service that you are going to grant delegated management to.
Click the Hosts subtab on the far right of the host/service entry. This is the tab which lists hosts which can manage the selected host/service.
Click the Add link at the top of the list.
Click the checkbox by the names of the hosts to which to delegate management for the host/service. Click the right arrows button, >>, to move the hosts to the selection box.
Click the Add button to close the selection box and to save the delegation settings.
6.7.4. Accessing Delegated Services
For both services and hosts, if a client has delegated authority, it can obtain a keytab for that principal on the local machine. For services, this has the format service/hostname@REALM. For hosts, the service is host.
With kinit, use the -k option to load a keytab and the -t option to specify the keytab.
For example, to access a host:
# kinit -kt /etc/krb5.keytab host/ipa.example.com@EXAMPLE.COM
To access a service:
# kinit -kt /etc/httpd/conf/krb5.keytab http/ipa.example.com@EXAMPLE.COM
6.8. Managing Public SSH Keys for Hosts
OpenSSH uses public-private key pairs to authenticate hosts. One machine attempts to access another machine and presents its key pair. The first time the host authenticates, the administrator on the target machine has to approve the request manually. The machine then stores the host's public key in a known_hosts file. Any time that the remote machine attempts to access the target machine again, the target machine simply checks its known_hosts file and then grants access automatically to approved hosts.
There are a few problems with this system:
The known_hosts file stores host entries in a triplet of the host IP address, hostname, and key. This file can rapidly become out of date if the IP address changes (which is common in virtual environments and data centers) or if the key is updated.
SSH keys have to be distributed manually and separately to all machines in an environment.
Administrators have to approve host keys to add them to the configuration, but it is difficult to verify either the host or key issuer properly, which can create security problems.
On Fedora, the System Security Services Daemon (SSSD) can be configured to cache and retrieve host SSH keys so that applications and services only have to look in one location for host keys. Because SSSD can use FreeIPA as one of its identity information providers, FreeIPA provides a universal and centralized repository of keys. Administrators do not need to worry about distributing, updating, or verifying host SSH keys.
6.8.1. About the SSH Key Format
When keys are uploaded to the FreeIPA entry, the key format can be either an
OpenSSH-style key or a raw
RFC 4253-style blob. Any RFC 4253-style key is automatically converted into an OpenSSH-style key before it is imported and saved into the FreeIPA LDAP server.
The FreeIPA server can identify the type of key, such as an RSA or DSA key, from the uploaded key blob. However, in a key file such as ~/.ssh/known_hosts, a key entry is identified by the hostname and IP address of the server, its type, then lastly the key itself. For example:
host.example.com,1.2.3.4 ssh-rsa AAA...ZZZ==
This is slightly different than a user public key entry, which has the elements in the order type key== comment:
"ssh-rsa ABCD1234...== ipaclient.example.com"
All three parts from the key file can be uploaded to and viewed for the host entry. In that case, the host public key entry from the ~/.ssh/known_hosts file needs to be reordered to match the format of a user key, type key== comment:
ssh-rsa AAA...ZZZ== host.example.com,1.2.3.4
The key type can be determined automatically from the content of the public key, and the comment is optional, to make identifying individual keys easier. The only required element is the public key blob itself.
6.8.2. About ipa-client-install and OpenSSH
The ipa-client-install script, by default, configures an OpenSSH server and client on the FreeIPA client machine. It also configures SSSD to perform host and user key caching. Essentially, simply configuring the client does all of the configuration necessary for the host to use SSSD, OpenSSH, and FreeIPA for key caching and retrieval.
There is an additional client configuration option, --ssh-trust-dns, which can be run with ipa-client-install and automatically configures OpenSSH to trust the FreeIPA DNS records, where the host keys are stored.
Alternatively, it is possible to disable OpenSSH at the time the client is installed, using the --no-sshd option. This prevents the install script from configuring the OpenSSH server.
Another option, --no-dns-sshfp, prevents the host from creating DNS SSHFP records with its own DNS entries. This can be used with or without the --no-sshd option.
6.8.3. Uploading Host SSH Keys Through the Web UI
The key for a host can probably be retrieved from a ~/.ssh/known_hosts. For example:
server.example.com,1.2.3.4 ssh-rsa AAAAB3NzaC1yc2EAAAABIwAAAQEApvjBvSFSkTU0WQW4eOweeo0DZZ08F9Ud21xlLy6FOhzwpXFGIyxvXZ52+siHBHbbqGL5+14N7UvElruyslIHx9LYUR/pPKSMXCGyboLy5aTNl5OQ5EHwrhVnFDIKXkvp45945R7SKYCUtRumm0Iw6wq0XD4o+ILeVbV3wmcB1bXs36ZvC/M6riefn9PcJmh6vNCvIsbMY6S+FhkWUTTiOXJjUDYRLlwM273FfWhzHK+SSQXeBp/zIn1gFvJhSZMRi9HZpDoqxLbBB9QIdIw6U4MIjNmKsSI/ASpkFm2GuQ7ZK9KuMItY2AoCuIRmRAdF8iYNHBTXNfFurGogXwRDjQ==
If necessary, generate a host key. When using the OpenSSH tools, make sure to use a blank passphrase and to save the key to a different location than the user's ~/.ssh/ directory, so it will not overwrite any existing keys.
[jsmith@server ~]$ ssh-keygen -t rsa -C "server.example.com,1.2.3.4"
Generating public/private rsa key pair.
Enter file in which to save the key (/home/jsmith/.ssh/id_rsa): /home/jsmith/.ssh/host_keys
Enter passphrase (empty for no passphrase):
Enter same passphrase again:
Your identification has been saved in /home/jsmith/.ssh/host_keys.
Your public key has been saved in /home/jsmith/.ssh/host_keys.pub.
The key fingerprint is:
4f:61:ee:2c:f7:d7:da:41:17:93:de:1d:19:ac:2e:c8 server.example.com
The key's randomart image is:
+--[ RSA 2048]----+
| .. |
| .+|
| o .* |
| o . .. *|
| S + . o+|
| E . .. .|
| . = . o |
| o . ..o|
| .....|
+-----------------+
Copy the public key from the key file. The full key entry has the form hostname,IP type key==. Only the key== is required, but the entire entry can be stored. To use all elements in the entry, rearrange the entry so it has the order type key== [hostname,IP]
[jsmith@server ~]$ cat /home/jsmith/.ssh/host_keys.pub
ssh-rsa AAAAB3NzaC1yc2E...tJG1PK2Mq++wQ== server.example.com,1.2.3.4
Open the Identity tab, and select the Hosts subtab.
Click the name of the host to edit.
In the Host Settings area of the Settings tab, click the SSH public keys: Add link.
The UI opens a new link, New: key not set Show/Set key. Click the Show/Set key link.
Paste in the public key for the host, and click the Set button.
The SSH public keys field now shows New: key set. Clicking the Show/Set key link opens the submitted key.
To upload multiple keys, click the Add link below the list of public keys, and upload the other keys.
When all the keys have been submitted, click the Update link at the top of the host's page to save the changes.
When the public key is saved, the entry is displayed as the key fingerprint, the comment (if one was included), and the key type
.
After uploading the host keys, configure SSSD to use FreeIPA as one of its identity domains and set up OpenSSH to use the SSSD tooling for managing host keys.
6.8.4. Adding Host Keys from the Command Line
Host SSH keys are added to host entries in FreeIPA, either when the host is created using host-add or by modifying the entry later.
Host keys are not created by the ipa-client-install command.
Run the host-mod command with the --sshpubkey option to upload the 64 bit-encoded public key to the host entry.
Adding a host key also changes the DNS SSHFP entry for the host, so also use the --updatedns option to update the host's DNS entry.
For example:
[jsmith@server ~]$ ipa host-mod --sshpubkey="ssh-rsa 12345abcde== ipaclient.example.com" --updatedns host1.example.com
With a real key, the key is longer and usually ends with an equals sign (=).
To upload multiple keys, pass a comma-separated list of keys with a single --sshpubkey option:
--sshpubkey="12345abcde==,key2==,key3=="
A host can have multiple public keys.
6.8.5. Removing Host Keys
Host keys can be removed once they expire or are no longer valid.
To remove an individual host key, it is easiest to remove the key through the web UI:
Open the Identity tab, and select the Hosts subtab.
Click the name of the host to edit.
Open the Host Settings area of the Settings tab.
Click the Delete link by the fingerprint of the key to remove.
Click the Update link at the top of the host's page to save the changes.
The command-line tools can be used to remove all keys. This is done by running ipa host-mod with the --sshpubkey= set to a blank value; this removes all public keys for the host. Also, use the --updatedns option to update the host's DNS entry. For example:
[jsmith@server ~]$ kinit admin
[jsmith@server ~]$ ipa host-mod --sshpubkey= --updatedns host1.example.com
6.9. Renaming Machines and Reconfiguring FreeIPA Client Configuration
The hostname of a system is critical for the correct operation of Kerberos and SSL. Both of these security mechanisms rely on the hostname to ensure that communication is occurring between the specified hosts. Infrastructures which use virtual machines or clustered servers will commonly have hosts which are renamed because systems are copied, moved, or renamed.
Fedora does not provide a simple rename command to facilitate the renaming of a FreeIPA host. Renaming a host in a FreeIPA domain involves deleting the entry in FreeIPA, uninstalling the client software, changing the hostname, and re-enrolling using the new name. Additionally, part of renaming hosts requires regenerating service principals.
To reconfigure the client:
Identify which services are running on the machine. These need to be re-created when the machine is re-enrolled.
# ipa service-find server.example.com
Each host has a default service which does not appear in the list of services. This service can be referred to as the "host service". The service principal for the host service is host/<hostname>, such as host/server.example.com. This principal can also be referred to as the host principal.
Identify all host groups to which the machine belongs.
# ipa hostgroup-find server.example.com
Identify which of the services have certificates associated with them. This can be done using the ldapsearch command to check the entries in the FreeIPA LDAP database directly:
# ldapsearch -x -b "cn=accounts,dc=example,dc=com" "(&(objectclass=ipaservice)(userCertificate=*))" krbPrincipalName
For any service principals (in addition to the host principal), determine the location of the corresponding keytabs on server.example.com. The keytab location is different for each service, and FreeIPA does not store this information.
Each service on the client system has a Kerberos principal in the form service name/hostname@REALM, such as ldap/server.example.com@EXAMPLE.COM.
Unenroll the client machine from the FreeIPA domain:
# ipa-client-install --uninstall
For each identified keytab other than
/etc/krb5.keytab, remove the old principals:
# ipa-rmkeytab -k /path/to/keytab -r EXAMPLE.COM
On another FreeIPA machine, as a FreeIPA administrator, remove the host entry. This removes all services and revokes all certificates issued for that host:
# ipa host-del server.example.com
At this point, the host is completely removed from FreeIPA.
Rename the machine.
Re-enroll the system with FreeIPA:
# ipa-client-install
This generates a host principal for with the new hostname in /etc/krb5.keytab.
For every service that needs a new keytab, run the following command:
# ipa service-add serviceName/new-hostname
To generate certificates for services, use either certmonger or the FreeIPA administration tools.
Re-add the host to any applicable host groups.
6.10. Managing Host Groups
Host groups are a way of centralizing control over important management tasks, particularly access control.
All groups in FreeIPA are essentially static groups, meaning that the members of the group are manually and explicitly added to the group. Tangentially, FreeIPA allows nested groups, where a group is a member of another group. In that case, all of the group members of the member group automatically belong to the parent group, as well.
Because groups are easy to create, it is possible to be very flexible in what groups to create and how they are organized. Groups can be defined around organizational divisions like departments, physical locations, or FreeIPA or infrastructure usage guidelines for access controls.
6.10.1. Creating Host Groups
6.10.1.1. Creating Host Groups from the Web UI
Open the Identity tab, and select the Host Groups subtab.
Click the Add link at the top of the groups list.
Enter the name and a description for the group.
Click the Add and Edit button to go immediately to the member selection page.
6.10.1.2. Creating Host Groups from the Command Line
New groups are created using the hostgroup-add command. (This adds only the group; members are added separately.)
Two attributes are always required: the group name and the group description. If those attributes are not given as arguments, then the script prompts for them.
$ ipa hostgroup-add groupName --desc="description"
6.10.2. Adding Group Members
6.10.2.1. Adding Group Members from the Web UI
Open the Identity tab, and select the Host Groups subtab.
Click the name of the group to which to add members.
Click the Add link at the top of the task area.
Click the checkbox by the names of the hosts to add, and click the right arrows button, >>, to move the hosts to the selection box.
Click the Add button.
6.10.2.2. Adding Group Members from the Command Line
Members are added to a host group using the hostgroup-add-member command. This command can add both hosts as group members and other groups as group members.
The syntax of the hostgroup-add-member command requires only the group name and a comma-separated list of hosts to add:
$ ipa hostgroup-add-member groupName [--hosts=list] [--hostgroups=list]
For example, this adds three hosts to the caligroup group:
$ ipa hostgroup-add-member caligroup --hosts=ipaserver.example.com,client1.example.com,client2.example.com
Group name: caligroup
Description: for machines in california
GID: 387115842
Member hosts: ipaserver.example.com,client1.example.com,client2.example.com
-------------------------
Number of members added 3
-------------------------
Likewise, other groups can be added as members, which creates nested groups:
$ ipa hostgroup-add-member caligroup --groups=mountainview,sandiego
Group name: caligroup
Description: for machines in california
GID: 387115842
Member groups: mountainview,sandiego
-------------------------
Number of members added 2
-------------------------
6.11. Troubleshooting Host Problems
6.11.1. Certificate Not Found/Serial Number Not Found Errors
The FreeIPA information is stored in a separate LDAP directory than the certificate information, and these two LDAP databases are replicated separately. It is possible for a replication agreement to be broken for one directory and working for another, which can cause problems with managing clients.
Specifically, if the replication agreement between the two CA databases is broken, then a server may not be able to find certificate information about a valid FreeIPA client, causing certificate errors:
Certificate operation cannot be completed: EXCEPTION (Certificate serial number 0x2d not found)
For example, a FreeIPA server and replica have a function replication agreement between their FreeIPA databases, but the replication agreement between their CA databases is broken. If a host is created on the server, the host entry is replicated over to the replica — but the certificate for that host is not replicated. The replica is aware of the client, but any management operations for that client will fail because the replica doesn't have a copy of its certificate.
6.11.2. Debugging Client Connection Problems
Client connection problems are apparent immediately. This can mean that users cannot log into a machine or attempts to access user and group information fails (for example, getent passwd admin).
Authentication in FreeIPA is managed with the SSSD daemon, which is described in the Red Hat Enterprise Linux Deployment Guide. If there are problems with client authentication, then check the SSSD information.
First, check the SSSD logs in /var/log/sssd/. There is a specific log file for the DNS domain, such as sssd_example.com.log. If there is not enough information in the logs at the default logging level, then increase the log level.
To increase the log level:
Open the sssd.conf file.
vim /etc/sssd/sssd.conf
In the [domain/example.com] section, set debug_level.
debug_level = 9
Restart the sssd daemon.
service sssd restart
Check the /var/log/sssd/sssd_example.com.log file for the debug messages.
Chapter 7. Identity: Integrating with NIS Domains and Netgroups
Network information service (NIS) is one of the most common ways to manage identities and authentication on Unix networks. It is simple and easy to use, but it also has inherent security risks and a lack of flexibility that can make administering NIS domains problematic.
FreeIPA supplies a way to integrate netgroups and other NIS data into the FreeIPA domain, which incorporates the stronger security structure of FreeIPA over the NIS configuration. Alternatively, administrators can simply migrate user and host identities from a NIS domain into the FreeIPA domain.
7.1. About NIS and FreeIPA
Network information service (NIS) centrally manages authentication and identity information such as users and passwords, hosts and IP addresses, and POSIX groups. This was originally called Yellow Pages (abbreviated YP) because of its simple focus on identity and authentication lookups.
NIS is considered too insecure for most modern network environments because it provides no host authentication mechanisms and it transmits all of its information over the network unencrypted, including password hashes. Still, while NIS has been falling out of favor with administrators, it is still actively used by many system clients. There are ways to work around those insecurities by integrating NIS with other protocols which offer enhanced security.
In FreeIPA, NIS objects are integrated into FreeIPA using the underlying LDAP directory. LDAP services offer support for NIS objects (as defined in
RFC 2307), which FreeIPA customizes to provide better integration with other domain identities. The NIS object is created inside the LDAP service and then a module like
nss_ldap or SSSD fetches the object using an encrypted LDAP connection.
NIS entities are stored in netgroups. A netgroup allows nesting (groups inside groups), which standard Unix groups don't support. Also, netgroups provide a way to group hosts, which is also missing in Unix group.
NIS groups work by defining users and hosts as members of a larger domain. A netgroup sets a trio of information — host, user, domain. This is called a
triple.
host,user,domain
A netgroup triple associates the user or the host with the domain; it does not associate the user and the host with each other. Therefore, a triple usually defines a host or a user for better clarity and management.
host.example.com,,nisdomain.example.com
-,jsmith,nisdomain.example.com
NIS distributes more than just netgroup data. It stores information about users and passwords, groups, network data, and hosts, among other information. FreeIPA can use a NIS listener to map passwords, groups, and netgroups to FreeIPA entries.
In FreeIPA LDAP entries, the users in a netgroup can be a single user or a group; both are identified by the memberUser parameter. Likewise, hosts can be either a single host or a host group; both are identified by the memberHost attribute.
dn: ipaUniqueID=d4453480-cc53-11dd-ad8b-0800200c9a66,cn=ng,cn=accounts,...
objectclass: top
objectclass: ipaAssociation
objectclass: ipaNISNetgroup
ipaUniqueID: d4453480-cc53-11dd-ad8b-0800200c9a66
cn: netgroup1
memberHost: fqdn=host1.example.com,cn=computers,cn=accounts,...
memberHost: cn=VirtGuests,cn=hostgroups,cn=accounts,...
memberUser: cn=jsmith,cn=users,cn=accounts,...
memberUser: cn=bjensen,cn=users,cn=accounts,...
memberUser: cn=Engineering,cn=groups,cn=accounts,...
nisDomainName: nisdomain.example.com
In FreeIPA, these netgroup entries are handled using the netgroup-* commands, which show the basic LDAP entry:
# ipa netgroup-show netgroup1
Netgroup name: netgroup1
Description: my netgroup
NIS domain name: nisdomain
Member User: jsmith
Member User: bjensen
Member User: Engineering
Member Host: host1.example.com
Member Host: VirtGuests
When a client attempts to access the NIS netgroup, then FreeIPA translates the LDAP entry into a traditional NIS map and sends it to a client over the NIS protocol (using a NIS plug-in) or it translates it into an LDAP format that is compliant with RFC 2307 or RFC 2307bis.
7.2. Setting the NIS Port for FreeIPA
The FreeIPA server binds to its NIS services over a random port that is selected when the server starts. It sends that port assignment to the portmapper so that NIS clients know what port to use to contact the FreeIPA server.
Administrators may need to open a firewall for NIS clients or may have other services that need to know the port number in advance and need that port number to remain the same. In that case, an administrator can specify the port to use.
Any available port number below 1024 can be used for the NIS Plug-in setting.
The NIS configuration is in the NIS Plug-in in FreeIPA's internal Directory Server instance. To specify the port:
Edit the plug-in configuration and add the port number as an argument. For example, to set the port to 514:
[root@ipaserver ~]# ldapmodify -x -D 'cn=directory manager' -w secret
dn: cn=NIS Server,cn=plugins,cn=config
changetype: modify
add: nsslapd-pluginarg0
nsslapd-pluginarg0: 514
modifying entry "cn=NIS Server,cn=plugins,cn=config"
Restart the Directory Server to load the new plug-in configuration.
[root@ipaserver ~]# service dirsrv restart
All netgroups in FreeIPA are essentially static groups, meaning that the members of the group are manually and explicitly added to the group. Tangentially, FreeIPA allows nested groups, where a group is a member of another group. In that case, all of the group members of the member group automatically belong to the parent group, as well.
Netgroups are added in two steps: the group itself is created, and then members are added to it.
Open the Identity tab, and select the Netgroups subtab.
Click the Add link at the top of the netgroups list.
Enter both a unique name and a description for the netgroup. Both the name and description are required.
The group name is the identifier used for the netgroup in the FreeIPA domain, and it cannot be changed after it is created. The name cannot contain spaces, but other separators like an underscore (_) are allowed.
Click the Add and Edit button to go immediately to the netgroup's edit pages.
Optionally, set the NIS domain for the netgroup. This defaults to the FreeIPA domain, but it can be changed.
Click the Settings tab.
Enter the name of the alternate NIS domain in the NIS domain name field.
The NIS domain name field sets the domain that appears in the netgroup triple. It does not affect which NIS domain the FreeIPA listener responds to.
7.3.1.2. With the Command Line
New netgroups are added using the netgroup-add command. This adds only the group; members are added separately. Two attributes are always required: the group name and the group description. If those attributes are not given as arguments, then the script prompts for them. There is also an option to set the NIS domain name to use for the group; this defaults to the FreeIPA domain, but it can be set to something different, depending on the network configuration.
$ ipa netgroup-add --desc="description" [--nisdomain=domainName] groupName
For example:
# ipa netgroup-add --desc="my new netgroup" example-netgroup
# ipa netgroup-add-member --hosts=ipa.example.com example-netgroup
# ypcat -d example.com -h ipa.example.com netgroup
(ipa.example.com,-,example.com)
The --nisdomain option sets the domain that appears in the netgroup triple. It does not affect which NIS domain the FreeIPA listener responds to.
7.3.2. Adding Netgroup Members
Netgroups can contain user groups, host groups, and other netgroups as their members. These are nested groups.
It can take up to several minutes for the members of the child group to show up as members of the parent group. This is especially true on virtual machines where the nested groups have more than 500 members.
When creating nested groups, be careful not to create recursive groups. For example, if GroupA is a member of GroupB, do not add GroupB as a member of GroupA. Recursive groups are not supported and can cause unpredictable behavior.
Open the Identity tab, and select the Netgroups subtab.
Click the name of the netgroup to which to add members.
Select the tab for the type of netgroup member to add. Netgroups can have users, user groups, hosts, host groups, and other netgroups as members.
Click the Add link at the top of the task area.
Click the checkbox by the names of the users to add, and click the right arrows button, >>, to move the names to the selection box.
Click the Add button.
7.3.2.2. With the Command Line
Once the group is configured, begin adding netgroup members with the netgroup-add-member command. Users, groups, hosts, host groups, and other netgroups can all be added to the netgroup entry. The entry name of the NIS group being edited usually comes at the end of the command:
# ipa netgroup-add-member --users=users --groups=groups --hosts=hosts --hostgroups=hostGroups --netgroups=netgroups groupName
To set more than one member, use a comma-separated list with the option. For example, this sets two users and two hosts with the other configuration:
# ipa netgroup-add-member --users=jsmith,bjensen --groups=ITadmin --hosts=host1.example.com,host2.example.com --hostgroups=EngDev --netgroups=nisgroup2 example-group
7.4. Exposing Automount Maps to NIS Clients
When the NIS service is enabled on a system, the FreeIPA server is automatically configured to set the NIS domain to the FreeIPA domain's name, and to include FreeIPA users, groups, and netgroups as passwd, group, and netgroup maps in the NIS domain.
If any automount maps are already defined, these maps need to be manually added to the NIS configuration in FreeIPA for them to be exposed to NIS clients. The NIS server is managed by a special plug-in entry in the FreeIPA LDAP directory; this is a container entry, and each NIS domain and map used by the NIS server is configured as a child entry beneath that container. The NIS domain entry in the must have the name of the NIS domain, the name of the NIS map, how to find the directory entries to use as the NIS map's contents, and which attributes to use as the NIS map's key and value. Most of these settings will be the same for every map.
The FreeIPA server stores the automount maps, grouped by automount location, in the cn=automount branch of the FreeIPA directory tree.
The NIS domain and map is added using LDAP tools, like
ldapadd, and editing the directory directly. For example, this adds an automount map that is named
auto.example in a location named
default and for a server named
nisserver:
ldapadd -h nisserver.example.com -x -D "cn=Directory Manager" -w secret
dn: nis-domain=example.com+nis-map=auto.example,cn=NIS Server,cn=plugins,cn=config
objectClass: extensibleObject
nis-domain: example.com
nis-map: auto.example
nis-filter: (objectclass=automount)
nis-key-format: %{automountKey}
nis-value-format: %{automountInformation}
nis-base: automountmapname=auto.example,cn=default,cn=automount,dc=example,dc=com
A similar add operation needs to be run for every map that is configured.
7.5. Migrating from NIS to FreeIPA
There is no direct migration path from NIS to FreeIPA. This is a manual process with three major steps: setting up netgroup entries in FreeIPA, exporting the existing data from NIS, and importing that data into FreeIPA. There are several options for how to set up the FreeIPA environment and how to export data; the best option depends on the type of data and the overall network environment that you have.
7.5.1. Preparing Netgroup Entries in FreeIPA
The first step is to identify what kinds of identities are being managed by NIS. Frequently, a NIS server is used for either user entries or host entries, but not for both, which can simplify the data migration process.
Create an entry for every user referenced in a netgroup.
Create an entry for every host referenced in a netgroup.
Create a netgroup with the same name as the original netgroup.
Add the users and hosts as direct members of the netgroup. Alternatively, put add the users and hosts into FreeIPA groups or other netgroups, and then add those groups as members to the netgroup.
7.5.2. Enabling the NIS Listener in FreeIPA
The FreeIPA Directory Server can function as a limited NIS server. The slapi-nis plug-in sets up a special NIS listener that receives incoming NIS requests and manages the NIS maps within the Directory Server. FreeIPA uses three NIS maps:
Using FreeIPA as an intermediate NIS server offers a reasonable way to handle NIS requests while migrating NIS clients and data.
The slapi-nis plug-in is not enabled by default. To enable NIS for FreeIPA:
Obtain new Kerberos credentials as a FreeIPA admin user.
[root@ipaserver ~]# kinit admin
Enable the NIS listener and compatibility plug-ins:
[root@ipaserver ~]# ipa-nis-manage enable
[root@ipaserver ~]# ipa-compat-manage enable
Restart the DNS and Directory Server service:
[root@server ~]# service restart rpcbind
[root@server ~]# service restart dirsrv
7.5.3. Setting Weak Password Encryption for NIS User Authentication to FreeIPA
A NIS server can handle CRYPT password hashes. Once an existing NIS server is migrated to FreeIPA (and its underlying LDAP database), it may still be necessary to preserve the NIS-supported CRYPT passwords. However, the LDAP server does not use CRYPT hashes by default. It uses SSHA or SSHA-256. If the 389 Directory Server password hash is not changed, then NIS users cannot authenticate to the FreeIPA domain, and kinit fails with password failures.
To set the underlying 389 Directory Server to use CRYPT as the password hash, change the passwordStorageScheme attribute using ldapmodify:
[root@server ~]# ldapmodify -D "cn=directory server" -w secret -p 389 -h ipaserver.example.com
dn: cn=config
changetype: modify
replace: passwordStorageScheme
passwordStorageScheme: crypt
Changing the password storage scheme only applies the scheme to new passwords; it does not retroactively change the encryption method used for existing passwords.
If weak crypto is required for password hashes, it is better to change the setting as early as possible so that more user passwords use the weaker password hash.
Chapter 8. Identity: Integrating with Active Directory Through Cross-Realm Kerberos Trusts
Active Directory is a Microsoft's implementation of LDAP, Kerberos, SMB, and few other protocol families. While there are many differences in the way these protocol families are implemented, in its core trusting one Active Directory domain to another means establishing relationships between the two domains on the Kerberos protocol level. Kerberos allows the configuration of trusted realms. Each realm has its own resources and users, yet the trust relationship allows users of any trusted realm to obtain tickets and connect to machines or services in a peer realm as if they were members of that peer realm.
Because of differences in the way that Windows and Linux domains implement LDAP services, DNS management, and even Kerberos realms, it is difficult to establish a direct trust between Active Directory and Linux domains manually. A trust relationship using FreeIPA centrally defines and establishes the Kerberos trust and DNS mappings so that Active Directory users can access Linux hosts and services completely transparently, using one set of credentials.
Active Directory was implemented on top of existing domain membership as provided by SMB protocol. In order to give short transition path for existing deployments, many of the concepts from SMB protocol are used internally by Active Directory. Trust relationship is one of these; establishing trust between two domains, in fact, requires execution of an SMB command sequence that leads to creation of specialized accounts in LDAP storage of domain controllers in both domains. When this step is performed, resulting accounts can be used to perform Kerberos authentication against the other domain and represent shared ticket and key for Kerberos cross-realm trust.
8.1. The Meaning of "Trust"
Kerberos has the ability to create a relationship between two otherwise separate realms. This is called a
cross-realm trust. This is described in some detail in
Managing Single Sign-On and Smart Cards. These realms create a shared ticket and key so a member of one realm is perceived as a member of both realms. One realm
trusts another.
8.1.1. How Trust Works: Transparency Between Kerberos and DNS Realms
Both Active Directory and FreeIPA manage a variety of core services: Kerberos, LDAP, DNS, certificate services. For these two disparate domains to be integrated transparently, all of these core services need to be able to interact cleanly with one another.
Those services can be broken into two major points of interaction: a Kerberos realm and a DNS domain. Certificate services, LDAP entries, and other services can be managed independently for Active Directory and FreeIPA. The place where they interect is where identities need to be authenticated (Kerberos) and a mechanism to route queries between domains (DNS).
8.1.1.1. Components Involved in Trusts
FreeIPA cross-realm trusts leverage four primary components:
Active Directory
Samba, to perform SMB protocol operations against domain controllers in Active Directory and represent points of communication that Active Directory domain controllers expect to exist in another Active Directory domain
SSSD, to query and cache user, group, and Kerberos ticket information for users from Active Directory. SSSD also maps Security Identifiers of user and group objects on Active Directory side to user and group identifiers on FreeIPA side
FreeIPA
8.1.1.2. Active Directory and FreeIPA Directories
One of the most common backends for user identities is Active Directory, and many environments — even primarily Linux or heterogenous environments — rely on Active Directory for user management. In many environments, however, that means that an entirely different set of users must be defined to access Linux systems.
Trusts allows a natural division of labor in an IT environment between user administration (in Active Directory) and Linux or data center management (through FreeIPA). All user accounts can be stored in Active Directory, without needing to recreate user accounts on Linux systems, while all Linux systems can still be centrally managed using native Linux tools.
Trust relationship is unidirectional. Active Directory users can access FreeIPA resources and services, but FreeIPA users cannot access Active Directory resources. Trust allows Windows administrators and users to be able to access and manage Linux resources
.
It is worth to note that while a single trust relationship is unidirectional, FreeIPA technically establishes bidirectional trust relationships with Active Directory and internally uses FreeIPA to Active Directory trust path to query Active Directory users membership information from Active Directory domain controllers. However, to allow full access for FreeIPA users to Active Directory resources, FreeIPA needs to advance its implementation of Global Catalog service as required by Active Directory-compliant domains.
Active Directory may contain a number of subordinated domains. In this case one Active Directory domain is called forest root domain. A trust relationship is established between a single Active Directory forest root domain and a single FreeIPA domain.
No Windows machine can be a client of the FreeIPA domain in a trust environment. All Windows machines must be in an Active Directory domain.
A relationship is established between the Active Directory environment and the FreeIPA environment through a trust agreement, which identifies the involved domains and the settings for the trust environment.
Both Active Directory and FreeIPA can define DNS services, and those DNS domains must interact cleanly with each other. There are two potential DNS configurations:
In both cases, the different domains forward requests to each other as necessary and maintain different DNS namespaces. It is just a matter of defining how they recognize each other for forwarding queries.
Both Active Directory and FreeIPA must be configured with integrated DNS servers.
8.1.1.4. Kerberos Realms, Authentication, and Authorization
Each object in Active Directory can be addressed using its Security Identifier (SID). Users, groups, machine accounts, and other objects all have associated SIDs. In some cases there could be more than one SID associated with an object. For performance purposes information about SIDs of related objects, including group membership, is stored in each Kerberos ticket for Active Directory users in a special dataset called privileged access certificates or MS-PAC. MS-PAC is issued as part of the Kerberos ticket and digitally signed by the Active Directory domain controller that issued the ticket. Digital signature allows to verify authenticity of the information and avoid requesting it over and over again from Active Directory domain controllers, making more efficient use of network and computing resources.
Understanding the group mapping for trusts can help clarify how groups should be structured in trust environments.
When an Active Directory user requests a ticket for a service in FreeIPA domain, it presents a cross-realm ticket granting ticket (TGT) issued to an Active Directory user by an Active Directory domain controller. This ticket contains MS-PAC information, signed by the Active Directory domain controller.
FreeIPA KDC, upon request to issue a ticket for a service in FreeIPA domain, verifies MS-PAC information in an Active Directory user ticket. If it contains any security identifier that should be filtered out and not allowed to access FreeIPA domain, the request to obtain a service ticket is rejected.
If FreeIPA KDC issues the ticket to the service in FreeIPA domain, Active Directory security identifiers from MS-PAC are used to map an Active Directory user membership to FreeIPA groups. If an Active Directory SID is an external member of FreeIPA group, FreeIPA KDC will track down any POSIX group this group is included into and will add its SID to MS-PAC structure.
Resulting ticket to the FreeIPA service is digitally signed by FreeIPA KDC. Its MS-PAC information will contain SIDs of original Active Directory objects and FreeIPA POSIX groups of which these objects are external members. This ticket is then used by a software that requested the ticket to connect to the actual FreeIPA service.
When the connection request comes to a FreeIPA client, FreeIPA (through SSSD, as a FreeIPA client), extracts the Active Directory security identifiers from the PAC and maps them to POSIX group and user identifiers. The user is granted access to the FreeIPA-hosted services according to their access rules. Additionally, the FreeIPA group information in the SSSD user cache is updated to include the mapped FreeIPA groups for the Active Directory user.
All Kerberos communication for both Active Directory and FreeIPA for trusts uses GSS-API.
A simpler way of saying this is that Active Directory supplies a list of groups for each user, based on an identifier for the group. FreeIPA compares that list of Active Directory groups to memberships in FreeIPA groups (where each group member is identified by that SID, rather than by a name or DN). If the Active Directory groups to which the user belongs are known to the FreeIPA domain, then the user is recognized by the FreeIPA domain.
The crucial factor to realize in this is that Active Directory users are recognized to the FreeIPA domain not by their Active Directory user entry, but by their FreeIPA group memberships. In a sense, Active Directory users are not trusted by the FreeIPA domain — FreeIPA groups are.
Since in POSIX environment every running process should be running under some user and have some group membership to access files, it is important that every user of FreeIPA services has corresponding POSIX identifier and user belongs to some groups which have POSIX identifiers. Each Active Directory user, therefore, should have membership in some POSIX group to be able to access files and run processes in FreeIPA domain.
Additionally, Active Directory user entries are never stored in FreeIPA LDAP and cannot be addressed by a DN. Group members in FreeIPA LDAP always addressed by their DNs. This means Active Directory users cannot be directly added to FreeIPA POSIX groups.
FreeIPA LDAP schema supports nested group membership. Each FreeIPA group may include another FreeIPA group as its member. When membership information is processed by FreeIPA KDC or SSSD, nested groups are unrolled and whole set of members is flattened.
FreeIPA has introduced an intermediary, non-POSIX group type, external groups, which allow entities outside FreeIPA or a Linux system to be added as a member. That external group can then be added to a standard FreeIPA (POSIX) group as a member.
When Active Directory objects are added to a FreeIPA group, they can be identified by their SID or by name, in the formats DOMAIN\group_name or group_name@domain. FreeIPA then resolves the object name to the SID and stores the SID as the group member entry, to be compared to any offered user PAC.
All sessions in a trust environment are ultimately secured with Kerberos tickets, but users have different login options:
Ticket-based authentication through kinit
Simple username/password authentication that is negotiated into a ticket
Passwordless authentication that is negotiated into a ticket (depending on the Kerberos client configuration
8.1.2. Trust in Contrast to Synchronization
Trusts and synchronization are fundamentally different approaches to integrating a FreeIPA domain and Active Directory domain. The structure of trust domains (outlined in
Section 8.1.1, “How Trust Works: Transparency Between Kerberos and DNS Realms”), the complexity of group assignments, the location of user and group entries, and other factors all influence which solution is most effective for a given environment.
Synchronization has a certain simplicity in its overall topology and setup, and it has a relatively small administrative footprint. However, it is much more limited in how it handles directory data, its ability to map entries, its overall performance, and its security.
Table 8.1. Positives and Negatives of Using Sync
|
Positives of Sync
|
Negatives of Sync
|
|---|
Simple setup procedures
Few rules about the Active Directory configuration, including being agnostic about DNS and Kerberos domains
Users and groups can originate in both Active Directory and FreeIPA domains
Active Directory users can function as FreeIPA users, including as administrators
Windows machines can be added as clients to the FreeIPA domain
|
Limited set of synchronized attributes and problematic data mapping
Potential data inconsistency between Active Directory and FreeIPA entries for the same user
Different LDAP versions, synchronization protocols, and other technology differences
Delays in relaying updates between directories
Decreased performance
Security implications of syncing passwords — or administrative complexity for maintaining different passwords for the same user account in difference locations
|
The initial environment configuration for trusts is much more complex than synchronization, but it has advantages in simplifying single sign-on to systems, web applications, or terminals; not requiring additional directory administration; and preserving data integrity.
Table 8.2. Positives and Negatives of Using Trusts
|
Positives of Trusts
|
Negatives of Trusts
|
|---|
Pulls in authentication, group, and authorization data automatically when a user logs in
Allows true single sign-on, with a single stored password and without having to synchronize passwords between directories
Caches data in a local database
Allows users to be entirely defined in a single domain, yet have access to multiple domains
Can be configured without having to know or restructure the underlying directory trees
Allows Kerberos authentication, username/password authentication (which generates a Kerberos ticket), or passwordless logins
|
Has very specific DNS configuration requirements
Can potentially have long wait times to retrieve data when a user initially logs in
Prefers that users be located in a single directory and resources in another
Windows machines cannot be clients of the FreeIPA domain
|
There is no clear migration path from using synchronization to using trusts because the entries already exist in the backend FreeIPA LDAP directory. This means that Active Directory user entries (or all user entries, if FreeIPA users are also synced) are duplicated, which can lead to odd behavior with logins, group associations, and lookups.
8.1.3. Active Directory Users and FreeIPA Features: sudo and Host-Based Access Control Policies
Active Directory users cannot be added directly to a FreeIPA group as a member. This means that policies and configuration in FreeIPA which rely on group associations — such as host-based access control rules and sudo policies — must be configured in a kind of daisy-chain.
The Active Directory user is added to an Active Directory group, then that Active Directory group is added to a FreeIPA external group, which is added as a member to a POSIX group. The sudo, host-based access controls, and other policies are applied against that POSIX group and, ultimately, through nesting memberships applied to the Active Directory user when accessing FreeIPA domain resources.
Active Directory Server Identity Management Server
--------------------------- -------------------------------------------------------
| AD user -> AD group ->| -> | External Group -> POSIX Group -> sudo/HBAC policies |
--------------------------- -------------------------------------------|-----------
^ V
|--------------------------------------------------------------------
Testing tools such as hbactest will not work with trusted users because the trusted user group associations are resolved dynamically, not stored in the FreeIPA directory.
8.1.4. Potential Issues with Group Mapping and SIDs
The wbinfo command can be used to obtain a SID for an Active Directory user and then to display groups associated with that SID.
[root@ipaserver ~]# wbinfo -n ADDOMAIN\\jsmith
S-1-5-21-1689615952-3716327440-3249090444-1104 SID_USER (1)
[root@ipaserver ~]# wbinfo --user-domgroups=S-1-5-21-1689615952-3716327440-3249090444-1104
S-1-5-21-1689615952-3716327440-3249090444-513
S-1-5-21-1689615952-3716327440-3249090444-1106
The same query using id shows only the user information, not the Active Directory group membership information.
[root@ipaserver ~]# id ADDOMAIN\\jsmith
uid=1921801104(jsmith@adexample.com) gid=1921801104(jsmith@adexample.com) groups=1921801104(jsmith@adexample.com)
To work around this, ssh into a FreeIPA client machine as the given Active Directory user. After the first successful login, the Active Directory group memberships are detected and returned in the id search.
[root@ipaserver ~]# id ADDOMAIN\jsmith
uid=1921801107(jsmith@adexample.com) gid=1921801107(jsmith@adexample.com) groups=1921801107(jsmith@adexample.com),129600004(ad_users),1921800513(domain users@adexample.com)
8.1.5. Active Directory Users and FreeIPA Administration
Trust relationships are unidirectional. Active Directory users exist only within the Active Directory domain and are limited to what resources within the FreeIPA domain they can access. Active Directory users are not administrators for FreeIPA because they do not exist within FreeIPA.
Active Directory users, then, cannot use any FreeIPA administrative tools, including the web UI and command-line tools.
8.2. Environment and Machine Requirements to Set Up Trusts
Make sure that both the Active Directory and FreeIPA servers, machines, and environments meet the requirements and settings in this section before configuring a trust agreement.
8.2.1. Domain and Realm Names
The FreeIPA DNS domain name and Kerberos realm name must be different than the Active Directory DNS domain name and Kerberos realm name.
The NetBIOS name is the far-left component of the domain name. For example, if the domain is linux.example.com, the NetBIOS name is linux, while if the domain name is simply example.com, it is example. The NetBIOS name is critical for identifying the Active Directory domain and, if the FreeIPA domain is within a subdomain of Active Directory DNS, for identifying the FreeIPA domain and services.
The FreeIPA domain and Active Directory domain must have different NetBIOS names.
Both the Active Directory server and the FreeIPA server must be configured to run their own respective DNS services.
8.2.4. Firewalls and Ports
Table 8.3. FreeIPA Ports
|
Service
|
Ports
|
Type
|
|---|
|
HTTP/HTTPS
|
|
TCP
|
|
Kerberos
|
|
TCP and UDP
|
|
DNS
|
53
|
TCP and UDP
|
|
NTP
|
123
|
UDP
|
The FreeIPA backend LDAP server must not be reachable by the Active Directory domain controller. The associated ports — 389 and 636 — on the FreeIPA server host must be shut down for the Active Directory domain controller.
[root@ipaserver ]# chkconfig iptables on
Open the iptables configuration file.
[root@ipaserver ~]# vim /etc/sysconfig/iptables
Add the rule to restrict access to LDAP ports for the Active Directory host.
-A INPUT -s ad_ip_address -p tcp -m multiport --dports 389,636 -m state --state NEW,ESTABLISHED -j REJECT
Make sure that there lines to allow access to the TCP and UDP ports required by FreeIPA.
-A INPUT -p tcp -m multiport --dports 80,443,389,636,88,464,53,138,139,445 -m state --state NEW,ESTABLISHED -j ACCEPT
-A INPUT -p udp -m multiport --dports 88,464,53,123,138,139,389,445 -m state --state NEW,ESTABLISHED -j ACCEPT
Save the file.
Restart the iptables service:
[root@ipaserver ]# service iptables restart
Example 8.1. Example iptables Configuration File
*filter
:INPUT ACCEPT [0:0]
:FORWARD ACCEPT [0:0]
:OUTPUT ACCEPT [0:0]
-A INPUT -m state --state ESTABLISHED,RELATED -j ACCEPT
-A INPUT -p icmp -j ACCEPT
-A INPUT -i lo -j ACCEPT
-A INPUT -m state --state NEW -m tcp -p tcp --dport 22 -j ACCEPT
-A INPUT -s ad_ip_address -p tcp -m multiport --dports 389,636 -m state --state NEW,ESTABLISHED -j REJECT
-A INPUT -p tcp -m multiport --dports 80,443,389,636,88,464,53,138,139,445 -m state --state NEW,ESTABLISHED -j ACCEPT
-A INPUT -p udp -m multiport --dports 88,464,53,123,138,139,389,445 -m state --state NEW,ESTABLISHED -j ACCEPT
-A INPUT -p udp -j REJECT
-A INPUT -p tcp -j REJECT
-A FORWARD -j REJECT --reject-with icmp-host-prohibited
COMMIT
Both the Active Directory server and the FreeIPA server must have their clocks in sync.
8.2.6. Supported Username Formats
Username mapping is performed in the local SSSD client. A Python regular expression is used by SSSD to identify the username and the domain to which it belongs.
By default in SSSD, the username format is defined in the form name@domain. This uses the regular expression:
re_expression = (?P<name>[^@]+)@?(?P<domain>[^@]*$)
Active Directory can support several different kinds of name formats, however, so the re_expression parameter in the SSSD configuration file for FreeIPA backends or Active Directory backends uses a more complex expression:
re_expression = (((?P<domain>[^\\]+)\\(?P<name>.+$))|((?P<name>[^@]+)@(?P<domain>.+$))|(^(?P<name>[^@\\]+)$))
This supports usernames in multiple formats:
username
username@domain.name
DOMAIN\username
An additional SSSD parameter, default_domain_suffix, can be used to supply a default domain value for usernames. For example, if all users are in a trusted Active Directory domain of adexample.com and the identity backend is the FreeIPA domain of ipa.example.com, the default_domain_suffix parameter can be set with the value adexample.com. All users are automatically assumed to belong to that user domain unless the domain value is explicitly given with the username.
8.2.7. Trust Can Only Be Configured Once
The ipa-ad-trust-install command can only be run once. If any information is entered incorrectly — particularly the NetBIOS name for the FreeIPA server, but also the administrative credentials or other settings — then the trust services and all FreeIPA packages must be uninstalled and then reinstalled and rerun.
It is not possible to rerun the ipa-ad-trust-install command to change the settings.
8.3. Setting up Trust with FreeIPA as a DNS Subdomain of Active Directory
Stop the Windows firewall service.
Stop iptables and ip6tables on the FreeIPA server.
[root@ipaserver ]# service iptables stop
Install the required trust packages, updated Samba4 packages, and LDAP-DNS packages for FreeIPA DNS management.
[root@ipaserver ]# yum install ipa-server "*ipa-server-trust-ad" samba4-winbind-clients bind-dyndb-ldap samba4-client
The Samba4 packages conflict with the default Samba3 packages on the Red Hat Enterprise Linux system. There may be dependency issues with other applications as the Samba3 packages are removed.
The cifs-utils package is removed when Samba3 is removed. This must be re-installed.
[root@ipaserver ]# yum install cifs-utils
It is recommended that you remove the samba4-winbind-krb5-locator package to improve Kerberos performance.
[root@ipaserver ]# yum remove samba4-winbind-krb5-locator
For a new FreeIPA server. Set up the FreeIPA server to use its own, integrated DNS services (--setup-dns), its own DNS domain (-n), and the Active Directory DNS server as a forwarder (--forwarder). For example:
[root@ipaserver ]# ipa-server-install --setup-dns --forwarder=2555.255.255.255 -p secret -a secret -r IPAEXAMPLE -n linux.adexample.com --hostname ipaserver.linux.adexample.com -U
If the FreeIPA server was set up without using Active Directory as a forwarder. If a FreeIPA server was configured without using Active Directory as a forwarder, then the Active Directory server can be added as a confitional forwarder. This requires the IP address of the Active Directory DNS server.
[root@ipaserver ]# ipa dnsconfig-mod --forwarder=255.255.255.255 --forward-policy=first
Using a first policy means that queries are sent to the forwarder first and then to the local named process. Alternatively, this can be set to only, so that only the DNS forwarder is queried, never named.
Add the FreeIPA domain as a subdomain entry in the Active Directory configuration and add an NS record for the FreeIPA DNS. For this exmaple, the FreeIPA configuration has a NetBIOS name of linux (the subdomain) and the domain name is adexample.com.
Open the command prompt, using Run as Administrator.
Use the dnscmd command to add the A record for the FreeIPA server, using the FreeIPA hostname, NetBIOS name, and IP address.
/RecordAdd ad_domain ipa_hostname.ipa_netbios A ipa_ip_address
For example:
C:\> dnscmd 127.0.0.1 /RecordAdd adexample.com ipaserver.linux A 255.255.255.0
Then add the NS record for the FreeIPA server. This has the format:
/RecordAdd ad_domain ipa_netbios NS ipa_hostname.ipa_subdomain
For example:
C:\> dnscmd 127.0.0.1 /RecordAdd adexample.com linux NS ipaserver.linux.adexample.com
Check the SRV records for both domains from both servers.
On the FreeIPA server, use the dig SRV command to list the records for the Active Directory domain and the FreeIPA domain.
[root@ipaserver ~]# dig SRV _ldap._tcp.adexample.com
;; ANSWER SECTION:
_ldap._tcp.adexample.com. 600 IN SRV 0 100 389 adserver.adexample.com.
;; ADDITIONAL SECTION:
adserver.adexample.com. 3600 IN A 192.168.2.161
;; ADDITIONAL SECTION:
adserver.adexample.com. 3600 IN A 192.168.2.161
[root@ipaserver ~]# dig SRV _ldap._tcp.linux.adexample.com
;; ANSWER SECTION:
_ldap._tcp.linux.adexample.com. 86400 IN SRV 0 100 389 ipaserver.linux.adexample.com.
;; AUTHORITY SECTION:
linux.adexample.com. 86400 IN NS ipaserver.linux.adexample.com.
;; ADDITIONAL SECTION:
ipaserver.linux.adexample.com. 1200 IN A 192.168.2.158
On the Active Directory server, open the nslookup tool and check the corresponding SRV records.
> nslookup
> set type=srv
> _ldap._tcp.adexample.com
> _ldap._tcp.linux.adexample.com
> quit
Enable DNS lookups in the Kerberos realm for the Kerberos client.
Open the /etc/krb5.conf configuration file.
[root@ipaserver ]# vim /etc/krb5.conf
In the [libdefaults] section, add or set the dns_lookup_kdc value to true.
[libdefaults]
....
dns_lookup_kdc = true
Configure the FreeIPA server to enable trust services. This requires the NetBIOS name of the FreeIPA server and the password of the FreeIPA administrator with the -a. Optionally, use the -U argument to run the script non-interactively.
[root@ipaserver ]# ipa-adtrust-install --netbios-name=IPAEXAMPLE -a secret -U
To verify the FreeIPA configuration at this point, use the Samba tools to check that the Windows-related services are running and accessible. The smbclient command shows whether the domain is in the Samba registry.
[root@ipaserver ~]# smbclient -L ipaserver.ipaexample.com -k
lp_load_ex: changing to config backend registry
Domain=[IPAEXAMPLE] OS=[Unix] Server=[Samba 4.0.0rc4]
Sharename Type Comment
--------- ---- -------
IPC$ IPC IPC Service (Samba 4.0.0rc4)
Domain=[IPAEXAMPLE] OS=[Unix] Server=[Samba 4.0.0rc4]
Server Comment
--------- -------
Workgroup Master
--------- -------
The wbinfo command shows whether the FreeIPA domain is online.
[root@ipaserver ~]# wbinfo --online-status
BUILTIN : online
IPAEXAMPLE : online
If there are existing FreeIPA users and groups. For existing FreeIPA users, it is required that all users and groups have an Active Directory-style security identifier (SID). A new ipaNTSecurityIdentifier containing a SID can be created automatically for each entry by running a special ipa-sidgen-task operation on the backend LDAP directory.
If there are no existing FreeIPA users or groups, then this step can be skipped.
[root@ipaserver ]# ldapmodify -x -H ldap://ipaserver.ipaexample.com:389 -D "cn=directory manager" -w Passwd -f
dn: cn=sidgen,cn=ipa-sidgen-task,cn=tasks,cn=config
changetype: add
objectClass: top
objectClass: extensibleObject
cn: sidgen
nsslapd-basedn: dc=ipadomain,dc=com
delay: 0
adding new entry "cn=sidgen,cn=ipa-sidgen-task,cn=tasks,cn=config"
When the task completes successfully, there will be a message in the error logs that the SID generation task (Sidgen task) finished with a status of zero (0).
[root@ipaserver ]# grep "sidgen_task_thread" /var/log/dirsrv/slapd-IPALAB-QE/errors
[20/Jul/2012:18:17:16 +051800] sidgen_task_thread - [file ipa_sidgen_task.c, line 191]: Sidgen task starts ...
[20/Jul/2012:18:17:16 +051800] sidgen_task_thread - [file ipa_sidgen_task.c, line 196]: Sidgen task finished [0].
Create a trust agreement for the Active Directory domain and the FreeIPA domain. This command requires the Active Directory domain and the credentials of an administrative user to use to connect to the domain.
ipa trust-add --type=type ad_domain_name --admin ad_admin_username --password
For example:
[root@ipaserver ~]# ipa trust-add --type=ad adexample.com --admin Administrator --password
Active directory domain administrator's password:
------------------------------------------------------
Added Active Directory trust for realm "adexample.com"
------------------------------------------------------
Realm name: adexample.com
Domain NetBIOS name: ADEXAMPLE
Domain Security Identifier: S-1-5-21-1689615952-3716327440-3249090444
Trust direction: Two-way trust
Trust type: Active Directory domain
Trust status: Established and verified
Request a ticket for a FreeIPA user to check the Kerberos configuration, and then check that that user can request service tickets.
[root@ipaserver ~]# kinit jsmith
First, request service tickets for services within the FreeIPA domain.
[root@ipaserver ]# kvno host/ipaserver.ipaexample.com@IPA.DOMAIN
Then, request service tickets for services within the Active Directory domain.
[root@ipaserver ]# kvno cifs/adserver.adexample.com@AD.DOMAIN
If the Active Directory service ticket is succcessfully granted, then there will be a cross-realm TGT listed with all of the other requested tickets. This will have the name krbtgt/AD.DOMAIN@IPA.DOMAIN.
[root@ipaserver ]# klist
Ticket cache: FILE:/tmp/krb5cc_0
Default principal: jsmith@IPA.DOMAIN
Valid starting Expires Service principal
06/15/12 12:13:04 06/16/12 12:12:55 krbtgt/IPA.DOMAIN@IPA.DOMAIN
06/15/12 12:13:13 06/16/12 12:12:55 host/ipaserver.ipaexample.com@IPA.DOMAIN
06/15/12 12:13:23 06/16/12 12:12:55 krbtgt/AD.DOMAIN@IPA.DOMAIN
06/15/12 12:14:58 06/15/12 22:14:58 cifs/adserver.adexample.com@AD.DOMAIN
This ticket is requested as a FreeIPA user because Kerberos realm mappings are not yet configured to allow Active Directory users to use Kerberos authentication to the realm.
Configure realm mapping in the Kerberos configuration. This allows Kerberos authentication for Active Directory users.
Open the /etc/krb5.conf configuration file.
[root@ipaserver ]# vim /etc/krb5.conf
In the [libdefaults] section, enable DNS lookups in the Kerberos realm.
[libdefaults]
....
dns_lookup_kdc = true
In the [realms] section, identify the FreeIPA realm by name, and then add two auth_to_local lines to define the Kerberos principal name mapping. One rule should have a value of DEFAULT, for standard Unix usernames, and the other should include a rule which maps different Active Directory username formats and the specific Active Directory domain. For example, this rule allows usernames in the format first.last@ADDOMAIN, username@ADDOMAIN[.something], or username@addomain[.something]; the last two expressions allow upper-case or lower-case domain names, since Kerberos is case-sensitive.
[realms]
IDM = {
....
auth_to_local = RULE:[1:$1@$0](^.*@ADDOMAIN$)s/@ADDOMAIN/@addomain/
auth_to_local = DEFAULT
}
Restart the KDC service.
[root@ipaserver ~]# service krb5kdc restart
Configure domain mapping in SSSD.
Open the /etc/sssd/sssd.conf.
[root@ipaserver ]# vim /etc/sssd/sssd.conf
In the [sssd] section, add pac to the services list to enable the SSSD service to request and use Kerberos tickets with PAC data.
[sssd]
services = nss, pam, ssh, pac
....
In the FreeIPA domain section, add the subdomains_provider parameter to explicitly enable SSSD to refer from the configured FreeIPA domain to any domains trusted by that domain.
[domain/ipa.lan]
cache_credentials = True
krb5_store_password_if_offline = True
ipa_domain = example2b.com
id_provider = ipa
auth_provider = ipa
access_provider = ipa
ipa_hostname = ipa2.example.com
chpass_provider = ipa
ipa_server = ipa2.example.com
ldap_tls_cacert = /etc/ipa/ca.crt
subdomains_provider = ipa
The trusted Active Directory domain is not explicitly defined in the SSSD configuration. The FreeIPA domain is automatically created in the SSSD configuration when the client is installed; adding this line makes it possible to use the existing configuration.
Save the changes to the sssd.conf file.
Restart SSSD.
[root@ipaserver ]# service sssd restart
Restart the iptables and ip6tables services on the FreeIPA server.
[root@ipaserver ]# service iptables start
Restart the Windows firewall.
8.4. Setting up Trust with FreeIPA and Active Directory in Different DNS Domains
Stop the Windows firewall service.
Stop iptables and ip6tables on the FreeIPA server.
[root@ipaserver ]# service iptables stop
Install the required trust packages, updated Samba4 packages, and LDAP-DNS packages for FreeIPA DNS management.
[root@ipaserver ]# yum install ipa-server "*ipa-server-trust-ad" samba4-winbind-clients bind-dyndb-ldap samba4-client
The Samba4 packages conflict with the default Samba3 packages on the Red Hat Enterprise Linux system. There may be dependency issues with other applications as the Samba3 packages are removed.
The cifs-utils package is removed when Samba3 is removed. This must be re-installed.
[root@ipaserver ]# yum install cifs-utils
It is recommended that you remove the samba4-winbind-krb5-locator package to improve Kerberos performance.
[root@ipaserver ]# yum remove samba4-winbind-krb5-locator
For a new FreeIPA server. Set up the FreeIPA server to use its own, integrated DNS services (--setup-dns), its own DNS domain (-n), and the Active Directory DNS server as a forwarder (--forwarder). For example:
[root@ipaserver ]# ipa-server-install --setup-dns --forwarder=ad-dns.adserver.example.com -p secret -a secret -r IPAEXAMPLE -n ipaexample.com --hostname ipaserver.ipaexample.com -U
If the FreeIPA server was set up without using Active Directory as a forwarder. If a FreeIPA server was configured without using Active Directory as a forwarder, then the Active Directory server can be added as a confitional forwarder. This requires the IP address of the Active Directory DNS server.
[root@ipaserver ]# ipa dnsconfig-mod --forwarder=255.255.255.255 --forward-policy=first
Using a first policy means that queries are sent to the forwarder first and then to the local named process. Alternatively, this can be set to only, so that only the DNS forwarder is queried, never named.
Add the FreeIPA server as a conditional forwarder in the Active Directory DNS configuration.
Open the Administrative Tools menu, and select the DNS item.
Right-click the Conditional Forwarders item in the left column of the window.
Select the New Conditional Forwarder... button.
Enter the DNS domain name of the FreeIPA domain and the IP address of the FreeIPA DNS server.
Save the new forwarder.
Alternatively, use the dnscmd command-line utility to add the forwarder entry:
> dnscmd 127.0.0.1 /ZoneAdd IPAEXAMPLE.COM /Forwarder 255.255.255.0
Check the SRV records for both domains from both servers.
On the FreeIPA server, use the dig SRV command to list the records for the Active Directory domain and the FreeIPA domain.
[root@ipaserver ~]# dig SRV _ldap._tcp.adexample.com
;; ANSWER SECTION:
_ldap._tcp.adexample.com. 600 IN SRV 0 100 389 adserver.adexample.com.
;; ADDITIONAL SECTION:
adserver.adexample.com. 3600 IN A 192.168.2.161
;; ADDITIONAL SECTION:
adserver.adexample.com. 3600 IN A 192.168.2.161
[root@ipaserver ~]# dig SRV _ldap._tcp.ipaexample.com
;; ANSWER SECTION:
_ldap._tcp.ipaexample.com. 86400 IN SRV 0 100 389 ipaserver.ipaexample.com.
;; AUTHORITY SECTION:
ipaexample.com. 86400 IN NS ipaserver.ipaexample.com.
;; ADDITIONAL SECTION:
ipaserver.ipaexample.com. 1200 IN A 192.168.2.158
On the Active Directory server, open the nslookup tool and check the corresponding SRV records.
> nslookup
> set type=srv
> _ldap._tcp.adexample.com
> _ldap._tcp.ipaexample.com
> quit
Enable DNS lookups in the Kerberos realm for the Kerberos client.
Open the /etc/krb5.conf configuration file.
[root@ipaserver ]# vim /etc/krb5.conf
In the [libdefaults] section, add or set the dns_lookup_kdc value to true.
[libdefaults]
....
dns_lookup_kdc = true
Configure the FreeIPA server to enable trust services. This requires the NetBIOS name of the FreeIPA server and the password of the FreeIPA administrator with the -a. Optionally, use the -U argument to run the script non-interactively.
[root@ipaserver ]# ipa-adtrust-install --netbios-name=IPAEXAMPLE -a secret -U
To verify the FreeIPA configuration at this point, use the Samba tools to check that the Windows-related services are running and accessible. The smbclient command shows whether the domain is in the Samba registry.
[root@ipaserver ~]# smbclient -L ipaserver.ipaexample.com -k
lp_load_ex: changing to config backend registry
Domain=[IPAEXAMPLE] OS=[Unix] Server=[Samba 4.0.0rc4]
Sharename Type Comment
--------- ---- -------
IPC$ IPC IPC Service (Samba 4.0.0rc4)
Domain=[IPAEXAMPLE] OS=[Unix] Server=[Samba 4.0.0rc4]
Server Comment
--------- -------
Workgroup Master
--------- -------
The wbinfo command shows whether the FreeIPA domain is online.
[root@ipaserver ~]# wbinfo --online-status
BUILTIN : online
IPAEXAMPLE : online
If there are existing FreeIPA users and groups. For existing FreeIPA users, it is required that all users and groups have an Active Directory-style security identifier (SID). A new ipaNTSecurityIdentifier containing a SID can be created automatically for each entry by running a special ipa-sidgen-task operation on the backend LDAP directory.
If there are no existing FreeIPA users or groups, then this step can be skipped.
[root@ipaserver ]# ldapmodify -x -H ldap://ipaserver.ipaexample.com:389 -D "cn=directory manager" -w Passwd -f
dn: cn=sidgen,cn=ipa-sidgen-task,cn=tasks,cn=config
changetype: add
objectClass: top
objectClass: extensibleObject
cn: sidgen
nsslapd-basedn: dc=ipadomain,dc=com
delay: 0
adding new entry "cn=sidgen,cn=ipa-sidgen-task,cn=tasks,cn=config"
When the task completes successfully, there will be a message in the error logs that the SID generation task (Sidgen task) finished with a status of zero (0).
[root@ipaserver ]# grep "sidgen_task_thread" /var/log/dirsrv/slapd-IPALAB-QE/errors
[20/Jul/2012:18:17:16 +051800] sidgen_task_thread - [file ipa_sidgen_task.c, line 191]: Sidgen task starts ...
[20/Jul/2012:18:17:16 +051800] sidgen_task_thread - [file ipa_sidgen_task.c, line 196]: Sidgen task finished [0].
Create a trust agreement for the Active Directory domain and the FreeIPA domain. This command requires the Active Directory domain and the credentials of an administrative user to use to connect to the domain.
ipa trust-add --type=type ad_domain_name --admin ad_admin_username --password
For example:
[root@ipaserver ~]# ipa trust-add --type=ad adexample.com --admin Administrator --password
Active directory domain administrator's password:
------------------------------------------------------
Added Active Directory trust for realm "adexample.com"
------------------------------------------------------
Realm name: adexample.com
Domain NetBIOS name: ADEXAMPLE
Domain Security Identifier: S-1-5-21-1689615952-3716327440-3249090444
Trust direction: Two-way trust
Trust type: Active Directory domain
Trust status: Established and verified
Request a ticket for a FreeIPA user to check the Kerberos configuration, and then check that that user can request service tickets.
[root@ipaserver ~]# kinit jsmith
First, request service tickets for services within the FreeIPA domain.
[root@ipaserver ]# kvno host/ipaserver.ipaexample.com@IPA.DOMAIN
Then, request service tickets for services within the Active Directory domain.
[root@ipaserver ]# kvno cifs/adserver.adexample.com@AD.DOMAIN
If the Active Directory service ticket is succcessfully granted, then there will be a cross-realm TGT listed with all of the other requested tickets. This will have the name krbtgt/AD.DOMAIN@IPA.DOMAIN.
[root@ipaserver ]# klist
Ticket cache: FILE:/tmp/krb5cc_0
Default principal: jsmith@IPA.DOMAIN
Valid starting Expires Service principal
06/15/12 12:13:04 06/16/12 12:12:55 krbtgt/IPA.DOMAIN@IPA.DOMAIN
06/15/12 12:13:13 06/16/12 12:12:55 host/ipaserver.ipaexample.com@IPA.DOMAIN
06/15/12 12:13:23 06/16/12 12:12:55 krbtgt/AD.DOMAIN@IPA.DOMAIN
06/15/12 12:14:58 06/15/12 22:14:58 cifs/adserver.adexample.com@AD.DOMAIN
This ticket is requested as a FreeIPA user because Kerberos realm mappings are not yet configured to allow Active Directory users to use Kerberos authentication to the realm.
Configure realm mapping in the Kerberos configuration. This allows Kerberos authentication for Active Directory users.
Open the /etc/krb5.conf configuration file.
[root@ipaserver ]# vim /etc/krb5.conf
In the [libdefaults] section, enable DNS lookups in the Kerberos realm.
[libdefaults]
....
dns_lookup_kdc = true
In the [realms] section, identify the FreeIPA realm by name, and then add two auth_to_local lines to define the Kerberos principal name mapping. One rule should have a value of DEFAULT, for standard Unix usernames, and the other should include a rule which maps different Active Directory username formats and the specific Active Directory domain. For example, this rule allows usernames in the format first.last@ADDOMAIN, username@ADDOMAIN[.something], or username@addomain[.something]; the last two expressions allow upper-case or lower-case domain names, since Kerberos is case-sensitive.
[realms]
IDM = {
....
auth_to_local = RULE:[1:$1@$0](^.*@ADDOMAIN$)s/@ADDOMAIN/@addomain/
auth_to_local = DEFAULT
}
Restart the KDC service.
[root@ipaserver ~]# service krb5kdc restart
Configure domain mapping in SSSD.
Open the /etc/sssd/sssd.conf.
[root@ipaserver ]# vim /etc/sssd/sssd.conf
In the [sssd] section, add pac to the services list to enable the SSSD service to request and use Kerberos tickets with PAC data.
[sssd]
services = nss, pam, ssh, pac
....
In the FreeIPA domain section, add the subdomains_provider parameter to explicitly enable SSSD to refer from the configured FreeIPA domain to any domains trusted by that domain.
[domain/ipa.lan]
cache_credentials = True
krb5_store_password_if_offline = True
ipa_domain = example2b.com
id_provider = ipa
auth_provider = ipa
access_provider = ipa
ipa_hostname = ipa2.example.com
chpass_provider = ipa
ipa_server = ipa2.example.com
ldap_tls_cacert = /etc/ipa/ca.crt
subdomains_provider = ipa
The trusted Active Directory domain is not explicitly defined in the SSSD configuration. The FreeIPA domain is automatically created in the SSSD configuration when the client is installed; adding this line makes it possible to use the existing configuration.
Save the changes to the sssd.conf file.
Restart SSSD.
[root@ipaserver ]# service sssd restart
Restart the iptables and ip6tables services on the FreeIPA server.
[root@ipaserver ]# service iptables start
Restart the Windows firewall.
8.5. Creating FreeIPA Groups for Active Directory Users
User groups are required to set access permissions, host-based access control, sudo rules, and other controls on FreeIPA users. These groups are what grant access to FreeIPA domain resources, as well as restricting access.
However, Active Directory users cannot be added directly to FreeIPA user groups. This means that Active Directory users require special configuration in order to access FreeIPA domain resources.
As described in
Section 8.1.1.4, “Kerberos Realms, Authentication, and Authorization”, Active Directory users are added to the FreeIPA domain in a kind of daisy chain. They are added to a group on the Active Directory side, then that group is added to a FreeIPA external group (meaning, a non-POSIX group), and then that external group is added to a local POSIX group as a member. The FreeIPA POSIX group can then be used for user/role management of Active Directory users.
Create or select the group in the Active Directory domain to use to manage Active Directory users in the FreeIPA realm. (Multiple groups can be used and added to different groups on the FreeIPA side.)
Create an external group in the FreeIPA domain for the Active Directory users. This correlates to the Active Directory group. Using the --external argument indicates that this group will contain members from outside the FreeIPA domain. For example:
[root@ipaserver ~]# ipa group-add --desc='AD users external map' ad_users_external --external
-------------------------------
Added group "ad_users_external"
-------------------------------
Group name: ad_users_external
Description: AD users external map
Create the POSIX group for actually administering the FreeIPA policies.
[root@ipaserver ~]# ipa group-add --desc='AD users' ad_users
----------------------
Added group "ad_users"
----------------------
Group name: ad_users
Description: AD users
GID: 129600004
Add the Active Directory group to the FreeIPA external group as an external member. The Active Directory group is identified by the name DOMAIN\group_name. The group name is then mapped to the Active Directory SID for the group. For example:
[root@ipaserver ~]# ipa group-add-member ad_users_external --external "AD\Domain Users"
[member user]:
[member group]:
Group name: ad_users_external
Description: AD users external map
External member: S-1-5-21-3655990580-1375374850-1633065477-513 SID_DOM_GROUP (2)
-------------------------
Number of members added 1
-------------------------
Add the external FreeIPA group to the POSIX FreeIPA group as a member. For example:
[root@ipaserver ~]# ipa group-add-member ad_users --groups ad_users_external
Group name: ad_users
Description: AD users
GID: 129600004
Member groups: ad_users_external
-------------------------
Number of members added 1
-------------------------
8.6. Using SSH from Active Directory Machines for FreeIPA Resources
When a trust is configured, Active Directory users can access machines, services, and files on FreeIPA hosts using SSH and their Active Directory credentials.
One critical factor when using SSH is the username. The username must meet several criteria:
The username must have the format ad_user@ad_domain.
The domain name itself must be lower-case. This is required for Kerberos principal mapping.
The case of the username must match, exactly, the case of the username in Active Directory. jsmith and JSmith are considered different users because of the different cases.
When using PuTTY on the Windows machine, make sure that GSS-API credential delegation is enabled.
8.7. Using Trust with Kerberized Web Applications
Any existing web application can be configured to use Kerberos authentication, which references the trusted Active Directory and FreeIPA Kerberos realms.
For example, for an Apache server, set the KrbAuthRealms directive for the application location to the name of the FreeIPA domain and set the location for the keytab (Krb5Keytab). Also set other paramters to enable Kerberos authentication, the service name used for the keytab (HTTP), and the Kerberos methods (which enables password-based authentication for valid users).
<Location "/mywebapp">
AuthType Kerberos
AuthName "IPA Kerberos authentication"
KrbMethodNegotiate on
KrbMethodK5Passwd on
KrbServiceName HTTP
KrbAuthRealms IDM_DOMAIN
Krb5Keytab /etc/httpd/conf/ipa.keytab
KrbSaveCredentials off
Require valid-user
</Location>
The Kerberos configuration directives are covered in the
mod_auth_kerb module man pages.
After changing the Apache application configuration, restart the Apache service:
[root@ipaserver ~]# service httpd restart
Chapter 9. Identity: Integrating with Microsoft Active Directory Through Synchronization
FreeIPA uses active synchronization to integrate user data stored in an Active Directory domain and the user data stored in the FreeIPA domain. Critical user attributes, including passwords, are synchronized between the services.
The capability to sync Active Directory and FreeIPA domains is inherent when a FreeIPA server is first installed. The synchronization process is configured by creating agreements between the FreeIPA server and the Active Directory domain controller.
This chapter describes how to configure synchronization, how to configure Active Directory for integration with FreeIPA, and how to configure Windows systems within the Active Directory domain to be aware of the FreeIPA domain.
9.1. About Active Directory and FreeIPA
Within the FreeIPA domain, information is shared among servers and replicas by copying that information, reliably and predictably, between the data masters (servers) and other data masters. This process is replication.
A similar process can be used to share data between the FreeIPA domain and a Microsoft Active Directory domain. This is synchronization.
Synchronization is the process of copying data back and forth between Active Directory and FreeIPA.
Synchronization is defined in an agreement between a FreeIPA server and an Active Directory domain controller. The sync agreement defines all of the information required to identify sync-able user entries (like the subtree to synchronize and requisite object classes in the user entries) as well as defining how account attributes are handled. The sync agreements are created with default values which can be tweaked to meet the needs of a specific domain. When two servers are involved in synchronization, they are called peers.
Synchronization is most commonly bi-directional. Information is sent back and forth between the FreeIPA domain and the Windows domain in a process that is very similar to how FreeIPA servers and replicas share information among themselves. It is possible to configure synchronization — or certain data areas — to only sync one way. That is uni-directional synchronization.
To prevent the risk of data conflicts, synchronization is configured between one FreeIPA server and one Active Directory domain controller. The FreeIPA server propagates changes back to the FreeIPA domain, while the domain controller propagates changes back to the Windows domain.
There are some key features to FreeIPA synchronization:
A synchronization operation runs every five minutes.
Synchronization can only be configured with one Active Directory domain. Multiple domains are not supported.
Synchronization can only be configured with one Active Directory domain controller. However, it is possible to have a list of failover Active Directory domain controllers. Likewise, there can be a list of failover FreeIPA servers to keep synchronization uninterrupted.
Only user information is synchronized.
Both user attributes and passwords can be synchronized.
While modifications are bi-directional (going both from Active Directory to FreeIPA and from FreeIPA to Active Directory), creating or adding accounts are only uni-directional, from Active Directory to FreeIPA. New accounts created in Active Directory are synchronized over to FreeIPA automatically. However, user accounts created in FreeIPA must also be created in Active Directory before they will be synchronized.
Account lock information is synchronized by default, so a user account which is disabled in one domain is disabled in the other.
Password synchronization changes take effect immediately.
When Active Directory users are synchronized over to FreeIPA, certain attributes (including Kerberos and POSIX attributes) will have IPA attributes are automatically added to the user entries. These attributes are used by FreeIPA within its domain. They are not synchronized back over the corresponding Active Directory user entry.
Some of the data in synchronization can be modified as part of the synchronization process. For examples, certain attributes can be automatically added to Active Directory user accounts when they are synced over to the FreeIPA domain. These attribute changes are defined as part of the synchronization agreement and are described in
Section 9.4.3, “Changing the Behavior for Syncing User Account Attributes”.
9.2. About Synchronized Attributes
FreeIPA synchronizes a subset of user attributes between FreeIPA and Active Directory user entries. Any other attributes present in the entry, either in FreeIPA or in Active Directory, are ignored by synchronization.
Most POSIX attributes are not synchronized.
Although there are significant schema differences between the Active Directory LDAP schema and the 389 Directory Server LDAP schema used by FreeIPA, there are many attributes that are the same. These attributes are simply synchronized between the Active Directory and FreeIPA user entries, with no changes to the attribute name or value format.
User Schema That Are the Same in FreeIPA and Windows Servers
Some attributes have different names but still have direct parity between FreeIPA (which uses 389 Directory Server) and Active Directory. These attributes are mapped by the synchronization process.
Table 9.1. User Schema Mapped between FreeIPA and Active Directory
|
FreeIPA
|
Active Directory
|
|---|
|
cn
|
name
|
|
nsAccountLock
|
userAccountControl
|
|
ntUserDomainId
|
sAMAccountName
|
|
ntUserHomeDir
|
homeDirectory
|
|
ntUserScriptPath
|
scriptPath
|
|
ntUserLastLogon
|
lastLogon
|
|
ntUserLastLogoff
|
lastLogoff
|
|
ntUserAcctExpires
|
accountExpires
|
|
ntUserCodePage
|
codePage
|
|
ntUserLogonHours
|
logonHours
|
|
ntUserMaxStorage
|
maxStorage
|
|
ntUserProfile
|
profilePath
|
|
ntUserParms
|
userParameters
|
|
ntUserWorkstations
|
userWorkstations
|
9.2.1. User Schema Differences between FreeIPA and Active Directory
Even though attributes may be successfully synced between Active Directory and FreeIPA, there may still be differences in how Active Directory and FreeIPA define the underlying X.500 object classes. This could lead to differences in how the data are handled in the different LDAP services.
This section describes the differences in how Active Directory and FreeIPA handle some of the attributes which can be synchronized between the two domains.
9.2.1.1. Values for cn Attributes
In 389 Directory Server, the cn attribute can be multi-valued, while in Active Directory this attribute must have only a single value. When the FreeIPA cn attribute is synchronized, then, only one value is sent to the Active Directory peer.
What this means for synchronization is that,potentially, if a cn value is added to an Active Directory entry and that value is not one of the values for cn in FreeIPA, then all of the FreeIPA cn values are overwritten with the single Active Directory value.
One other important difference is that Active Directory uses the cn attribute as its naming attribute, where FreeIPA uses uid. This means that there is the potential to rename the entry entirely (and accidentally) if the cn attribute is edited in the FreeIPA. If that cn change is written over to the Active Directory entry, then the entry is renamed, and the new named entry is written back over to FreeIPA.
9.2.1.2. Values for street and streetAddress
Active Directory uses the attribute streetAddress for a user's postal address; this is the way that 389 Directory Server uses the street attribute. There are two important differences in the way that Active Directory and FreeIPA use the streetAddress and street attributes, respectively:
In 389 Directory Server, streetAddress is an alias for street. Active Directory also has the street attribute, but it is a separate attribute that can hold an independent value, not an alias for streetAddress.
Active Directory defines both streetAddress and street as single-valued attributes, while 389 Directory Server defines street as a multi-valued attribute, as specified in RFC 4519.
Because of the different ways that 389 Directory Server and Active Directory handle streetAddress and street attributes, there are two rules to follow when setting address attributes in Active Directory and FreeIPA:
The synchronization process maps streetAddress in the Active Directory entry to street in FreeIPA. To avoid conflicts, the street attribute should not be used in Active Directory.
Only one FreeIPA street attribute value is synced to Active Directory. If the streetAddress attribute is changed in Active Directory and the new value does not already exist in FreeIPA, then all street attribute values in FreeIPA are replaced with the new, single Active Directory value.
9.2.1.3. Constraints on the initials Attribute
For the initials attribute, Active Directory imposes a maximum length constraint of six characters, but 389 Directory Server does not have a length limit. If an initials attribute longer than six characters is added to FreeIPA, the value is trimmed when it is synchronized with the Active Directory entry.
9.2.1.4. Requiring the surname (sn) Attribute
Active Directory allows person entries to be created without a surname attribute. However, RFC 4519 defines the person object class as requiring a surname attribute, and this is the definition used in Directory Server.
If an Active Directory person entry is created without a surname attribute, that entry will not be synced over to FreeIPA since it fails with an object class violation.
9.2.2. Active Directory Entries and RFC 2307 Attributes
Windows uses unique, random security IDs (SIDs) to identify users. These SIDs are assigned in blocks or ranges, identifying different system user types within the Windows domain. When users are synchronized between FreeIPA and Active Directory, Windows SIDs for users are mapped to the Unix UIDs used by the FreeIPA entry. Another way of saying this is that the Windows SID is the only ID within the Windows entry which is used as an identifier in the corresponding Unix entry, and then it is used in a mapping.
When Active Directory domains interact with Unix-style applications or domains, then the Active Directory domain may use Services for Unix or IdM for Unix to enable Unix-style
uidNumber and
gidNumber attributes. This allows Windows user entries to follow the specifications for those attributes in
RFC 2307.
However, the uidNumber and gidNumber attributes are not actually used as the uidNumber and gidNumber attributes for the FreeIPA entry. The FreeIPA uidNumber and gidNumber attributes are generated when the Windows user is synced over.
The uidNumber and gidNumber attributes defined and used in FreeIPA are not the same uidNumber and gidNumber attributes defined and used in the Active Directory entry, and the numbers are not related.
9.3. Setting up Active Directory for Synchronization
Synchronizing user accounts alone is enabled within FreeIPA, so all that is necessary is to set up a sync agreement (
Section 9.4.2, “Creating Synchronization Agreements”). On the Windows server, it is necessary to create the user that the FreeIPA server will use to connect to the Active Directory domain.
Grant the sync user account Replicating directory changes rights to the synchronized Active Directory subtree. Replicator rights are required for the sync user to perform synchronization operations.
Add the sync user as a member of the Account Operator and Enterprise Read-Only Domain controller groups. It is not necessary for the user to belong to the full Domain Admin group.
9.4. Managing Synchronization Agreements
9.4.1. Trusting the Active Directory and FreeIPA CA Certificates
Both Active Directory and FreeIPA use certificates for server authentication. For the Active Directory and FreeIPA SSL server certificates to be trusted by each other, both servers need to trust the CA certificate for the CA which issued those certificates. This means that the Active Directory CA certificate needs to be imported into the FreeIPA database, and the FreeIPA CA certificate needs to be imported into the Active Directory database.
On the Active Directory server, download the FreeIPA server's CA certificate from http://ipa.example.com/ipa/config/ca.crt.
Install the FreeIPA CA certificate in the Active Directory certificate database. This can be done using the Microsoft Management Console or the
certutil utility. For example:
certutil -installcert -v -config "ipaserver.example.com\Example Domain CA" c:\path\to\ca.crt
For more details, see the Active Directory documentation.
Export the Active Directory CA certificate.
In My Network Places, open the CA distribution point. For example, the location on Windows Server 2003 is C:\WINDOWS\system32\certsrv\CertEnroll\.
Double-click the security certificate file (.crt file) to display the Certificate dialog box.
On the Details tab, click Copy to File to start the Certificate Export Wizard.
Click Next, and then select Base-64 encoded X.509 (.CER).
Specify a suitable directory and file name for the exported file. Click Next to export the certificate, and then click Finish.
Copy the Active Directory certificate over to the FreeIPA server machine.
Download the FreeIPA server's CA certificate from http://ipa.example.com/ipa/config/ca.crt.
Copy both the Active Directory CA certificate and the FreeIPA CA certificate into the /etc/openldap/cacerts/ directory.
Update the hash symlinks for the certificates.
cacertdir_rehash /etc/openldap/cacerts/
Edit the /etc/openldap/ldap.conf file, and add the information to point to and use the certificates in the /etc/openldap/cacerts/ directory.
TLS_CACERTDIR /etc/openldap/cacerts/
TLS_REQCERT allow
9.4.2. Creating Synchronization Agreements
Synchronization agreements are created on the FreeIPA server using the
ipa-replica-manage connect command because it creates a
connection to the Active Directory domain. The options to create the synchronization agreement are listed in
Table 9.2, “Synchronization Agreement Options”.
Remove any existing Kerberos credentials on the FreeIPA server.
$ kdestroy
Use the ipa-replica-manage command to create a Windows synchronization agreement. This requires the --winsync option. If passwords will be synchronized as well as user accounts, then also use the --passsync option and set a password to use for Password Sync.
The --binddn and--bindpwd options give the username and password of the system account on the Active Directory server that FreeIPA will use to connect to the Active Directory server.
$ ipa-replica-manage connect --winsync
--binddn cn=administrator,cn=users,dc=example,dc=com
--bindpw Windows-secret
--passsync secretpwd
--cacert /etc/openldap/cacerts/windows.cer
adserver.example.com -v
When prompted, enter the Directory Manager password.
Table 9.2. Synchronization Agreement Options
|
Option
|
Description
|
|---|
|
--winsync
|
Identifies this as a synchronization agreement.
|
|
--binddn
|
Gives the full user DN of the synchronization identity. This is the user DN that the FreeIPA LDAP server uses to bind to Active Directory. This user must exist in the Active Directory domain and must have replicator, read, search, and write permissions on the Active Directory subtree.
|
|
--bindpw
|
Gives the password for the sync user.
|
|
--passsync
|
Gives the password for the Windows user account which is involved in synchronization.
|
|
--cacert
|
Gives the full path and file name of the Active Directory CA certificate. This certificate is exported in Section 9.4.1, “Trusting the Active Directory and FreeIPA CA Certificates”.
|
|
--win-subtree
|
Gives the DN of the Windows subtree containing the users to synchronize. The default value is cn=Users,$SUFFIX.
|
|
AD_server_name
|
Gives the hostname of the Active Directory domain controller.
|
9.4.3. Changing the Behavior for Syncing User Account Attributes
When the sync agreement is created, it has certain default behaviors defined for how the synchronization process handled the user account attributes during synchronization. The types of behaviors are things like how to handle lockout attributes or how to handle different DN formats. This behavior can be changed by editing the synchronization agreement. The list of attribute-related parameters are in
Table 9.3, “Synced Attribute Settings”.
The sync agreement exists as a special plug-in entry in the LDAP server and each attribute behavior is set through an LDAP attribute. To change the sync behavior, use the ldapmodify command to modify the LDAP server entry directly.
For example, account lockout attributes are synchronized between FreeIPA and Active Directory by default, but this can be disabled by editing the ipaWinSyncAcctDisable attribute. (Changing this means that if an account is disabled in Active Directory, it is still active in FreeIPA and vice versa.)
[jsmith@ipaserver ~]$ ldapmodify -x -D "cn=directory manager" -w password
dn: cn=ipa-winsync,cn=plugins,cn=config
changetype: modify
replace: ipaWinSyncAcctDisable
ipaWinSyncAcctDisable: none
modifying entry "cn=ipa-winsync,cn=plugins,cn=config"
Table 9.3. Synced Attribute Settings
|
Parameter
|
Description
|
Possible Values
|
|---|
|
General User Account Parameters
| |
|
ipaWinSyncNewEntryFilter
|
Sets the search filter to use to find the entry which contains the list of object classes to add to new user entries.
|
The default is (cn=ipaConfig).
|
|
ipaWinSyncNewUserOCAttr
|
Sets the attribute in the configuration entry which actually contains the list of object classes to add to new user entries.
|
The default is ipauserobjectclasses.
|
|
ipaWinSyncHomeDirAttr
|
Identifies which attribute in the entry contains the default location of the POSIX home directory.
|
The default is ipaHomesRootDir.
|
|
ipaWinSyncUserAttr
|
Sets an additional attribute with a specific value to add to Active Directory users when they are synced over from the Active Directory domain. If the attribute is multi-valued, then it can be set multiple times, and the sync process adds all of the values to the entry.
This only sets the attribute value if the entry does not already have that attribute present. If the attribute is present, then the entry's value is used when the Active Directory entry is synced over.
|
ipaWinSyncUserAttr: attributeName attributeValue
|
|
ipaWinSyncUserFlatten
|
Sets whether to normalize the DN of Active Directory entries to conform with the FreeIPA directory structure. In FreeIPA, all users are stored under the cn=users,cn=accounts,$SUFFIX entry, but Active Directory can have more branches in its directory, which can result in DNs like cn=John Smith,ou=Development,ou=Engineering,cn=users,dc=example,dc=com. Flattening the DN discards any additional intervening organizational units in the Active Directory DN and creating a simple DN on the FreeIPA side.
Any ou attributes are stored in the FreeIPA user entry.
|
true | false
|
|
ipaWinSyncForceSync
|
Sets whether to check existing FreeIPA users which match an existing Active Directory user should be automatically edited so they can be synchronized. If a FreeIPA user account has a uid parameter which is identical to the samAccountName in an existing Active Directory user, then that account is not synced by default. This attribute tells the sync service to add the ntUser and ntUserDomainId to the FreeIPA user entries automatically, which allows them to be synchronized.
|
true | false
|
|
User Account Lock Parameters
| |
|
ipaWinSyncAcctDisable
|
Sets which way to synchronize account lockout attributes. It is possible to control which account lockout settings are in effect. For example, to_ad means that when account lockout attribute is set in FreeIPA, its value is synced over to Active Directory and overrides the local Active Directory value. By default, account lockout attributes are synced from both domains.
|
both (default)
to_ad
to_ds
none
|
|
ipaWinSyncInactivatedFilter
|
Sets the search filter to use to find the DN of the group used to hold inactivated (disabled) users. This does not need to be changed in most deployments.
|
The default is (&(cn=inactivated)(objectclass=groupOfNames)).
|
|
ipaWinSyncActivatedFilter
|
Sets the search filter to use to find the DN of the group used to hold active users. This does not need to be changed in most deployments.
|
The default is (&(cn=activated)(objectclass=groupOfNames)).
|
|
Group Parameters
| |
|
ipaWinSyncDefaultGroupAttr
|
Sets the attribute in the new user account to reference to see what the default group for the user is. The group name in the entry is then used to find the gidNumber for the user account.
|
The default is ipaDefaultPrimaryGroup.
|
|
ipaWinSyncDefaultGroupFilter
|
Sets the search filter to map the group name to the POSIX gidNumber.
|
The default is (&(gidNumber=*)(objectclass=posixGroup)(cn=groupAttr_value)).
|
|
Realm Parameters
| |
|
ipaWinSyncRealmAttr
|
Sets the attribute which contains the realm name in the realm entry.
|
The default is cn.
|
|
ipaWinSyncRealmFilter
|
Sets the search filter to use to find the entry which contains the FreeIPA realm name.
|
The default is (objectclass=krbRealmContainer).
|
9.4.4. Changing the Synchronized Windows Subtree
Creating a synchronization agreement automatically sets the two subtrees to use as the synchronized user database. In FreeIPA, the default is cn=users,cn=accounts,$SUFFIX, and for Active Directory, the default is CN=Users,$SUFFIX.
The value for the Active Directory subtree can be set to a non-default value when the sync agreement is created by using the --win-subtree option. After the agreement is created, the Active Directory subtree can be changed by using the ldapmodify command to edit the nsds7WindowsReplicaSubtree value in the sync agreement entry.
Get the name of the sync agreement, using ldapsearch. This search returns only the values for the dn and nsds7WindowsReplicaSubtree attributes instead of the entire entry.
[jsmith@ipaserver ~]$ ldapsearch -xLLL -D "cn=directory manager" -w password -p 389 -h ipaserver.example.com -b cn=config objectclass=nsdswindowsreplicationagreement dn nsds7WindowsReplicaSubtree
dn: cn=meToWindowsBox.example.com,cn=replica,cn=dc\3Dexample\2Cdc\3Dcom,cn=mapping tree,cn=config
nsds7WindowsReplicaSubtree: cn=users,dc=example,dc=com
... 8< ...
Modify the sync agreement
[jsmith@ipaserver ~]$ ldapmodify -x -D "cn=directory manager" -W -p 389 -h ipaserver.example.com <<EOF
dn: cn=meToWindowsBox.example.com,cn=replica,cn=dc\3Dexample\2Cdc\3Dcom,cn=mapping tree,cn=config
changetype: modify
replace: nsds7WindowsReplicaSubtree
nsds7WindowsReplicaSubtree: cn=alternateusers,dc=example,dc=com
EOF
modifying entry "cn=meToWindowsBox.example.com,cn=replica,cn=dc\3Dexample\2Cdc\3Dcom,cn=mapping tree,cn=config"
The new subtree setting takes effect immediately. If a sync operation is currently running, then it takes effect as soon as the current operation completes.
9.4.5. Configuring Uni-Directional Sync
By default, all modifications and deletions are bi-directional. A change in Active Directory is synced over to FreeIPA, and a change to an entry in FreeIPA is synced over to Active Directory. This is essentially an equitable, multi-master relationship, where both Active Directory and FreeIPA are equal peers in synchronization and are both data masters.
However, there can be some data structure or IT designs where only one domain should be a data master and the other domain should accept updates. This changes the sync relationship from a multi-master relationship (where the peer servers are equal) to a master-consumer relationship.
This is done by setting the oneWaySync parameter on the sync agreement. The possible values are fromWindows (for Active Directory to FreeIPA sync) and toWindows (for FreeIPA to Active Directory sync).
For example, to sync changes from Active Directory to FreeIPA:
[jsmith@ipaserver ~]$ ldapmodify -x -D "cn=directory manager" -w password -p 389 -h ipaserver.example.com
dn: cn=ipa-winsync,cn=plugins,cn=config
changetype: modify
add: oneWaySync
oneWaySync: fromWindows
Enabling uni-directional sync does not automatically prevent changes on the un-synchronized server, and this can lead to inconsistencies between the sync peers between sync updates. For example, uni-directional sync is configured to go from Active Directory to FreeIPA, so Active Directory is (in essence) the data master. If an entry is modified or even deleted on the FreeIPA, then the FreeIPA information is different then the information and those changes are never carried over to Active Directory. During the next sync update, the edits are overwritten on the Directory Server and the deleted entry is re-added.
9.4.6. Deleting Synchronization Agreements
Synchronization can be stopped by deleting the sync agreement which disconnects the FreeIPA and Active Directory servers. In the inverse of creating a sync agreement, deleting a sync agreement uses the ipa-replica-manage disconnect command and then the hostname of the Active Directory server.
Delete the sync agreement.
# ipa-replica-manage disconnect adserver.example.com
Remove the Active Directory CA certificate from the FreeIPA server database:
# certutil -D -d /etc/dirsrv/slapd-EXAMPLE.COM/ -n "Imported CA"
9.4.7. Winsync Agreement Failures
This can occur if the wrong Active Directory CA certificate was specified when the agreement was created. This creates duplicate certificates in the FreeIPA LDAP database (in the
/etc/dirsrv/slapd-DOMAIN/ directory) with the name
Imported CA. This can be checked using
certutil:
$ certutil -L -d /etc/dirsrv/slapd-DOMAIN/
Certificate Nickname Trust Attributes
SSL,S/MIME,JAR/XPI
CA certificate CTu,u,Cu
Imported CA CT,,C
Server-Cert u,u,u
Imported CA CT,,C
To resolve this issue, clear the certificate database:
# certutil -d /etc/dirsrv/slapd-DOMAIN-NAME -D -n "Imported CA"
This deletes the CA certificate from the LDAP database.
"Windows PassSync entry exists, not resetting password"
This is not an error. This message occurs when an exempt user, the Password Sync user, it not being changed. The Password Sync user is the operational user which is used by the service to change the passwords in FreeIPA.
9.5. Managing Password Synchronization
Password synchronization is configured separately from Windows Synchronization.
9.5.1. Setting up the Windows Server for Password Synchronization
Synchronizing passwords requires two things:
The Password Sync Service records password changes and synchronizes them, over a secure connection, to the FreeIPA entry.
Install the Microsoft Certificate System in Enterprise Root Mode. Active Directory will then automatically enroll to retrieve its SSL server certificate.
Make sure that the Active Directory password complexity policies are enabled so that the Password Sync service will run.
Run secpol.msc from the command line.
Select .
Open , and then open .
Enable the Password must meet complexity requirements option and save.
If SSL is not already enabled, set up SSL on the Active Directory server. Setting up LDAPS is explained in more detail in the Microsoft knowledgebase at
http://support.microsoft.com/kb/321051.
Install a certificate authority in the Windows Components section in Add/Remove Programs.
Select the Enterprise Root CA option.
Reboot the Active Directory server. If IIS web services are running, the CA certificate can be accessed by opening http://servername/certsrv.
Set up the Active Directory server to use the SSL server certificate.
Create a certificate request .inf, using the fully-qualified domain name of the Active Directory as the certificate subject. For example:
;----------------- request.inf -----------------
[Version]
Signature="$Windows NT$
[NewRequest]
Subject = "CN=ad.server.example.com, O=Engineering, L=Raleigh, S=North Carolina, C=US"
KeySpec = 1
KeyLength = 2048
Exportable = TRUE
MachineKeySet = TRUE
SMIME = False
PrivateKeyArchive = FALSE
UserProtected = FALSE
UseExistingKeySet = FALSE
ProviderName = "Microsoft RSA SChannel Cryptographic Provider"
ProviderType = 12
RequestType = PKCS10
KeyUsage = 0xa0
[EnhancedKeyUsageExtension]
OID=1.3.6.1.5.5.7.3.1
;-----------------------------------------------
Generate the certificate request.
certreq -new request.inf request.req
Submit the request to the Active Directory CA. For example:
certreq -submit request.req certnew.cer
If the command-line tool returns an error message, then use the Web browser to access the CA and submit the certificate request. If IIS is running, then the CA URL is http://servername/certsrv.
Accept the certificate request. For example:
certreq -accept certnew.cer
Make sure that the server certificate is present on the Active Directory server.
In the menu, click , then click and .
Import the CA certificate from Directory Server into Active Directory. Click Trusted Root CA, then Import, and browse for the Directory Server CA certificate.
Reboot the domain controller.
9.5.2. Setting up Password Synchronization
Install the Password Sync Service on every domain controller in the Active Directory domain in order to synchronize Windows passwords.
There are two PassSync packages available, one for 32-bit Windows servers and one for 64-bit. Make sure to select the appropriate packages for your Windows platform.
Double-click the PassSync.msi file to install it.
The Password Sync Setup window appears. Hit Next to begin installing.
Fill in the information to establish the connection to the FreeIPA server.
The FreeIPA server connection information, including the hostname and secure port number.
The username of the system user which Active Directory uses to connect to the FreeIPA machine. This account is configured automatically when sync is configured on the FreeIPA server. The default account is uid=passsync,cn=sysaccounts,cn=etc,dc=example,dc=com.
The password set in the --passsync option when the sync agreement was created.
The search base for the people subtree on the FreeIPA server. The Active Directory server connects to the FreeIPA server similar to an ldapsearch or replication operation, so it has to know where in the FreeIPA subtree to look for user accounts. The user subtree is cn=users,cn=accounts,dc=example,dc=com.
The certificate token is not used at this time, so that field should be left blank.
Hit Next, then Finish to install Password Sync.
Import the FreeIPA server's CA certificate into the Active Directory certificate store.
Download the FreeIPA server's CA certificate from http://ipa.example.com/ipa/config/ca.crt.
Copy the FreeIPA CA certificate to the Active Directory server.
Install the FreeIPA CA certificate in the Password Sync database. For example:
cd "C:\Program Files\389 Directory Password Synchronization"
certutil.exe -d . -A -n "IPASERVER.EXAMPLE.COM IPA CA" -t CT,, -a -i ipaca.crt
Reboot the Windows machine to start Password Sync.
The Windows machine must be rebooted. Without the rebooting, PasswordHook.dll is not enabled, and password synchronization will not function.
The first attempt to synchronize passwords, which happened when the Password Sync application is installed, will always fail because of the SSL connection between the Directory Server and Active Directory sync peers. The tools to create the certificate and key databases is installed with the .msi.
9.5.3. Exempting Active Directory Users from Password Synchronization
The passwords in password change operations are still subject to the password policy settings, such as password expiration times. For example, in FreeIPA every password change requires an immediate password reset. While normal user passwords need to be subject to password policies, administrative passwords should be exempt from any password rules. A list of user DNs can be set in the password synchronization configuration that are exempted from the password policy.
The Directory Manager password is always exempt from password policy.
Edit the password synchronization entry, cn=ipa_pwd_extop,cn=plugins,cn=config, and add the passSyncManagersDNs attribute with the name of the user. This attribute is multi-valued. For example:
$ ldapmodify -x -D "cn=Directory Manager" -w secret -h ldap.example.com -p 389
dn: cn=ipa_pwd_extop,cn=plugins,cn=config
changetype: modify
add: passSyncManagersDNs
passSyncManagersDNs: uid=admin,cn=users,cn=accounts,dc=example,dc=com
Chapter 10. Identity: Managing DNS
If the FreeIPA server was installed with DNS configured, then all of the DNS entries for the domain — host entries, locations, records — can be managed using the FreeIPA tools.
10.1. About DNS in FreeIPA
DNS is one of the services that can be configured and maintained by the FreeIPA domain. DNS is critical to the performance of the FreeIPA domain; DNS is used for the Kerberos services and SSL connections for all servers and clients and for connections to domain services like LDAP.
While FreeIPA can use an external DNS service, there is a lot more flexibility and control over FreeIPA — DNS interactions when the DNS service is configured within the domain. For example, DNS records and zones can be managed within the domain using FreeIPA tools, and clients can update their own DNS records dynamically. When a host is added to FreeIPA, a DNS record is automatically created in FreeIPA's DNS service for that host machine.
FreeIPA stores all DNS information as LDAP entries. Every resource record for each machine is stored for the domain. For example, the client1 resource has three IPv4 (A) records and one IPv6 (AAAA) record:
dn:
idnsname=client1,idnsname=example.com,cn=dns,dc=example,dc=com
idnsname: client1 arecord: 10.0.0.1 arecord: 10.0.0.2 arecord: 10.0.0.3
aaaarecord: fc00::1 objectclass: top objectclass: idnsrecord
The schema used to define the DNS entries is in the
/usr/share/ipa/60basev2.ldifschema file
.
The BIND service communicates with the Directory Server using the system
bind-dyndb-ldapplug-in. When FreeIPA is configured to manage DNS (
Section 10.3, “Setting up DNS After FreeIPA Server Installation”), FreeIPA creates a
dynamic-dbconfiguration section in the
/etc/named.conffile for the BIND service. This configures the
bind-dyndb-ldapplug-in for the BIND (
named) service.
When this plug-in is properly configured, it delivers the DNS records from the Directory Server to the namedservice. The configuration can be changed to adapt the behavior of the plug-in and, therefore, the LDAP-BIND interactions.
When DNS is configured either by the
--setup-dns option during the intial FreeIPA install or the
ipa-dns-install command after the intial FreeIPA setup, the FreeIPA server installer creates
/etc/resolv.conf with the FreeIPA server's IP address. For example:
nameserver 192.0.2.1
It is recommended to manually add other replicas to /etc/resolv.conf as new replicas are deployed.
The maximum number of nameserver lines currently allowed in /etc/resolv.conf is three (3). If more information is needed, man resolv.conf for more details.
In FreeIPA versions 3 and lower, wildcards cannot be used when configuring DNS names. Only explicit DNS domain names are supported.
10.2. The FreeIPA-Generated DNS File
To help create and configure a suitable DNS setup, the FreeIPA installation script creates a sample zone file. During the installation, FreeIPA displays a message similar to the following:
Sample zone file for bind has been created in
/tmp/sample.zone.F_uMf4.db
If a DNS server is already configured in the network, then the configuration in the FreeIPA-generated file can be added to the existing DNS zone file. This allows FreeIPA clients to find LDAP and Kerberos servers that are required for them to participate in the FreeIPA domain. For example, this DNS zone configuration is created for an FreeIPA server with the KDC and DNS servers all on the same machine in the EXAMPLE.COM realm:
; ldap servers _ldap._tcp IN SRV 0 100 389
ipaserver.example.com. ;kerberos realm _kerberos IN TXT EXAMPLE.COM ;
kerberos servers _kerberos._tcp IN SRV 0 100 88 ipaserver.example.com.
_kerberos._udp IN SRV 0 100 88 ipaserver.example.com.
_kerberos-master._tcp IN SRV 0 100 88 ipaserver.example.com.
_kerberos-master._udp IN SRV 0 100 88 ipaserver.example.com.
_kpasswd._tcp IN SRV 0 100 464 ipaserver.example.com. _kpasswd._udp IN
SRV 0 100 464 ipaserver.example.com.10.3. Setting up DNS After FreeIPA Server Installation
DNS can be configured as part of the FreeIPA server installation, simply by using the --setup-dnsoption. If DNS is not configured then, it can be configured later using the ipa-dns-installcommand. For example:
ipa-dns-install -p secret --ip-address=1.2.34.56
--no-forwarders [--zone-refresh=60 | --zone-notif]
-pgives the password for the Directory Manager user in the 389 Directory Server. All of the DNS entries are stored in the LDAP directory, so this directory must be accessed to add the DNS configuration.
--ip-addressgives the IP address for the master DNS server.
--no-forwardersmeans that there are no forwarders used with the DNS service, only root servers. Alternatively, use the --forwarderoption to define a forward to use; to specify multiple forwarders, use the --forwarderoption multiple times.
Reverse DNS is configured automatically. It is possible to disable reverse DNS by using the --no-reverseoption.
If an existing reverse DNS zone is already configured, using the --no-reverseoption uses the existing reverse zone rather than creating a new reverse zone.
The FreeIPA server can actively check to see when new DNS zones are added and to update its DNS server accordingly. If no value is explicitly given, the zone refresh period is 30 seconds. The refresh interval can be set to another value using the --zone-refreshoption, which sets the polling interval in seconds.
Similar to refreshing the zones, the FreeIPA server can leave a persistent search open with its Directory Server and capture any new zone changes immediately. This is enabled with the --zone-notifoption.
If the --zone-notifoption is used to configure DNS, then the automatic zone refresh is disabled.
10.4. Managing DNS Zone Entries
10.4.1.1. Adding DNS Zones from the Web UI
Open the Identitytab, and select the DNSsubtab.
Click the Addlink at the top of the list of DNS zones.
Fill in the information about the new DNS zone. The Zone Nameis required; this is the actual domain name. The other information about the administrator email and the authoritative name server are optional.
10.4.1.2. Adding DNS Zones from the Command Line
The ipa dnszone-addcommand adds a new zone to the DNS domain. At a minimum, this requires the name of the new subdomain:
$ ipa dnszone-add
domainName
If the name is not given, the script prompts for it. Other command-line options can also be passed with the ipa dnszone-addcommand.
To add a zone entry:
Add the new zone. For example:
$ ipa dnszone-add newserver.example.com
--admin-email=admin@example.com --minimum=3000
--dynamic-update=TRUE
Reload the name service.
# rndc reload
10.4.2. Modifying DNS Zones
A zone is created with a certain amount of configuration, set to default values.
Example 10.1. Default DNS Zone Entry Settings
dn: idnsname=example.com,cn=dns,dc=example,dc=com
idnsname: example.com idnssoamname: server.example.com.
idnssoarname: root.server.example.com. idnssoaserial:
2011130701 idnssoarefresh: 3600 idnssoaretry: 900
idnssoaexpire: 1209600 idnssoaminimum: 3600 idnsupdatepolicy:
grant EXAMPLE.COM krb5-self * A; grant EXAMPLE.COM krb5-self *
AAAA; idnszoneactive: TRUE idnsallowdynupdate: TRUE nsrecord:
server.example.com. objectclass: top objectclass: idnsrecord
objectclass: idnszone
All of the possible zone settings are listed in
Table 10.1, “Zone Attributes”. Along with setting the actual information for the zone, the settings define how the DNS server handles the
start of authority(SOA) record entries and how it updates its records from the DNS name server.
Table 10.1. Zone Attributes
|
Attribute
|
Command-Line Option
|
Description
|
|---|
|
Zone name
|
--name
|
Sets the name of the zone.
|
|
Authoritative nameserver
|
--name-server
|
Sets the fully-qualified domain name of the DNS name server.
|
|
Administrator e-mail address
|
--admin-email
|
Sets the email address to use for the zone administrator. This defaults to the root account on the host.
|
|
SOA serial
|
--serial
|
Sets a version number for the SOA record file.
|
|
SOA refresh
|
--refresh
|
Sets the interval, in seconds, for a secondary DNS server to wait before requesting updates from the primary DNS server.
|
|
SOA retry
|
--retry
|
Sets the time, in seconds, to wait before retrying a failed refresh operation.
|
|
SOA expire
|
--expire
|
Sets the time, in seconds, that a secondary DNS server will try to perform a refresh update before ending the operation attempt.
|
|
SOA minimum
|
--minimum
|
Sets the minimum amount of time, in seconds, that data are kept in cache.
|
|
SOA time to live
|
--ttl
|
Sets the maximum time, in seconds, that information is kept in the data cache.
|
|
SOA class
|
--class
|
Sets the type of record. This is almost always IN, which stands for Internet.
|
|
BIND update policy
|
--update-policy
|
Sets the permissions allowed to clients in the DNS zone.
|
|
Dynamic update
|
--dynamic-update=TRUE|FALSE
|
Enables dynamic updates to DNS records for clients.
|
|
Name server
|
--ip-address
|
Adds the DNS name server by its IP address.
|
|
Allow transfer
|
--allow-transfer= string
|
Gives a semi-colon-separated listed of IP addresses or network names which are allowed to transfer the given zone.
|
|
Allow query
|
--allow-query
|
Gives a semi-colon-separated listed of IP addresses or network names which are allowed to issue DNS queries.
|
|
Allow PTR sync
|
--allow-sync-ptr=TRUE|FALSE
|
Sets whether A or AAAA records (forward records) for the zone will be automatically synchronized with the PTR (reverse) records.
|
|
Zone forwarders
|
--forwarder= string
|
Specifies a forwarder specifically configured for the DNS zone. This is separate from any global forwarders used in the FreeIPA domain.
To specificy multiple forwarders, use the option multiple times.
|
|
Forward policy
|
--forward-policy=only|first
|
Sets whether the zone will only forward requests to configured the DNS name servers (a forward-onlyzone) or whether it will check the forwarders first for DNS records and then check its own local records.
|
10.4.2.1. Editing the Zone Configuration in the Web UI
Open the Identitytab, and select the DNSsubtab.
Click the name of the DNS zone to edit.
Open the Settingstab.
Change any of the DNS zone settings. The full list of attributes is described in
Table 10.1, “Zone Attributes”. There are some common attributes to change:
Authoritative name server, the fully-qualified domain name of the DNS name server.
Dynamic update, to enable dynamic updates to DNS records for clients.
SOA refresh, the interval, in seconds, for a secondary DNS server to wait before requesting updates from the primary DNS server.
Click the Updatelink at the top of the settings page.
10.4.2.2. Editing the Zone Configuration in the Command Line
The zone can be created with additional attributes and values different from the default by passing additional options with the
dnszone-addcommand. Likewise, attributes can be added or modified in the zone entry by passing the same attribute options with the
dnszone-modcommand. These are listed in
Table 10.1, “Zone Attributes”.
If an attribute does not exist in the DNS zone entry, than the dnszone-modcommand adds the attribute. If the attribute exists, then it overwrites the current value with the specified value.
For example, to set a time to live for SOA records:
$ ipa dnszone-mod server.example.com
--ttl=1800
This adds a new attribute to the DNS zone entry:
dn: idnsname=example.com,cn=dns,dc=example,dc=com
idnsname: example.com idnssoamname: server.example.com.
idnssoarname: root.server.example.com. idnssoaserial:
2011130701 idnssoarefresh: 3600 idnssoaretry: 900
idnssoaexpire: 1209600 idnssoaminimum: 3600
dnsttl: 1800idnsupdatepolicy: grant
EXAMPLE.COM krb5-self * A; grant EXAMPLE.COM krb5-self * AAAA;
idnszoneactive: TRUE idnsallowdynupdate: TRUE nsrecord:
server.example.com. objectclass: top objectclass: idnsrecord
objectclass: idnszone10.4.3. Enabling and Disabling Zones
Active zones can have clients added to them, are available for lookups, and are used by FreeIPA services like Kerberos. Deleting a DNS zone removes the zone entry and all the associated configuration.
There can be situations when it is necessary to remove a zone from activity without permanently removing the zone. This is done by disablingthe zone.
10.4.3.1. Disabling Zones in the Web UI
Open the Identitytab, and select the DNSsubtab.
Click the name of the DNS zone to edit.
Open the Settingstab.
Scroll down to the Active zonefield. To disable the zone, set the value to Disabled.
Click the Updatelink at the top of the settings page.
10.4.3.2. Disabling Zones in the Command Line
Disabling a zone is done by using the dnszone-disablecommand.
For example:
$ ipa dnszone-disable server.example.com
When the zone needs to be brought back online, it can be re-enabled using the dnszone-enablecommand.
10.5. Managing DNS Record Entries
10.5.1. Adding Records to DNS Zones
Table 10.2. DNS Record Types
|
A
|
CERT
|
KX
|
NS
|
SIG
|
|
AAAA
|
CNAME
|
LOC
|
NSEC
|
SRV
|
|
A6
|
DNAME
|
MX
|
PTR
|
SSHFP
|
|
AFSDB
|
DS
|
NAPTR
|
RRSIG
|
TXT
|
10.5.1.1. Adding DNS Resource Records from the Web UI
Open the Identitytab, and select the DNSsubtab.
Click the name of the DNS zone to which to add records.
In the DNS Resource Recordstab, click the Addlink.
Select the type of record to create in the Record Typedrop-down menu. The required data is different, depending on the record type. For example, a CNAME record requires a hostname. The data field name updates automatically to indicate what kind of information to supply.
Although FreeIPA supports many different record types, there are four frequent record types that are used:
A. This is a basic map for a hostname and an ordinary IPv4 address. The Record Nameis a hostname, such as www. The IP Addressvalue is a standard IPv4 address, such as 192.168.1.2.
More information about A records is in
RFC 1035.
AAAA. This is a basic map for a hostname and an IPv6 address. The Record Nameis a hostname, such as www. The IP Addressvalue is a standard hexadecimal IPv6 address, such as fe80::20c:29ff:fe02:a1b3.
More information about AAAA records is in
RFC 3596.
SRV. Service (SRV) resource recordsmap service names to the DNS name of the server that is providing that particular service. The Record Namehas the format _service._protocol, such as _ldap._tcp. There are individual fields to set the priority, weight, port number, and hostname for the target service.
More information about SRV records is in
RFC 2782.
PTR. A pointer record type (PTR) record adds a reverseDNS record, which maps an IP address to a domain name. In this case, the Record Nameis the record ID number for the DNS entry of the resource and the Hostnamevalue is the hostname with a terminal period, such as server.example.com..
More information about PTR records is in
RFC 1035.
Click the Addbutton to save the new resource record.
10.5.1.2. Adding DNS Resource Records from the Command Line
The same script, ipa dnsrecord-add, is used to add resource records of any type, but the options for the script and the required data are different, based on the resource record type.
10.5.1.2.1. About the Commands to Add DNS Records
The ipa dnsrecord-addcommand adds records to DNS zones, based on the type. Adding a record has the same basic command format:
$ ipa dnsrecord-add
zoneName recordName--
recordType-option=data
The zoneNameis the name of the DNS zone to which the record is being added. The recordNameis an identifier for the new DNS resource record.
Table 10.3, “Common dnsrecord-add Options”lists options for the most common resource record types: A (IPv4), AAAA (IPv6), SRV, and PTR. Options for other supported record types are listed in the
ipa dnsrecord-addhelp and manpages.
The ipa dnsrecord-addcommand only creates forward entries, not reverse entries.
Table 10.3. Common dnsrecord-add Options
|
General Record Options
|
|---|
|
Option
|
Description
|
|
--ttl= number
|
Sets the time to live for the record.
|
|
--class=IN | CS | CH | HS
|
Sets the class of the record. This is usually IN, for Internet protocol.
|
|
--structured
|
Parses the raw DNS records and returns them in a structured format.
|
|
"A" Record Options
|
|---|
|
Option
|
Description
|
|
--a-rec= ARECORD
|
Passes a comma-separated list of A records.
|
|
--a-ip-address= string
|
Gives the IP address for the record.
|
|
"AAAA" Record Options
|
|---|
|
Option
|
Description
|
|
--aaaa-rec= AAAARECORD
|
Passes a comma-separated list of AAAA (IPv6) records.
|
|
--aaaa-ip-address= string
|
Gives the IPv6 address for the record.
|
|
"PTR" Record Options
|
|---|
|
Option
|
Description
|
|
--ptr-rec= PTRRECORD
|
Passes a comma-separated list of PTR records.
|
|
--ptr-hostname= string
|
Gives the hostname for the record.
|
|
"SRV" Record Options
|
|---|
|
Option
|
Description
|
|
--srv-rec= SRVRECORD
|
Passes a comma-separated list of SRV records.
|
|
--srv-priority= number
|
Sets the priority of the record. There can be multiple SRV records for a service type. The priority (0 - 65535) sets the rank of the record; the lower the number, the higher the priority. A service has to use the record with the highest priority first.
|
|
--srv-weight= number
|
Sets the weight of the record. This helps determine the order of SRV records with the same priority.
|
|
--srv-port= number
|
Gives the port for the service on the target host.
|
|
--srv-target= string
|
Gives the domain name of the target host. This can be a single period (.) if the service is not available in the domain.
|
10.5.1.2.2. Examples of Adding DNS Resource Records
Example 10.2. IPv4 Record
Type A resource records map hostnames to IPv4 addresses. The
recordvalue for these commands, then, is a standard IPv4 address. The URL label is usually www.
$ ipa dnsrecord-add example.com www --a-rec
10.64.14.165
This creates the record
www.example.comwith the IP address 10.64.14.165.
More information about A records is in
RFC 1035.
Example 10.3. IPv6 Record
Type AAAA resource records (
quad-A records)map hostnames to IPv6 addresses. The
recordvalue for these commands is an IPv6 address. As with Type A records, the URL label is usually www.
$ ipa dnsrecord-add example.com www --aaaa-rec
fe80::20c:29ff:fe02:a1b3
This creates the record
www.example.comwith the IP address fe80::20c:29ff:fe02:a1b3. More information about AAAA records is in
RFC 3596.
Example 10.4. SRV Record
Service (SRV) resource recordsmap service names to the DNS name of the server that is providing that particular service. For example, this record type can map a service like an LDAP directory to the DNS server which manages it.
As with Type A and Type AAAA records, SRV records specify a way to connect to and identify the service, but the record format is different.
The recordNameidentifies the service type and the connection protocol, in the format _service._protocol.
The recordinformation has the format "priority weight port target".
$ ipa dnsrecord-add server.example.com
_ldap._tcp --srv-rec="0 100 389 server1.example.com." $
ipa dnsrecord-add server.example.com _ldap._tcp
--srv-rec="1 100 389 server2.example.com."
More information about SRV records is in
RFC 2782.
Example 10.5. PTR Record
A pointer record type (PTR) record adds a reverseDNS record, which maps an IP address to a domain name, rather than the other way around.
All reverse DNS lookups for IPv4 addresses use reverse entries that are defined in the in-addr.arpa.domain. The reverse address, in human-readable form, is the exact reverse of the regular IP address, with the in-addr.arpa.domain appended to it. For example, for the IP address 192.0.1.2, the reverse address is 2.1.0.192.in-addr.arpa.
When adding the reverse DNS record, the format of the dnsrecord-addcommand is also reverse, compared to the usage for adding regular DNS entries:
$ ipa dnsrecord-add
reverseIpAddress recordId--ptr-rec
FQDN
The recordIdis the numeric identifier to use for the entry in the zone.
For example, this adds a record with an ID of 4 for server2.example.com:
$ ipa dnsrecord-add 2.1.0.192.in-addr.arpa 4
--ptr-rec server2.example.com.
More information about PTR records is in
RFC 1035.
Reverse zones can also be configured for IPv6 addresses, with zones in the
.ip6.arpa.domain. For more information about IPv6 reverse zones, see
RFC 3596.
10.5.2. Deleting Records from DNS Zones
10.5.2.1. Deleting Records with the Web UI
To delete only a specific record from the resource record:
Open the Identitytab, and select the DNSsubtab.
Click the name of the DNS zone.
In the DNS Resource Recordstab, click the name of the resource record.
Click the checkbox by the name of the record type to delete, and then click the active Deletelink at the top of the list.
This deletes only that record type while leaving the other configuration intact.
Alternatively, delete all of the records for the resource in the zone:
Open the Identitytab, and select the DNSsubtab.
Click the name of the DNS zone.
In the DNS Resource Recordstab, select the checkbox by the name of the resource record to delete. This deletes the entire record.
Click the Deletelink at the top of the zone records page.
10.5.2.2. Deleting Records with the Command Line
Records are removed from the zone using the ipa dnsrecord-delcommand. As with adding records, records are deleted using an option that specifies the type of record ( -- recordType -rec) and the record value.
For example, to remove the A type record:
$ ipa dnsrecord-del example.com www --a-rec
10.64.14.213
If you run the ipa dnsrecord-delcommand without any options, the command prompts for information about the record to delete.
Alternatively, using the --del-alloption removes all associated records for the zone.
10.6. Configuring the bind-dyndb-ldap Plug-in
The bind-dyndb-ldapsystem plug-in contains a DNS record cache for zones and a history of successful DNS resolutions. Maintaining the cache improves lookup performance in the Directory Server because it is not necessary to query the directory services every time there is a new DNS request.
When this plug-in is installed and FreeIPA is configured to manage DNS, then a new configuration section is added to the plug-in configuration.
Example 10.6. Default dynamic-db Configuration
dynamic-db "ipa" { library "ldap.so"; arg "uri
ldapi://%2fvar%2frun%2fslapd-EXAMPLE-COM.socket"; arg "base cn=dns,
dc=example,dc=com"; arg "fake_mname server.example.com."; arg
"auth_method sasl"; arg "sasl_mech GSSAPI"; arg "sasl_user
DNS/server.example.com"; };
This configuration uses implied default values for otherplug-in behaviors, like how long it maintains the cache. The assumed, default configuration can be changed by adding arguments to the dynamic-db "ipa"entry.
arg "
argument value";
Both cache updates and new zone detection can be forced by reloading the name server:
# rndc reload
Table 10.4. Additional bind-dyndb-ldap Configuration Parameters
|
Parameter
|
Description
|
Default Value
|
|---|
|
cache_ttl
|
Checks the DNS configuration in the Directory Server for new zones.
|
120 (seconds); this is defined in the bind-dyndb-ldapplug-in.
|
|
zone_refresh
|
Checks frequency, in seconds, that the server checks the DNS configuration in the Directory Server for new zones.
|
60 (seconds); this is configured by ipa-dns-install. If this is not set in the /etc/named.conffile, the bind-dyndb-ldapplug-in sets this value to zero (0), which disables zone refresh.
|
|
psearch
|
Enables persistent searches for the Directory Server so the BIND service immediately receives an update notification when a new DNS zone is added.
|
no
|
10.6.1. Changing the DNS Cache Setting
To improve DNS performance, it may be necessary to change the cache setting. By default, DNS records are kept in cache and considered valid for 120 seconds. This means that if a DNS record changes, it will not (necessarily) be propagated to the name server for up to 120 seconds. If the Directory Server has a high traffic volume or if records do not change frequently, then the cache time can be increased to improve performance by adding the cache_ttlparameter.
dynamic-db "ipa" { ... arg "cache_ttl 1800"; };10.6.2. Enabling Zone Refreshes and Persistent Searches
The DNS service receives its information through the bind-dyndb-ldapplug-in. The plug-in resolves only zones which were configured and enabled in the Directory Server when the name server started. When the name service restarts, the plug-in reloads its configuration and identifies any new zones or any new resource records.
However, the bind-dyndb-ldapplug-in pulls zone and resource record information from the FreeIPA LDAP directory, and it is possible to pull information from that directory apart from simply restarting the plug-in. The bind-dyndb-ldapplug-in search for zone changes actively either by refreshing the zone data or by keeping a persistent connection open to the Directory Server and immediately catching any changes.
Periodically checking for new zones is the same as refreshingthe zone configuration. This is set in the zone_refreshargument.
dynamic-db "ipa" { ... arg "zone_refresh 30"; };
Alternatively, the plug-in can maintain an open connection to the server through a persistent search. Persistent searches provide immediate notification of changes, unlike polling, and maintain a local cache of the configuration data.
A persistent search catches updates both to zones and to zone resource records.
Persistent searches are disabled by default but can be enabled in the psearchargument:
dynamic-db "ipa" { ... arg "psearch yes"; };
Because persistent searches leave an ongoing, open connection with the Directory Server, there can be some performance issues. Performance implications are covered in the
389 Directory Server Administrator's Guide.
10.7. Changing Recursive Queries Against Forwarders
The ipa-client-installscript sets a configuration statement in the /etc/named.conffile that allows name resolution against hosts that are outside the FreeIPA DNS domain. (This requires that the FreeIPA server be set up with DNS configured and with forwarders configured.) What this means is that any host is permitted to issue recursive queries against configured forwarders.
By default, any host is permitted to issue recursive queries against configured forwarders. The FreeIPA installation script automatically adds a line to the /etc/named.conffile to allow these recursive queries.
forward first; forwarders { 10.16.36.29; };
allow-recursion { any; };
This behavior can be changed in the allow-recursionstatement.
Open the /etc/named.conffile.
Reset the allow-recursionstatement. This is set to anyby default, which allows all hosts to resolve names against all forwarders.
forward first; forwarders { 10.16.36.29; };
allow-recursion { any; };
Restart the namedservice.
service named restart
The name server documentation has more details on editing configuration statements.
10.8. Enabling Dynamic DNS Updates
Dynamic DNS updates are not enabled by default for new DNS zones in FreeIPA. If dynamic updates are not allowed, then it may not be possible for the ipa-client-installscript to join a client to the domain because it cannot add a DNS record pointing to the new client.
10.8.1. Enabling Dynamic DNS Updates in the Web UI
Open the Identitytab, and select the DNSsubtab.
Click the name of the DNS zone to edit.
Open the Settingstab.
Scroll down to the Dynamic updatefield, and set the value to True.
Click the Updatelink at the top of the settings page.
10.8.2. Enabling Dynamic DNS Updates in the Command Line
To allow dynamic updates to the DNS zones, set the --dynamic-updateoption.
$ ipa dnszone-mod server.example.com
--dynamic-update=TRUE10.9. Configuring Forwarders and Forward Policy
A DNS forwarderis a server which passes DNS queries on to another, external DNS name server for resolution. Within the FreeIPA DNS domain, there are three configuration properties that define how forwarders are used:
A list of global forwarders which are used by all zones in FreeIPA
A list of forwarders which are used by a single, specific zone (as part of the zone configuration)
A policy which defines how the zone sends requests to the forwarders
10.9.1. Configuring Global Forwarders
Global forwarders are configured as part of the FreeIPA server configuration itself. Forwarders are (optionally) set up when the server is installed with the setup-dnsoption or when the ipa-dns-installscript is used.
After server configuration, the list of global forwarders can be edited using the dnsconfig-modcommand. For example:
[jsmith@server ~]$ ipa dnsconfig-mod --forwarder=0.9.8.7
Global forwarders: 0.9.8.710.9.2. Configuring Zone Forwarders
Forwarders can be configured to be used with a specific DNS zone as part of the zone configuration. The --forwarderoption sets a semi-colon-separated list of forwarders to use with the zone.
For example:
[jsmith@server ~]$ ipa dnszone-mod
--forwarder=1.2.3.4;255.255.255.255 example.com Zone name:
example.com ... Zone forwarders: 1.2.3.4;255.255.255.255
DNS forwarders must be specified as IP addresses, not as hostnames.
10.9.3. Configuring Forwarder Policy for a Zone
Once forwarders are configured, there are different ways that the zone can use them to service requests.
The zone can use the forwarders only for servicing name resolution requests; this is called a forward-only zone. A forward-only zone does not check its own name records. Only the forwarder server records are checked. If the record does not exist on the configured forwarders, then the zone returns a negative response to the client.
Alternatively, the zone can check the forwarder records first, and then fallback on its own resource records. This has a firstpolicy.
This configuration is set in the --forward-policyoption, using a policy of either onlyor first. For example:
[jsmith@server ~]$ ipa dnszone-mod --forward-policy=only
example.com Zone name: example.com ... Zone forwarders:
1.2.3.4;255.255.255.255 Forward policy: only10.10. Enabling Zone Transfers
Name servers maintain authoritative data for the zones; as changes are made to the zones, those changes must be sent to and distributed among the name servers for the DNS domain. A zone transfermoves resource records from one name server to another. An authoritative transfer(AXFR) is a zone transfer which includes that authoritative data for the zone (as opposed to an incremental transfer, which only delivers the most immediate zone change).
Zone transfers can be enabled when the zone is created or when it is modified by using the --allow-transferoption to set a list of name servers to which the zone records can be transferred.
For example:
[jsmith@server ~]$ ipa dnszone-mod
--allow-transfer=255.255.255.255;0.0.0.0;1.2.3.4 example-zone
The default is any, which the zone to be transferred anywhere in the DNS domain.
Once it is enabled in the bindservice, FreeIPA DNS zones can be transferred, by name, by clients like dig:
[root@server ~]# dig @ipa-server
zone_nameAXFR10.11. Defining DNS Queries
To resolve hostnames within the DNS domain, a DNS client issues a query to the DNS name server. For some security contexts or for performance, it may be advisable to restrict what clients can query DNS records in the zone.
DNS queries can be configured when the zone is created or when it is modified by using the --allow-queryoption to set a list of clients which are allowed to issue queries.
For example:
[jsmith@server ~]$ ipa dnszone-mod
--allow-query=255.255.255.255;0.0.0.0;1.2.3.4 example-zone
The default is any, which allows the zone to be queried by any client.
10.12. Synchronizing Forward and Reverse Zone Entries
Forward entries (A and AAAA) are configured separately from reverse entries (PTR). Because these entries are configured independently, it is possible for forward entries to exist without corresponding reverse entries, and vice versa.
A DNS zone can be configured to allow its forward and reverse entries to be synchronized automatically, by setting the --allow-sync-ptroption to true. This can be done when the zone is created or when it is edited.
For example, for editing an existing entry:
[jsmith@server ~]$ ipa dnszone-mod --allow-sync-ptr
example-zone
The default is false, which disables synchronization and has better server performance.
10.13. Setting DNS Access Policies
The FreeIPA DNS domain can define access controls, based on grant/deny rules, for zones. This creates an update-policystatement in the /etc/named.conffile, which defines the DNS access rule.
--update-policy "grant|deny
zoneName
policyName
recordName
recordType"
zoneNameis the FreeIPA DNS zone to which to apply the rule.
policyNameis the name to use for the BIND rule.
recordNamesets the resource records to which to apply the rule. Using an asterisk (*) is used for self rules.
recordTypeis the record type the rule applies to. Update access rules are applied individually for each record type, even within the same DNS zone entry.
For example, to grant the EXAMPLE.COMzone the ability to edit its own A and AAAA resource record entries:
$ ipa dnszone-mod example.com --update-policy="grant
EXAMPLE.COM krb5-self * A; grant EXAMPLE.COM krb5-self *
AAAA;"10.14. Resolving Hostnames in the FreeIPA Domain
It is possible to check the DNS entries for FreeIPA domain members using the dns-resolvecommand. If the record exists and is properly formatted in the DNS configuration, then the command returns the DNS record. If not, the command returns an error, that the hostname is not recognized within the DNS service.
$ipa dns-resolve server1.example.com
This can be helpful with troubleshooting connection problems between servers, clients, and services.
10.15. Changing Load Balancing for FreeIPA Servers and Replicas
As
Section 1.3.1, “About FreeIPA Servers and Replicas”touches on, FreeIPA servers and replicas in the domain automatically share the load among instances to maintain performance. The load balancing is defined first by the
priorityset for the server or replica in its SRV entry, and then by the
weightof that instance for servers/replicas with the same priority. Clients contact servers/replicas with the highest priority and then work their way down.
Load balancing is done automatically by servers, replicas, and clients. The configuration used for load balancing can be altered by changing the priority and the weight given to a server or replica.
(All replicas are initially created with the same priority.)
For example, this gives server1 a higher priority than server 2, meaning it will be contacted first:
$ ipa dnsrecord-add server.example.com _ldap._tcp --srv-rec="0
100 389 server1.example.com." $ ipa dnsrecord-add server.example.com
_ldap._tcp --srv-rec="1 100 389 server2.example.com."
More information about SRV records is in
RFC 2782.
Chapter 11. Policy: Using Automount
Automount is a way of making directories on different servers available, automatically, when requested by a user. This works exceptionally well within an FreeIPA domain since it allows directories on clients within the domain to be shared easily. This is especially important with user home directories (
Section 5.1, “Setting up User Home Directories”).
In FreeIPA, automount works with the internal LDAP directory and, if it is configured, DNS services.
11.1. About Automount and FreeIPA
Automount is a way to manage, organize, and access directories across multiple systems. Automount automatically mounts a directory whenever that resource is requested. Automount also provides a coherent structure to the way that these directories are organized. Every single directory, or mount point is called a key. Multiple keys that are grouped together are a map, and maps are associated according to their physical or conceptual location.
The base configuration file for autofs is the auto.master file in the /etc/ directory. There can be multiple auto.master configuration files in separate server locations, if necessary.
When autofs is configured on a server and that server is a client in a FreeIPA domain, then all of the configuration information for automount is stored in the FreeIPA directory. Rather than being stored in separate text files, the autofs configuration — maps, locations, and keys — are stored as LDAP entries. For example, the default map file, auto.master, is stored as:
dn: automountmapname=auto.master,cn=default,cn=automount,dc=example,dc=com
objectClass: automountMap
objectClass: top
automountMapName: auto.master
FreeIPA does not set up or configure autofs. That must be done separately. FreeIPA works with an existing autofs deployment.
Each new location is added as a container entry under cn=automount,dc=example,dc=com, and each map and each key are stored beneath that location.
As with other FreeIPA domain services, automount works with FreeIPA natively. The automount configuration can be managed by FreeIPA tools:
Locations, using ipa automountlocation* commands
Both direct and indirect maps, using ipa automountmap* commands
Keys, using ipa automountkey* commands
11.2. Configuring Automount
FreeIPA does not set up or configure autofs. That must be done separately, as described in these procedures. FreeIPA works with an existing autofs deployment.
Test that the /home directory can be mounted from the command line successfully before changing the automount configuration. Making sure that NFS is already working properly makes it easier to troubleshoot any potential FreeIPA automount configuration errors later.
11.2.1. Configuring autofs on Fedora
Edit the /etc/sysconfig/autofs file to specify the schema attributes that autofs searches for:
#
# Other common LDAP naming
#
MAP_OBJECT_CLASS="automountMap"
ENTRY_OBJECT_CLASS="automount"
MAP_ATTRIBUTE="automountMapName"
ENTRY_ATTRIBUTE="automountKey"
VALUE_ATTRIBUTE="automountInformation"
Specify the LDAP configuration. There are two ways to do this. The simplest is to let the automount service discover the LDAP server and locations on its own:
LDAP_URI="ldap:///dc=example,dc=com"
Alternatively, explicitly set which LDAP server to use and the base DN for LDAP searches:
LDAP_URI="ldap://ipa.example.com"
SEARCH_BASE="cn=location,cn=automount,dc=example,dc=com"
Edit the
/etc/autofs_ldap_auth.conf file so that autofs allows client authentication with the FreeIPA LDAP server.
Change authrequired to yes.
Set the principal to the Kerberos host principal for the NFS client server, host/fqdn@REALM. The principal name is used to connect to the FreeIPA directory as part of GSS client authentication.
<autofs_ldap_sasl_conf
usetls="no"
tlsrequired="no"
authrequired="yes"
authtype="GSSAPI"
clientprinc="host/server.example.com@EXAMPLE.COM"
/>
If necessary, run klist -k to get the exact host principal information.
Check the /etc/nssswitch.conf file, so that LDAP is listed as a source for automount configuration:
automount: files ldap
Restart autofs:
# service autofs restart
Test the configuration by listing a user's /home directory:
# ls /home/userName
If this does not mount the remote file system, check the /var/log/messages file for errors. If necessary, increase the debug level in the /etc/sysconfig/autofs file by setting the LOGGING parameter to debug.
If there are problems with automount, then cross-reference the automount attempts with the 389 Directory Server access logs for the FreeIPA instance, which will show the attempted access, user, and search base.
It is also simple to run automount in the foreground with debug logging on.
automount -f -d
This prints the debug log information directly, without having to cross-check the LDAP access log with automount's log.
11.2.2. Configuring Automount on Solaris
Solaris uses a different schema for autofs configuration than the schema used by FreeIPA. FreeIPA uses the 2307bis-style automount schema which is defined for 389 Directory Server (and used in FreeIPA's internal Directory Server instance).
If the NFS server is running on Fedora, specify on the Solaris machine that NFSv3 is the maximum supported version. Edit the /etc/default/nfs file and set the following parameter:
NFS_CLIENT_VERSMAX=3
Use the ldapclient command to configure the host to use LDAP:
ldapclient -v manual -a authenticationMethod=none
-a defaultSearchBase=dc=example,dc=com
-a defaultServerList=ipa.example.com
-a serviceSearchDescriptor=passwd:cn=users,cn=accounts,dc=example,dc=com
-a serviceSearchDescriptor=group:cn=groups,cn=compat,dc=example,dc=com
-a serviceSearchDescriptor=auto_master:automountMapName=auto.master,cn=location,cn=automount,dc=example,dc=com?one
-a serviceSearchDescriptor=auto_home:automountMapName=auto_home,cn=location,cn=automount,dc=example,dc=com?one
-a objectClassMap=shadow:shadowAccount=posixAccount
-a searchTimelimit=15
-a bindTimeLimit=5
Enable automount:
# svcadm enable svc:/system/filesystem/autofs
Test the configuration.
Check the LDAP configuration:
# ldapclient -l auto_master
dn: automountkey=/home,automountmapname=auto.master,cn=location,cn=automount,dc=example,dc=com
objectClass: automount
objectClass: top
automountKey: /home
automountInformation: auto.home
List a user's /home directory:
11.3. Setting up a Kerberized NFS Server
FreeIPA can be used to set up a Kerberized NFS server.
The NFS server does not need to be running on Fedora.
11.3.1. Setting up a Kerberized NFS Server
Obtain a Kerberos ticket before running FreeIPA tools.
[jsmith@server ~]$ kinit admin
Create the NFS service entry in the FreeIPA domain. For example:
[jsmith@server ~]$ ipa service-add nfs/nfs-server.example.com
Generate an NFS service keytab for the NFS server using the ipa-getkeytab command.
The NFS server may be on a Fedora machine in the FreeIPA domain or a different Unix machine. For a Fedora machine, the ipa-getkeytab command can be run on the NFS server machine. Otherwise, the ipa-getkeytab command should be run on a Fedora machine in the FreeIPA domain and then copied over to the NFS server.
If ipa-getkeytab command is run on the NFS server, then save the keys directly to the host keytab. For example:
[jsmith@server ~]$ ipa-getkeytab -s server.example.com -p nfs/nfs-server.example.com -k /etc/krb5.keytab
For a Fedora machine, that's all you need to do.
Only DES keys are supported on Red Hat Enterprise Linux 5.
When generating keys to copy over to another system, then generate the key but do not save it in the host keytab. The key must be added separately to the keytab after it is copied to the NFS server:
Save the keytab to a temporary file. For example:
[jsmith@server ~]$ ipa-getkeytab -s server.example.com -p nfs/nfs-server.example.com -k /tmp/nfs.keytab
Copy the keytabs over to the NFS server.
Set the file permissions to 0700.
Add the service key to the keytab file.
[root@nfs-server ~]# ( echo rkt /tmp/nfs.keytab; echo wkt /etc/krb5.keytab) |ktutil
Verify that the NFS service has been properly configured in FreeIPA, with its keytab, by checking the service entry:
[jsmith@server ~]$ ipa service-show nfs/ipaclient2.example.com
Principal: NFS/ipaclient2.example.com@EXAMPLE.COM
Keytab: True
Install the NFS packages. For example:
[root@nfs-server ~]# yum install nfs-utils
Configure weak crypto support. This is required for every NFS client if any client (such as a Red Hat Enterprise Linux 5 client) in the domain will use older encryption options like DES.
Edit the krb5.conf file to allow weak crypto.
[root@nfs-server ~]# vim /etc/krb5.conf
allow_weak_crypto = true
Update the FreeIPA server Kerberos configuration to support the DES encryption type.
[jsmith@ipaserver ~]$ ldapmodify -x -D "cn=directory manager" -w password -h ipaserver.example.com -p 389
dn: cn=EXAMPLEREALM,cn=kerberos,dc=example,dc=com
changetype: modify
add: krbSupportedEncSaltTypes
krbSupportedEncSaltTypes: des-cbc-crc:normal
-
add: krbSupportedEncSaltTypes
krbSupportedEncSaltTypes: des-cbc-crc:special
-
add: krbDefaultEncSaltTypes
krbDefaultEncSaltTypes: des-cbc-crc:special
If the NFS server and client are in different DNS domains, then configure the NFS domain.
[root@nfs-server ~]# vim /etc/idmapd.conf
Domain = example.com
Edit the /etc/exports file and add the Kerberos information:
/export *(rw,sec=sys:krb5:krb5i:krb5p)
Restart the NFS server and related services.
[root@nfs-server ~]# service nfs restart
[root@nfs-server ~]# service nfs-server restart
[root@nfs-server ~]# service nfs-secure restart
[root@nfs-server ~]# service nfs-secure-server restart
11.3.2. Setting up a Kerberized NFS Client
Obtain a Kerberos ticket before running FreeIPA tools.
[jsmith@server ~]$ kinit admin
Generate an NFS service keytab for the NFS client using the ipa-getkeytab command.
The NFS client may be on a Fedora machine in the FreeIPA domain or a different Unix machine. For a Fedora machine, the ipa-getkeytab command can be run on the NFS client machine. Otherwise, the ipa-getkeytab command should be run on a Fedora machine in the FreeIPA domain and then copied over to the NFS client.
If ipa-getkeytab command is run on the NFS client, then save the keys directly to the host keytab. For example:
For a Fedora machine, that's all you need to do.
When generating keys to copy over to another system, then generate the key but do not save it in the host keytab. The key must be added separately to the keytab after it is copied to the NFS server:
Save the keytab to a temporary file. For example:
[jsmith@server ~]$ ipa-getkeytab -p host/nfs-client-server.example.com@EXAMPLE.COM -k /tmp/nfs.keytab
Copy the keytabs over to the NFS client.
Set the file permissions to 0700.
Add the service key to the keytab file.
[root@nfs-client-server ~]# ( echo rkt /root/nfs-client.keytab; echo wkt /etc/krb5.keytab) |ktutil
If the NFS server and client are in different DNS domains, then configure the NFS domain. The idmapd.conf must be the same on the NFS client as it is on the NFS server.
[root@nfs-client-server ~]# vim /etc/idmapd.conf
Domain = example.com
Start the GSS daemon.
[root@nfs-client-server ~]# service rpcgssd start
[root@nfs-client-server ~]# service rpcbind start
[root@nfs-client-server ~]# service rpcidmapd start
Mount the directory.
[root@nfs-client-server ~]# echo "$NFSSERVER:/this /mnt/this nfs4 sec=krb5i,rw,proto=tcp,port=2049" >>/etc/fstab
[root@nfs-client-server ~]# mount -av
11.4. Configuring Locations
A location is a set of maps, which are all stored in auto.master, and a location can store multiple maps. The location entry only works as a container for map entries; it is not an automount configuration in and of itself.
FreeIPA does not set up or configure autofs. That must be done separately. FreeIPA works with an existing autofs deployment.
11.4.1. Configuring Locations through the Web UI
Click the Policy tab.
Click the Automount subtab.
Click the Add link at the top of the list of automount locations.
Enter the name for the new location.
11.4.2. Configuring Locations through the Command Line
To create a map, using the
automountlocation-add and give the location name.
$ ipa automountlocation-add location
For example:
$ ipa automountlocation-add raleigh
----------------------------------
Added automount location "raleigh"
----------------------------------
Location: raleigh
When a new location is created, two maps are automatically created for it, auto.master and auto.direct. auto.master is the root map for all automount maps for the location. auto.direct is the default map for direct mounts and is mounted on /-.
To view all of the maps configured for a location as if they were deployed on a filesystem, use the automountlocation-tofiles command:
$ ipa automountlocation-tofiles raleigh
/etc/auto.master:
/- /etc/auto.direct
---------------------------
/etc/auto.direct:
Configuring maps not only creates the maps, it associates mount points through the keys and it assigns mount options that should be used when the directory is accessed. FreeIPA supports both direct and indirect maps.
Different clients can use different map sets. Map sets use a tree structure, so maps cannot be shared between locations.
FreeIPA does not set up or configure autofs. That must be done separately. FreeIPA works with an existing autofs deployment.
11.5.1. Configuring Direct Maps
Direct maps define exact locations, meaning absolute paths, to the file mount. In the location entry, a direct map is identified by the preceding forward slash:
---------------------------
/etc/auto.direct:
/shared/man server.example.com:/shared/man
11.5.1.1. Configuring Direct Maps from the Web UI
Click the Policy tab.
Click the Automount subtab.
Click name of the automount location to which to add the map.
In the Automount Maps tab, click the + Add link to create a new map.
In pop-up window, select the Direct radio button and enter the name of the new map.
In the Automount Keys tab, click the + Add link to create a new key for the map.
Enter the mount point. The key defines the actual mount point in the key name. The Info field sets the network location of the directory, as well as any mount options to use.
Click the Add button to save the new key.
11.5.1.2. Configuring Direct Maps from the Command Line
The key defines the actual mount point (in the key name) and any options. A map is a direct or indirect map based on the format of its key.
Each location is created with an auto.direct item. The simplest configuration is to define a direct mapping by adding an automount key the existing direct map entry. It is also possible to create different direct map entries.
Add the key for the direct map to the location's auto.direct file. The --key option identifies the mount point, and --info gives the network location of the directory, as well as any mount options to use. For example:
$ ipa automountkey-add raleigh auto.direct --key=/share --info="-ro,soft, ipaserver.example.com:/home/share"
Key: /share
Mount information: -ro,soft, ipaserver.example.com:/home/share
On Solaris, add the direct map and key using the ldapclient command to add the LDAP entry directly:
ldapclient -a serviceSearchDescriptor=auto_direct:automountMapName=auto.direct,cn=location,cn=automount,dc=example,dc=com?one
11.5.2. Configuring Indirect Maps
An indirect map essentially specifies a relative path for maps. A parent entry sets the base directory for all of the indirect maps. The indirect map key sets a sub directory; whenever the indirect map location is loaded, the key is appended to that base directory. For example, if the base directory is /docs and the key is man, then the map is /docs/man.
11.5.2.1. Configuring Indirect Maps from the Web UI
Click the Policy tab.
Click the Automount subtab.
Click name of the automount location to which to add the map.
In the Automount Maps tab, click the + Add link to create a new map.
In pop-up window, select the Indirect radio button and enter the required information for the indirect map:
The name of the new map
The mount point. The Mount field sets the base directory to use for all the indirect map keys.
Optionally, a parent map. The default parent is auto.master, but if another map exists which should be used, that can be specified in the Parent Map field.
Click the Add button to save the new key.
11.5.2.2. Configuring Indirect Maps from the Command Line
The primary difference between a direct map and an indirect map is that there is no forward slash in front of an indirect key.
---------------------------
/etc/auto.share:
man ipa.example.com:/docs/man
---------------------------
Create an indirect map to set the base entry using the
automountmap-add-indirect command. The
--mount option sets the base directory to use for all the indirect map keys. The default parent entry is
auto.master, but if another map exists which should be used, that can be specified using the
--parentmap option.
$ ipa automountmap-add-indirect location mapName --mount=directory [--parentmap=mapName]
For example:
$ ipa automountmap-add-indirect raleigh auto.share --mount=/share
--------------------------------
Added automount map "auto.share"
--------------------------------
Add the indirect key for the mount location:
$ ipa automountkey-add raleigh auto.share --key=docs --info="ipa.example.com:/export/docs"
-------------------------
Added automount key "docs"
-------------------------
Key: docs
Mount information: ipa.example.com:/export/docs
To verify the configuration, check the location file list using automountlocation-tofiles:
$ ipa automountlocation-tofiles raleigh
/etc/auto.master:
/- /etc/auto.direct
/share /etc/auto.share
---------------------------
/etc/auto.direct:
---------------------------
/etc/auto.share:
man ipa.example.com:/export/docs
On Solaris, add the indirect map using the ldapclient command to add the LDAP entry directly:
ldapclient -a serviceSearchDescriptor=auto_share:automountMapName=auto.share,cn=location,cn=automount,dc=example,dc=com?one
11.5.3. Importing Automount Maps
If there are existing automount maps, these can be imported into the FreeIPA automount configuration.
ipa automountlocation-import location map_file [--continuous]
The only required information is the FreeIPA automount location and the full path and name of the map file. The --continuous option tells the automountlocation-import command to continue through the map file, even if the command encounters errors.
For example:
$ ipa automountlocation-import raleigh /etc/custom.map
Chapter 12. Policy: Defining Password Policies
All users must have a password which they use to authenticate to the Kerberos domain. FreeIPA defines and enforces rules about password complexity, password histories, and account lockouts in order to maintain security.
FreeIPA, by default, does not expose passwords to clients, even hashed passwords, for system security.
12.1. About Password Policies and Policy Attributes
A password policy sets certain standards for passwords, such as the password complexity and the rules for changing passwords. A password policy minimizes the inherent risk of using passwords by ensuring that they meet adequate complexity standards to thwart brute force attacks and they are changed frequently enough to mitigate the risk of someone revealing or discovering a password.
There are three main configuration areas that are defined within the password policy:
The FreeIPA password policy is enforced jointly by the KDC and the LDAP server. While the password policy is set in the LDAP directory and is based on 389 Directory Server password policy attributes, the policy is ultimately constrained by the KDC password policy framework. The KDC policy is less flexible than the 389 Directory Server policy framework, so the FreeIPA password policy can only implement password policy elements supported in the KDC. Any other policy settings made within the 389 Directory Server are not visible or enforced in FreeIPA.
Password policies are assigned either globally or to groups in FreeIPA, not to individual users. The password policy is assigned a weight, so that if a user belongs to multiple groups with different password policies, the policy with the highest priority will take precedence.
Table 12.1. Password Policy Settings
|
Configuration Property
|
Command-Line Option
|
Description
|
|---|
|
Options for both the UI and CLI
|
|
Minimum Password Lifetime
|
--minlife
|
Sets the minimum period of time, in hours, that a user's password must be in effect before the user can change it. This can prevent a user from changing a password and then immediately changing it to the original value. The default value is one hour.
|
|
Maximum Password Lifetime
|
--maxlife
|
Sets the maximum period of time, in days, that a user's password can be in effect before it must be changed. The default value is 90 days.
|
|
Minimum Number of Character Classes
|
--minclasses
|
Sets the minimum number of different classes, or types, of character that must exist in a password before it is considered valid. For example, setting this value to 3 requires that any password must have characters from at least three categories in order to be approved. The default value is zero (0), meaning there are no required classes.
There are six character classes:
Upper-case characters
Lower-case characters
Digits
Special characters (for example, punctuation)
8-bit characters (characters whose decimal code starts at 128 or below)
Number of repeated characters
This weights in the opposite direction, so that if you have too many repeated characters you will not meet the quorum to satisfy the "level" expressed by krbPwdMinDiffChars.
|
|
Minimum Length of Password
|
--minlength
|
Sets the minimum number of characters for a password. The default value is eight characters.
|
|
Password History
|
--history
|
Sets the number of previous passwords that are stored and which a user is prevented from using. For example, if this is set to ten, FreeIPA prevents a user from reusing any of their previous ten passwords. The default value is zero (0), which disables password history.
Even with the password history set to zero, users cannot reuse a current password.
|
|
Options for the CLI only
|
|
Priority
|
--priority
|
Sets the priority which determines which policy is in effect. The lower the number, the higher priority.
Although this priority is required when the policy is first created in the UI, it cannot be reset in the UI. It can only be reset using the CLI.
|
|
Maximum Consecutive Failures
|
--maxfail
|
Specifies the maximum number of consecutive failures to input the correct password before the user's account is locked.
|
|
Fail Interval
|
--failinterval
|
Specifies the period (in seconds) after which the failure count will be reset.
|
|
Lockout Time
|
--lockouttime
|
Specifies the period (in seconds) for which a lockout is enforced. If the value is set to zero (0), the account is permanently locked after the maximum number of failures is reached.
|
12.2. Viewing Password Policies
There can be multiple password policies configured in FreeIPA. There is always a global policy, which is created with the server. Additional policies can be created for groups in FreeIPA.
The UI lists all of the group password policies and the global policy on the Password Policies page.
Using the CLI, both global and group-level password policies can be viewed using the pwpolicy-show command. The CLI can also display the password policy in effect for a user.
12.2.1. Viewing the Global Password Policy
The global password policy is created with the server. This applies to every user until a group-level password policy supersedes it.
Table 12.2. Default Global Password Policy
|
Attribute
|
Value
|
|---|
|
Max lifetime
|
90 (days)
|
|
Min lifetime
|
1 (hour)
|
|
History size
|
0 (unset)
|
|
Character classes
|
0 (unset)
|
|
Min length
|
8
|
|
Max failures
|
6
|
|
Failure reset interval
|
60
|
|
Lockout duration
|
600
|
12.2.1.1. With the Web UI
Click the Policy tab, and then click the Password Policies subtab.
All of the policies in the UI are listed by group. The global password policy is defined by the global_policy group. Click the group link.
The global policy is displayed.
12.2.1.2. With the Command Line
To view the global policy, simply run the pwpolicy-show command with no arguments:
$ ipa pwpolicy-show
Group: global_policy
Max lifetime (days): 90
Min lifetime (hours): 1
History size: 0
Character classes: 0
Min length: 8
Max failures: 6
Failure reset interval: 60
Lockout duration: 600
12.2.2. Viewing Group-Level Password Policies
12.2.2.1. With the Web UI
Click the Policy tab, and then click the Password Policies subtab.
All of the policies in the UI are listed by group. Click the name of the group which is assigned the policy.
The group policy is displayed.
12.2.2.2. With the Command Line
For a group-level password policy, specify the group name with the command:
$ ipa pwpolicy-show examplegroup
Group: global_policy
Max lifetime (days): 90
Min lifetime (hours): 1
History size: 3
Character classes: 4
Min length: 8
Max failures: 3
Failure reset interval: 15
Lockout duration: 150
12.2.3. Viewing the Password Policy in Effect for a User
A user may belong to multiple groups, each with their own separate password policies. These policies are not additive. Only one policy is in effect at a time and it applies to all password policy attributes. To see which policy is in effect for a specific user, the pwpolicy-show command can be run for a specific user. The results also show which group policy is in effect for that user.
$ ipa pwpolicy-show --user=jsmith
Group: admins
Max lifetime (days): 90
Min lifetime (hours): 1
History size: 0
Character classes: 0
Min length: 8
Max failures: 6
Failure reset interval: 60
Lockout duration: 600
12.3. Creating and Editing Password Policies
A password policy can be selective; it may only define certain elements. A global password policy sets defaults that are used for every user entry, unless a group policy takes priority.
A global policy always exists, so there is no reason to add a global password policy.
Group-level policies override the global policies and offer specific policies that only apply to group members. Password policies are not cumulative. Either a group policy or the global policy is in effect for a user or group, but not both simultaneously.
Group-level policies do not exist by default, so they must be created manually.
It is not possible to set a password policy for a non-existent group.
12.3.1. Creating Password Policies in the Web UI
Click the Policy tab, and then click the Password Policies subtab.
All of the policies in the UI are listed by group. The global password policy is defined by the global_policy group. Click the group link.
Click the Add link at the top.
In the pop-up box, select the group for which to create the password policy.
Set the priority of the policy. The higher the number, the lower the priority.
Only one password policy is in effect for a user, and that is the highest priority policy.
The priority cannot be changed in the UI once the policy is created.
Click the Add and Edit button so that the policy form immediately opens.
Set the policy fields. Leaving a field blank means that attribute is not added the password policy configuration.
Max lifetime sets the maximum amount of time, in days, that a password is valid before a user must reset it.
Min lifetime sets the minimum amount of time, in hours, that a password must remain in effect before a user is permitted to change it. This prevents a user from attempting to change a password back immediately to an older password or from cycling through the password history.
History size sets how many previous passwords are stored. A user cannot re-use a password that is still in the password history.
Character classes sets the different categories of character that must be used in the password. For example, a character class can be a number, special character, or capital; the complete list of categories is in
Table 12.1, “Password Policy Settings”. This is part of setting the complexity requirements.
Min length sets how many characters must be in a password. This is part of setting the complexity requirements.
12.3.2. Creating Password Policies with the Command Line
Password policies are added with the pwpolicy-add command.
$ ipa pwpolicy-add groupName --attribute-value
For example:
$ ipa pwpolicy-add exampleGroup --minlife=7 --maxlife=49 --history= --priority=1
Group: exampleGroup
Max lifetime (days): 49
Min lifetime (hours): 7
Priority: 1
Setting an attribute to a blank value effectively removes that attribute from the password policy.
12.3.3. Editing Password Policies with the Command Line
As with most FreeIPA entries, a password policy is edited by using a *-mod command, pwpolicy-mod, and then the policy name. However, there is one difference with editing password policies: there is a global policy which always exists. Editing a group-level password policy is slightly different than editing the global password policy.
Editing a group-level password policy follows the standard syntax of *-mod commands. It uses the pwpolicy-mod command, the name of the policy entry, and the attributes to change. For example:
[jsmith@ipaserver ~]$ ipa pwpolicy-mod exampleGroup --lockouttime=300 --history=5 --minlength=8
To edit the global password policy, use the pwpolicy-mod command with the attributes to change, but without specifying a password policy name. For example:
[jsmith@ipaserver ~]$ ipa pwpolicy-mod --lockouttime=300 --history=5 --minlength=8
12.4. Managing Password Expiration Limits
Password policies are applied at the time a password is changed. So, when a password is set, it conforms to the password policy in effect at that time. If the password policy is changed later, that change is not applied, retroactively, to the password.
With password expiration periods, there are two attributes that are related:
Changing the password expiration time in the password policy does not affect the expiration date for a user, until the user password is changed. If the password expiration date needs to be changed immediately, it can be changed by editing the user entry.
To force the expiration date to change, reset the krbPasswordExpiration attribute value for the user. This can be done using the FreeIPA CLI with the --setattr option:
[bjensen@ipaserver ~]$ kinit
[bjensen@ipaserver ~]$ ipa user-mod jsmith --setattr=krbPasswordExpiration=20121231011529Z
If the new expiration date should be applied to multiple entries, it may be simpler to use ldapmodify and edit multiple entries simultaneously through an LDIF file in the -f option. For example, editing a single entry (with a modify statement similar to the LDIF file in -f):
[bjensen@ipaserver ~]$ ldapmodify -Y GSSAPI -h ipaserver.example.com -p 389 -vv
dn: uid=jsmith,cn=users,cn=accounts,dc=example,dc=com
changetype: modify
replace: krbpasswordexpiration
krbpasswordexpiration: 20140202203734Z
-
If an administrator resets a password, it expires the previous password and forces the user to update the password. When the user updates the password, it automatically uses the new password policies, including a new expiration date.
12.5. Changing the Priority of Group Password Policies
A user may belong to multiple groups, each with different password policies. Since only one policy can be in effect for a user, there has to be a method to assign precedence to policies. That is done through priority.
The highest priority is zero (0). The lower the number, the higher the priority.
This is set initially when the password policy is created. It can be modified after the policy is created by resetting the --priority option.
$ ipa pwpolicy-mod examplegroup --priority=10
When a user belongs to multiple groups, the group password policy with the lowest number has the priority.
The password policy priority cannot be changed in the UI after it is created.
12.6. Setting Account Lockout Policies
A brute force attack occurs when a malefactor attempts to guess a password by simply slamming the server with multiple login attempts. An account lockout policy prevents brute force attacks by blocking an account from logging into the system after a certain number of login failures — even if the correct password is subsequently entered.
These attributes are available in the password policy form when a group-level password policy is created or when any password policy (including the global password policy) is edited.
Click the Policy tab, and then click the Password Policies subtab.
Click the name of the policy to edit.
Set the account lockout attribute values.
There are three parts to the account lockout policy:
The number of failed login attempts before the account is locked (Max Failures).
The time after a failed login attempt before the counter resets (Failure reset interval). Since mistakes do happen honestly, the count of failed attempts is not kept forever; it naturally lapses after a certain amount of time. This is in seconds.
How long an account is locked after the max number of failures is reached (Lockout duration). This is in seconds.
There are three parts to the account lockout policy:
The number of failed login attempts before the account is locked (--maxfail).
How long an account is locked after the max number of failures is reached (--lockouttime). This is in seconds.
The time after a failed login attempt before the counter resets (--failinterval). Since mistakes do happen honestly, the count of failed attempts is not kept forever; it naturally lapses after a certain amount of time. This is in seconds.
These account lockout attributes can all be set when a password policy is created with pwpolicy-add or added later using pwpolicy-mod. For example:
[jsmith@ipaserver ~]$ kinit admin
[jsmith@ipaserver ~]$ ipa pwpolicy-mod examplegroup --maxfail=4 --lockouttime=600 --failinterval=30
12.7. Enabling a Password Change Dialog
There may be situations when a user exists in FreeIPA but does not have a valid Kerberos ticket, meaning he cannot authenticate to the FreeIPA domain. This is possible for new users or for users whose domain passwords have expired. Much like enabling password authentication in the web UI, it is possible to enable password-based authentication to the client. This opens up a password change dialog box to allow the user to reset the expired password.
The password change dialog is enabled by using OpenSSH's challenge-response authentication.
The challenge-response dialog is optional. In many environments, it is not necessary because SSSD can handle changing expired passwords by invoking the required PAM modules. However, using the challenge-response option in OpenSSH makes it possible to do password changes directly in PAM and to support full PAM conversations.
This is not enabled by default, but it can be enabled by editing the OpenSSH configuration.
Open the /etc/ssh/sshd_config file.
Set ChallengeResponseAuthentication to yes.
Chapter 13. Policy: Managing the Kerberos Domain
Kerberos authentication is the core of authentication within the FreeIPA domain. The FreeIPA server actually runs a Kerberos server within it, and this Kerberos server can be configured for custom policies for managing tickets and keytabs.
FreeIPA has its own command-line tools to use to manage Kerberos policies. Do not use kadmin or kadmin.local to manage FreeIPA Kerberos settings.
Kerberos provides an authentication layer between services and users. Kerberos centralizes authentication into a single location; a user authenticates to the Kerberos server, and then when that user attempts to access any resource on the network, that resource can check the key distribution center (KDC) for the stored user credentials. This allows users to access multiple resources without having to supply credentials separately to each and every one.
All of the users and services, combined, and all of the KDCs and Kerberos servers that are aware of each other constitute a realm. Each user, machine, and service within the realm is identified by a unique name called the principal. The user or service uses the principal and a verifying credential (usually a password) to authenticate to the KDC. The credential that is shared with the KDC is a key and it is stored in a file called a key table or keytab.
When the KDC verifies the user's identity, it issues a ticket. The ticket is a long-term pass to any service and machine on the realm. The KDC issues the user a special kind of ticket called a ticket-granting ticket (TGT). Whenever the user tries to access a resource within the Kerberos realm, the resource sends a request for a ticket specifically for it. The TGT is used to issue a resource-specific ticket that the resource then uses to authenticate the user and grant access.
When a FreeIPA client is first configured, the host principal is automatically retrieved by the setup script and stored in the /etc/krb5.keytab file. This host principal is stored within the host record so that local service commands cannot be used with this principal. This prepares the client to function in the FreeIPA realm.
13.1.1. About Principal Names
The principal identifies not only the user or service, but also the realm that that entity belongs to. A principal name has two parts, the identifier and the realm:
identifier@REALM
For a user, the identifier is only the Kerberos username. For a service, the identifier is a combination of the service name and the hostname of the machine it runs on:
service/FQDN@REALM
The service name is a case-sensitive string that is specific to the service type, like host, ldap, http, and dns. Not all services have obvious principal identifiers; the sshd daemon, for example, uses the host service principal.
The host principal is usually stored in /etc/krb5.keytab.
When Kerberos requests a ticket, it always resolves the domain name aliases (DNS CNAME records) to the corresponding DNS address (A or AAAA records). The hostname from the address record is then used when service or host principals are created.
For example:
www.example.com CNAME web-01.example.com
web-01.example.com A 192.0.2.145
A service attempts to connect to the host using its CNAME alias:
$ ssh www.example.com
The Kerberos server requests a ticket for the resolved hostname, web-01.example.com@EXAMPLE.COM, so the host principal must be host/web-01.example.com@EXAMPLE.COM.
13.1.2. About Protecting Keytabs
To protect keytab files, reset the permissions and ownership to restrict access to the files to only the keytab owner. For example, set the owner of the Apache keytab (/etc/httpd/conf/ipa.keytab) to apache and the mode to 0600.
13.2. Setting Kerberos Ticket Policies
The Kerberos ticket policy sets basic restrictions on managing tickets within the Kerberos realm, such as the maximum ticket lifetime and the maximum renewal age (the period during which the ticket is renewable).
The Kerberos ticket policy is set globally so that it applies to every ticket issued within the realm. FreeIPA also has the ability to set user-level ticket policies which override the global policies. This can be used, for example, to set extended expiration times for administrators or to set shorter expiration times for some employees.
13.2.1. Setting Global Ticket Policies
13.2.1.1. From the Web UI
Click the Policy tab, and then click the Kerberos Ticket Policy subtab.
Change the ticket lifetime policies.
Click the Update link at the top of the policy page.
Restart the KDC.
# service krb5kdc restart
Any change to the global Kerberos ticket policy requires a restart of the KDC for the changes to take effect.
13.2.1.2. From the Command Line
The ipa krbtpolicy-mod command modifies the policy, while the ipa krbtpolicy-reset command resets the policy to the default values.
For example:
# ipa krbtpolicy-mod --maxlife=3600 --maxrenew=18000
Max life: 3600
Max renew: 18000
Any change to the global Kerberos ticket policy requires a restart of the KDC for the changes to take effect. Restart the KDC:
# service krb5kdc restart
13.2.2. Setting User-Level Ticket Policies
User-level Kerberos ticket policies are set using the same commands as global policies, but the user is specified in the command.
For example:
# ipa krbtpolicy-mod jsmith --maxlife=3600
Max life: 3600
User-level policies take effect immediately on the next requested ticket (such as running kinit), without having to restart the KDC service.
13.3. Refreshing Kerberos Tickets
Kerberos keys are analogous to passwords. Like passwords, security policies may require that Kerberos tickets are manually refreshed after a specified interval.
The version of the key is shown in its key version number (KVNO). Refreshing (also called rotating) the principal's key increments the KVNO in the keytab entry. When a key is refreshed, a new entry is added to the keytab with a higher KVNO. The original key remains in the keytab but is no longer used to issue tickets.
Each keytab for the FreeIPA realm has an entry in the FreeIPA LDAP server, which includes its last change time. The principals which need to be refreshed can be regenerated using the ipa-getkeytab command.
The ipa-getkeytab command does not delete the old keytab in case it already exists in the file.
Find all keytabs issued before the requisite date. For example, this looks for any principals created between midnight on January 1, 2010, and 11:59 PM on December 31, 2010:
# ldapsearch -x -b "cn=computers,cn=accounts,dc=example,dc=com" "(&(krblastpwdchange>=20100101000000)(krblastpwdchange<=20101231235959))" dn krbprincipalname
# ldapsearch -x -b "cn=services,cn=accounts,dc=example,dc=com" "(&(krblastpwdchange>=20100101000000)(krblastpwdchange<=20101231235959))" dn krbprincipalname
Host (machine) principals are stored under the cn=computers,cn=accounts,dc=example,dc=com subtree.
Service principals are stored under the cn=services,cn=accounts,dc=example,dc=com subtree.
Filter by the last change date (krblastpwdchange).
Limit the search result information to only the entry name and principal by specifying the dn krbprincipalname attributes.
Dates are expressed in YYYYMMDD format, and times in HHMMSS format (GMT).
Retrieve a new keytab for the principal using the ipa-getkeytab command. This requires the location of the original keytab for the service or host (-k), the principal (-p), and the FreeIPA server hostname (-s).
For example, this refreshes the host principal with a keytab in the default location of
/etc/krb5.keytab:
# ipa-getkeytab -p host/client.example.com@EXAMPLE.COM -s ipa.example.com -k /etc/krb5.keytab
This refreshes the keytab for the Apache service, with a keytab in the default location of
/etc/httpd/conf/ipa.keytab:
# ipa-getkeytab -p HTTP/client.example.com@EXAMPLE.COM -s ipa.example.com -k /etc/httpd/conf/ipa.keytab
Regenerate the keytab using ipa-getkeytab for every service.
The
klist command displays the new key version number for the refreshed keytab. The original keytab still exists in the database, and it is listed with the previous KVNO.
# klist -kt /etc/krb5.keytab
Keytab: WRFILE:/etc/krb5.keytab
KVNO Timestamp Principal
---- ----------------- --------------------------------------------------------
1 06/09/10 11:23:01 host/client.example.com@EXAMPLE.COM(aes256-cts-hmac-sha1-96)
2 06/09/11 05:58:47 host/client.example.com@EXAMPLE.COM(aes256-cts-hmac-sha1-96)
1 03/09/11 13:57:16 krbtgt/EXAMPLE.COM@EXAMPLE.COM(aes256-cts-hmac-sha1-96)
1 03/09/11 13:57:16 HTTP/ipa.example.com@EXAMPLE.COM(aes256-cts-hmac-sha1-96)
1 03/09/11 13:57:16 ldap/ipa.example.com@EXAMPLE.COM(aes256-cts-hmac-sha1-96)
Tickets issued against the old keytab continue to work, while new tickets are issued using the key with the highest KVNO. This avoids any disruption to system operations.
Some services, such as NFSv4, only support a limited set of encryption types. Pass the appropriate arguments to the ipa-getkeytab command to configure the keytab properly.
13.4. Caching Kerberos Passwords
A machine may not always be on the same network as the FreeIPA domain; for example, a machine may need to be logged into a VPN before it can access the FreeIPA domain. If a user logs into a system when it is offline and then later attempts to connect to FreeIPA services, then the user is blocked because there is no FreeIPA Kerberos ticket for that user. FreeIPA works around that limitation by using SSSD to store the Kerberos passwords in the SSSD cache.
This is configured by default by the ipa-client-install script. A configuration parameter is added to the /etc/sssd/sssd.conf file which specifically instructs SSSD to store those Kerberos passwords for the FreeIPA domain:
[domain/example.com]
cache_credentials = True
ipa_domain = example.com
id_provider = ipa
auth_provider = ipa
access_provider = ipa
chpass_provider = ipa
ipa_server = _srv_, server.example.com
krb5_store_password_if_offline = true
This default behavior can be disabled during the client installation by using the --no-krb5-offline-passwords option.
This behavior can also be disabled by editing the /etc/sssd/sssd.conf file and removing the krb5_store_password_if_offline line or changing its value to false.
[domain/example.com]
...
krb5_store_password_if_offline = false
Refreshing Kerberos tickets adds a new key to the keytab, but it does not clear the keytab. If a host is being unenrolled and re-added to the FreeIPA domain or if there are Kerberos connection errors, then it may be necessary to remove the keytab and create a new keytab.
This is done using the ipa-rmkeytab command. To remove all principals on the host, specify the realm with the -r option:
# ipa-rmkeytab -r EXAMPLE.COM -k /etc/krb5.keytab
To remove the keytab for a specific service, use the -p option to specify the service principal:
# ipa-rmkeytab -p ldap/client.example.com -k /etc/krb5.keytab
13.6. Troubleshooting Kerberos Errors
Kerberos errors frequently become apparent when trying to connect to the realm using kinit or a similar client. For information related to Kerberos, first check the Kerberos manpages, help files, and other resources.
FreeIPA has its own command-line tools to use to manage Kerberos policies. Do not use kadmin or kadmin.local to manage FreeIPA Kerberos settings.
There are several places to look for Kerberos error log information:
For kinit problems or other Kerberos server problems, look at the KDC log in /var/log/krb5kdc.log.
For FreeIPA-specific errors, look in /var/log/httpd/error_log.
It is possible to disable reverse DNS lookups in the SSH configuration. Rather than using reverse DNS records, SSH passes the given username directly to GSS-API.
To disable reverse DNS lookups with SSH, add or edit the GSSAPITrustDNS directive and set the value to no.
# vim /etc/ssh/ssh_config
GSSAPITrustDNS no
Failure to export an updated keytab can cause problems that are difficult to isolate. For example, existing service connections may continue to function, but no new connections may be possible.
Chapter 14. Policy: Using sudo
FreeIPA provides a mechanism for predictably and consistently apply sudo policies across the FreeIPA domain. The sudo policies apply to domain users and domain hosts.
The sudo command allows a system administrator to delegate authority to specific users to run specific commands as root or another specified user. sudo provides an audit trail of the commands and their arguments, so access can be tracked.
14.1.1. General sudo Configuration in FreeIPA
sudo uses a local configuration file, /etc/sudoers, which defines the commands and users with sudo access. While this file can be shared among machines, there's no native way to distribute sudo configuration files among machines.
FreeIPA uses its centralized LDAP database to contain the sudo configuration, which makes it globally available to all domain hosts. FreeIPA also has a specialized LDAP schema for sudo entries that allows a lot more flexible and simpler configuration. This schema adds two key features:
The FreeIPA schema supports host groups in addition to netgroups for sudo, while sudo only supports netgroups.
For every host group, FreeIPA also creates a corresponding shadow netgroup. This allows FreeIPA administrators to create sudo rules that reference host groups, while the local sudo command uses the corresponding netgroup.
FreeIPA introduces the concept of a sudo command group. The group contains multiple commands, and then the command group can be referenced in the sudo configuration.
Because sudo does not support host groups and command groups, FreeIPA translates the FreeIPA sudo configuration into native sudo configuration when the sudo rules are created.
Both sudo and FreeIPA support user groups as part of the sudo configuration. User groups can be either Unix or non-POSIX groups. Creating non-POSIX groups can create some access issues because any users in the group inherit non-POSIX rights from the group. Having the choice between Unix and non-POSIX groups allows administrators the choice in group formatting and to avoid problems with inherited permissions or GID information.
14.1.2. sudo and Netgroups
As
Section 14.1.1, “General sudo Configuration in FreeIPA” mentions, the LDAP schema used for sudo entries in FreeIPA supports host group-style groups in addition to netgroups. Really, FreeIPA creates two groups, a visible host group and a shadow netgroup.
sudo itself only supports NIS-style netgroups for group formats.
One important thing to consider is that even though sudo uses NIS netgroups, it is not necessary to have a NIS server installed or a NIS client configured. When any group is created for sudo, the NIS object is created in the Directory Server instance, and then the information is retrieved by NSS_LDAP or by SSSD. The client (in this case, sudo) then extracts the required NIS information from the information provided by FreeIPA's Directory Server.
In short, sudo configuration requires NIS-formatted netgroups. It does not require NIS.
The FreeIPA Directory Server instance uses the standard LDAP schema for NIS objects, defined in RFC 2307.
14.1.3. Supported sudo Clients
Any system which is supported as an FreeIPA client system can be configured as a sudo client in FreeIPA.
14.2. Setting up sudo Commands and Command Groups
Just as in regular sudo configuration, any command which will be governed by sudo access must be listed in the configuration. FreeIPA adds an extra control measure with sudo command groups, which allow a group of commands to be defined and then applied to the sudo configuration as one.
Adding a command or a command group makes it available to FreeIPA to be defined in a sudo rule; simply adding a command does not automatically include it in a sudo rule.
14.2.1. Adding sudo Commands
14.2.1.1. Adding sudo Commands with the Web UI
Click the Policy tab.
Click the Sudo subtab, and then select the Sudo Commands link.
Click the Add link at the top of the list of commands.
Enter the full system path and name of the command and, optionally, a description.
Click the Add and Edit button to go immediately to the edit pages for the command.
In the Sudo Command Groups tab, click the Add button to add the sudo command to a command group.
Click the checkbox by the groups for the command to join, and click the right arrows button, >>, to move the group to the selection box.
Click the Add button.
14.2.1.2. Adding sudo Commands with the Command Line
To add a single command, use the sudocmd-add command. This requires the full, local path to the command executable and a description of the command:
$ ipa sudocmd-add --desc "description" /local/path/to/command
For example:
$ ipa sudocmd-add --desc 'For reading log files' '/usr/bin/less'
----------------------------------
Added sudo command "/usr/bin/less"
----------------------------------
sudo Command: /usr/bin/less
Description: For reading log files
14.2.2. Adding sudo Command Groups
14.2.2.1. Adding sudo Command Groups with the Web UI
Click the Policy tab.
Click the Sudo subtab, and then select the Sudo Command Groups link.
Click the Add link at the top of the list of command groups.
Enter the name and description for the new command group.
Click the Add and Edit button to go immediately to the edit pages for the group.
In the Sudo Commands tab, click the Add button to add a sudo command to the group.
In the Sudo Commands tab, click the Add button to add a sudo command to the group.
Click the checkbox by the names of the commands to add, and click the right arrows button, >>, to move the command to the selection box.
Click the Add button.
14.2.2.2. Adding sudo Command Groups with the Command Line
Creating a command group requires creating two entries, one for the group and one for the command itself:
Create the command group using the sudocmdgroup-add command:
$ ipa sudocmdgroup-add --desc 'File editing commands' files
-----------------------------------
Added sudo command group "files"
-----------------------------------
sudo Command Group: files
Description: File editing commands
Create a command entry using the sudocmd-add command:
$ ipa sudocmd-add --desc 'For editing files' '/usr/bin/vim'
----------------------------------
Added sudo command "/usr/bin/vim"
----------------------------------
sudo Command: /usr/bin/vim
Description: For editing files
Add the command, using its full directory location as its name, to the command group using the
sudocmdgroup-add-member command:
$ ipa sudocmdgroup-add-member --sudocmds '/usr/bin/vim' files
sudo Command Group: files
Description: File editing commands
Member sudo commands: /usr/bin/vim
-------------------------
Number of members added 1
-------------------------
14.3. Defining sudo Rules
sudo rules are in a sense similar to access control rules: they define users who are granted access, the commands which are within the scope of the rule, and then the target hosts to which the rule applies. In FreeIPA, additional information can be configured in the rule, such as sudoers options and run-as settings, but the basic elements always define who, what (services), and where (hosts).
14.3.1. About External Users and Hosts
sudo rules define four elements: who can do what, where, and as whom. The who is the regular user, and the as whom is the system or other user identity which the user uses to perform tasks. Those tasks are system commands that can be run (or specifically not run) on a target machine.
Three of those elements — who, as whom, and where — are identities. They are uses and hosts. Most of the time, those identities are going to be entities within the FreeIPA domain because there will be overlap between the system users and machines in the environment and the users and hosts belonging to the FreeIPA domain.
However, that is not necessarily the case with all identities that a sudo policy may realistically cover. For example, sudo rules could be used to grant root access to member of the IT group in FreeIPA, and that root user is not a user in FreeIPA. Or, for another example, administrators may want to block access to certain hosts that are on a network but are not part of the FreeIPA domain.
The sudo rules in FreeIPA support the concept of external users and hosts — meaning, hosts and users which are stored and exist outside of the FreeIPA configuration.
When configuring a sudo rule, the user, run-as, and host settings all can point to an external identity to be included and evaluated in the sudo rule.
14.3.2. About sudo Options Format
However, the sudo rule configuration in FreeIPA does not allow the same formatting as the configuration in the /etc/sudoers file. Specifically, FreeIPA does not allow whitespaces in the options parameter, whether it is set in the UI or the CLI.
For example, in the /etc/sudoers file, it is permissible to list options in a comma-separate list with spaces between:
mail_badpass, mail_no_host, mail_no_perms, syslog = local2
However, in FreeIPA, that same configuration would be interpreted as different arguments — including the equals sign (=) since it has spaces around it.
Likewise, linebreaks that are ignored in the /etc/sudoers file are not allowed in the FreeIPA configuration:
env_keep = "COLORS DISPLAY EDITOR HOSTNAME HISTSIZE INPUTRC
KDEDIR LESSSECURE LS_COLORS MAIL PATH PS1 PS2
QTDIR USERNAME LANG LC_ADDRESS LC_CTYPE LC_COLLATE
LC_IDENTIFICATION LC_MEASUREMENT LC_MESSAGES
LC_MONETARY LC_NAME LC_NUMERIC LC_PAPER LC_TELEPHONE
LC_TIME LC_ALL LANGUAGE LINGUAS _XKB_CHARSET
XAUTHORITY"
To use multiple sudoers options in FreeIPA, configure each one as a separate option setting, rather than all on one line, as is allowed in the /etc/sudoers file.
14.3.3. Defining sudo Rules in the Web UI
Click the Policy tab.
Click the Sudo subtab, and then select the Sudo Rules link.
Click the Add link at the top of the list of sudo rules.
Enter the name for the rule.
Click the Add and Edit button to go immediately to set the configuration for the rule.
There are a number of configuration areas for the rule. The most basic elements are set in the Who, Access This Host, and Run Commands areas; the others are optional and are used to refine the rule.
Click the + Add link at the right of the options list.
Enter the sudoers option.
Click Add.
In the Who area, select the users or user groups to which the sudo rule is applied.
Click the + Add link at the right of the users list.
Click the checkbox by the users to add to the rule, and click the right arrows button, >>, to move the users to the selection box.
Click Add.
In the Access This Host area, select the hosts on which the sudo rule is in effect.
Click the + Add link at the right of the hosts list.
Click the checkbox by the hosts to include with the rule, and click the right arrows button, >>, to move the hosts to the selection box.
Click Add.
In the Run Commands area, select the commands which are included in the sudo rule. The sudo rule can grant access or deny access to commands, and it can grant allow access to one command and also deny access to another.
In the Allow/Deny area, click the + Add link at the right of the commands list.
Click the checkbox by the commands or command groups to include with the rule, and click the right arrows button, >>, to move the commands to the selection box.
Click Add.
Optional. The sudo rule can be configured to run the given commands as a specific, non-root user.
In the As Whom area, click the + Add link at the right of the users list.
Click the checkbox by the users to run the command as, and click the right arrows button, >>, to move the users to the selection box.
Click Add.
14.3.4. Defining sudo Rules in the Command Line
The basic outline of a sudo rule command is:
$ ipa sudorule-add* options ruleName
Example 14.1. Creating Basic sudo Rules
In the most basic case, the sudo configuration is going to grant the right to one user for one command on one host.
The first step is to add the initial rule entry.
$ ipa sudorule-add files-commands
-----------------------------------
Added sudo rule "files-commands"
-----------------------------------
Rule name: files-commands
Enabled: TRUE
Next, add the commands to
grant access to. This can be a single command, using
--sudocmd, or a group of commands, using
--sudocmdgroups.
$ ipa sudorule-add-allow-command --sudocmd "/usr/bin/vim" files-commands
Rule name: files-commands
Enabled: TRUE
sudo Commands: /usr/bin/vim
-------------------------
Number of members added 1
-------------------------
Add a host or a host group to the rule.
$ ipa sudorule-add-host --host server.example.com files-commands
Rule name: files-commands
Enabled: TRUE
Hosts: server.example.com
sudo Commands: /usr/bin/vim
-------------------------
Number of members added 1
-------------------------
Last, add the user or group to the rule. This is the user who is allowed to use sudo as defined in the rule; if no "run-as" user is given, then this user will run the sudo commands as root.
$ ipa sudorule-add-user --user jsmith files-commands
Rule name: files-commands
Enabled: TRUE
Users: jsmith
Hosts: server.example.com
sudo Commands: /usr/bin/vim"
-------------------------
Number of members added 1
-------------------------
Example 14.2. Allowing and Denying Commands
The sudo rule can grant access or deny access to commands. For example, this rule would allow read access to files but prevent editing:
$ ipa sudorule-add-allow-command --sudocmd "/usr/bin/less" readfiles
$ ipa sudorule-add-allow-command --sudocmd "/usr/bin/tail" readfiles
$ ipa sudorule-add-deny-command --sudocmd "/usr/bin/vim" readfiles
Example 14.3. Using sudoers Options
The
sudoers file has a lot of potential flags that can be set to control the behavior of
sudo users, like requiring (or not requiring) passwords to offer a user to authenticate to
sudo or using fully-qualified domain names in the
sudoers file. The complete list of options is in the
sudoers manpage and at
http://www.gratisoft.us/sudo/sudoers.man.html#sudoers_options.
Any of these options can be set for the FreeIPA sudo rule using the sudorule-add-option command. When the command is run, it prompts for the option to add:
$ ipa sudorule-add-option readfiles
Sudo Option: !authenticate
-----------------------------------------------------
Added option "!authenticate" to Sudo rule "readfiles"
-----------------------------------------------------
Example 14.4. Running as Other Users
The sudo rule also has the option of specifying a non-root user or group to run the commands as. The initial rule has the user or group specified using the --sudorule-add-runasuser or --sudorule-add-runasgroup command, respectively.
$ ipa sudorule-add-runasuser --users=jsmith readfiles
$ ipa sudorule-add-runasgroup --groups=ITadmins readfiles
When creating a rule, the sudorule-add-runasuser or sudorule-add-runasgroup command can only set specific users or groups. However, when editing a rule, it is possible to run sudo as all users or all groups by using the --runasusercat or --runasgroupcat option. For example:
$ ipa sudorule-mod --runasgroupcat=all ruleName
The --sudorule-add-runasuser and --sudorule-add-runasgroup commands do not support an all option, only specific user or group names. Specifying all users or all groups can only be used with options with the sudorule-mod command.
Example 14.5. Referencing External Users or Hosts
The "who" in a sudo rule can be a FreeIPA user, but there are many logical and useful rules where one of the referents is a system user. Similarly, a rule may need to grant or deny access to a host machine on the network which is not an FreeIPA client.
There are three options to add an external identity to a sudo rule:
--externaluser
--runasexternaluser
--externalhost
For example:
$ ipa sudorule-add-host --externalhost=external-server.example.com readfiles
$ ipa sudorule-add-user --externaluser=ITAdmin readfiles
$ ipa sudorule-add-runasuser --runasexternaluser=root readfiles
Table 14.1. sudo Commands
|
Command
|
Description
|
|---|
|
sudorule-add
|
Adds a sudo rule entry.
|
|
sudorule-add-user
|
Adds a user or a user group to the sudo rule. This user (or every member of the group) is then entitled to sudo any of the commands in the rule.
|
|
sudorule-add-host
|
Adds a target host for the rule. These are the hosts where the users are granted sudo permissions.
|
|
sudorule-add-runasgroup
|
Sets a group to run the sudo commands as. This must be a specific user; to specify all users, modify the rule using sudo-rule.
|
|
sudorule-add-runasuser
|
Sets a user to run the sudo commands as. This must be a specific user; to specify all users, modify the rule using sudo-rule.
|
|
sudorule-add-allow-command
|
Adds a command that users in the rule have sudo permission to run.
|
|
sudorule-add-deny-command
|
Adds a command that users in the rule are explicitly denied sudo permission to run.
|
|
sudorule-add-option
|
Adds a sudoers flag to the sudo rule.
|
14.4. Applying the Configured sudo Policies to Hosts
Actually implementing sudo policies is more complicated than simply creating the rules in FreeIPA. Those rules need to be applied to every local machine, which means that each system in the FreeIPA domain has to be configured to refer to FreeIPA (as an LDAP server) for its policies.
This example specifically configures a Fedora 2.2 client for sudo rules. The configuration file in step
d is different, depending on the platform.
Set up a bind (authenticated) user by setting a password for the default FreeIPA sudo user. The user must be able to authenticate to the server; anonymous access is not supported for sudo policies.
Using LDAP tools, set the password for the sudo user, uid=sudo,cn=sysaccounts,cn=etc,dc=example,dc=com. For example:
$ ldappasswd -Y GSSAPI -S -h ipaserver.ipadocs.org uid=sudo,cn=sysaccounts,cn=etc,dc=example,dc=com
New password:
Re-enter new password:
Enter LDAP Password:
Configure every system in the FreeIPA domain.
Configure
sudo to look to LDAP for the
sudoers file.
vim /etc/nsswitch.conf
sudoers: files ldap
Leaving the files option in place allows sudo to check its local configuration before checking the LDAP-based FreeIPA configuration.
Enable debug logging for
sudo operations in the
/etc/ldap.conf file. If this file does not exist, it can be created.
vim /etc/ldap.conf
sudoers_debug: 1
Adding the
sudoers_debug parameter helps with troubleshooting. Valid values for this parameter are 0, 1, and 2. The
sudo documentation at
http://www.gratisoft.us/sudo/readme_ldap.html has more information on debugging the process.
Optionally, enable debugging in SSSD to show what LDAP settings it is using.
vim /etc/sssd/sssd.conf
[domain/IPADOMAIN]
debug_level = 6
....
The LDAP search base used by SSSD for operations is recorded in the sssd_DOMAINNAME.log log.
Edit the NSS/LDAP configuration file and add the following
sudo-related lines to the
/etc/sudo-ldap.conf file:
binddn uid=sudo,cn=sysaccounts,cn=etc,dc=example,dc=com
bindpw sudo_password
ssl start_tls
tls_cacertfile /etc/ipa/ca.crt
tls_checkpeer yes
bind_timelimit 5
timelimit 15
uri ldap://ipaserver.example.com ldap://backup.example.com:3890
sudoers_base ou=SUDOers,dc=example,dc=com
The uri directive must give the fully-qualified domain name of the LDAP server, not an IP address. Otherwise, sudo fails to connect to the LDAP server.
Set a name for the NIS domain in the
sudo configuration.
sudo uses NIS-style netgroups, so the NIS domain name must be set in the system configuration for
sudo to be able to find the host groups used in the FreeIPA
sudo configuration.
Open the /etc/rc.d/rc.local file. Setting the NIS domain name in this file allows the value to persist between reboots.
# vim /etc/rc.d/rc.local
Add the command to set the NIS domain name.
nisdomainname example.com
Even though sudo uses NIS-style netgroups, it is not necessary to have a NIS server installed. Netgroups require that a NIS domain be named in their configuration, so sudo requires that a NIS domain be named for netgroups. However, that NIS domain does not actually need to exist.
Chapter 15. Policy: Configuring Host-Based Access Control
FreeIPA can control access to both machines and the services on those machines within the FreeIPA domain. The rules define who can access what within the domain, not the level of access (which are defined by system or application settings). These access control rules grant access, with all other users and hosts implicitly denied.
This is called host-based access control because the rule defines what hosts (targets) within the domain a user is allowd to access. This access can be further broken down to users and services on those hosts.
Using host-based access control requires SSSD to be installed and configured on the FreeIPA client machine.
15.1. About Host-Based Access Control
Host-based access control rules (which are described in
Chapter 15, Policy: Configuring Host-Based Access Control) can be applied to individual hosts. However, using host groups allows centralized, and potentially simplified, access control management because an access control rule only needs to be defined once and then it is applied immediately and consistently to all the hosts within the group.
While access must be explicitly granted to users and hosts within the FreeIPA domain, FreeIPA servers are configured by default with an allow all access control rule which allows access for every host within the domain to every host within the domain.
To create an FreeIPA server without the default allow all rule, run ipa-server-install with the --no_hbac_allow option.
The rule first defines things that can be accessed, and there are two types of entities:
Hosts, or target hosts, within the FreeIPA domain.
Services on the target hosts. Multiple services can be combined into service groups. The service group can be modified without having to edit the access control rule itself.
The rule also sets who can have access (the FreeIPA domain user).
It is possible to use categories for users and target hosts instead of adding each one individually to the access control rule. The only supported category is all.
The entities in host-based access control rules follow the Kerberos principal entries: users, hosts (machines), and services. Users and target hosts can be added directly to host-based access control rules. However, services must be flagged first and then added to the access control rules.
15.2. Creating Host-Based Access Control Entries for Services and Service Groups
Any PAM service can be identified as to the host-based access control (HBAC) system in FreeIPA. The service entries used in host-based access control are separate from adding a service to the FreeIPA domain. Adding a service to the domain makes it a recognized resource which is available to other resources. Adding a domain resource to the host-based access control configuration allows administrators to exert defined control over what domain users and what domain clients can access that service.
Some common services are already configured as HBAC services, so they can be used in host-based access control rules. Additional services can be added, and services can be added into service groups for simpler management.
15.2.1. Adding HBAC Services
15.2.1.1. Adding HBAC Services in the Web UI
Click the Policy tab.
Click the Host-Based Access Control subtab, and then select the HBAC Services link.
Click the Add link at the top of the list of services.
Enter the service name and a description.
Click the Add button to save the new service.
15.2.1.2. Adding Services in the Command Line
The service is added to the access control system using the hbacsvc-add command, specifying the service by the name that PAM uses to evaluate the service.
For example, this adds the tftp service:
# ipa hbacsvc-add --desc="TFTP service" tftp
-------------------------
Added HBAC service "tftp"
-------------------------
15.2.2. Adding Service Groups
Once the individual service is added, it can be added to the access control rule. However, if there a large number of services, then it can require frequent updates to the access control rules as services change. FreeIPA also allows groups of services to added to access control rules. This makes it much easier to manage access control, because the members of the service group can be changed without having to edit the rule itself.
15.2.2.1. Adding Service Groups in the Web UI
Click the Policy tab.
Click the Host-Based Access Control subtab, and then select the HBAC Service Groups link.
Click the Add link at the top of the list of service groups.
Enter the service group name and a description.
Click the Add and Edit button to go immediately to the service configuration page.
At the top of the HBAC Services tab, click the Add link.
Click the checkbox by the names of the services to add, and click the right arrows button, >>, to move the command to the selection box.
Click the Add button to save the group membership.
15.2.2.2. Adding Service Groups in the Command Line
First create the service group entry, then create the service, and then add that service to the service group as a member. For example:
$ ipa hbacsvcgroup-add --desc="login services" login
--------------------------------
Added HBAC service group "login"
--------------------------------
Service group name: login
Description: login services
$ ipa hbacsvc-add --desc="SSHD service" sshd
-------------------------
Added HBAC service "sshd"
-------------------------
$ ipa hbacsvcgroup-add-member --hbacsvcs=sshd login
Service group name: login
Description: login services
-------------------------
Number of members added 1
-------------------------
FreeIPA defines two default service groups: SUDO for sudo services and FTP for services which provide FTP access.
15.3. Defining Host-Based Access Control Rules
Access controls, at a high level, define who has access to what. The who is a FreeIPA user, and the what can be either a host (target host), service, or service group, or a combination of the three.
15.3.1. Setting Host-Based Access Control Rules in the Web UI
Click the Policy tab.
Click the Host-Based Access Control subtab, and then select the HBAC Rules link.
Click the Add link at the top of the list of host-based access control rules.
Enter the name for the rule.
Click the Add and Edit button to go immediately to set the configuration for the rule.
There are a number of configuration areas for the rule. The four basic elements are who the rule applies to, what hosts allow access (the target), and, optionally, what services can be accessed.
In the Who area, select the users or user groups to which the access control rule is applied.
To apply the rule to all FreeIPA users, select the Anyone radio button.
To apply the rule to a specific set of users or user groups:
Select the Specified Users and Groups radio button.
Click the + Add link at the right of the users list.
Click the checkbox by the users to add to the rule, and click the right arrows button, >>, to move the users to the selection box.
Click Add.
In the Accessing area, select the target hosts which can be accessed through this access control rule.
To apply the rule to all FreeIPA hosts, select the Any Host radio button.
To apply the rule to a specific set of hosts or host groups:
Select the Specified Hosts and Groups radio button.
Click the + Add link at the right of the hosts list.
Click the checkbox by the hosts to include with the rule, and click the right arrows button, >>, to move the hosts to the selection box.
Click Add.
In the Via Service area, select specific services on the target hosts which the users are allowed to use to access target machines.
To apply the rule to all FreeIPA hosts, select the Any Service radio button.
To apply the rule to a specific set of hosts or host groups:
Select the Specified Services and Groups radio button.
Click the + Add link at the right of the commands list.
Click the checkbox by the services or groups to include with the rule, and click the right arrows button, >>, to move the services to the selection box.
Click Add.
15.3.2. Setting Host-Based Access Control Rules in the Command Line
The basic outline of all the access control commands is:
$ ipa hbacrule-add* options ruleName
To set every user or every host as a target, use the category options, such as --usercat=all.
Example 15.1. Granting All Access to One Host
One simple rule is to grant every user access to a single server. The first command creates the entry and uses the category options to apply every user.
$ ipa hbacrule-add --usercat=all allGroup
--------------------------
Added HBAC rule "allGroup"
--------------------------
Rule name: allGroup
User category: all
Enabled: TRUE
The second rule adds the target host, using the hbacrule-add-host command:
$ ipa hbacrule-add-host --hosts=server.example.com allGroup
Rule name: allGroup
User category: all
Enabled: TRUE
Successful hosts/hostgroups:
member host: server.example.com
-------------------------
Number of members added 1
-------------------------
Example 15.2. Adding Control for a Single User to a Service
Another access control method is to specify which services users are allowed to use to access the target hosts.
First, for the user to have access to every machine, every host must be added as both a host and target. This can be done using the category options:
$ ipa hbacrule-add --hostcat=all sshd-jsmith
Since the access control rule applies to a specific user, the user is added to the rule using the hbacrule-add-user command:
$ ipa hbacrule-add-user --users=jsmith sshd-jsmith
Then, the service is added to the access control rule. (The service should have already been added to the access control system using the hbacsvc-add command.) This is the service that the user can use to connect to the machine.
$ ipa hbacrule-add-service --hbacsvcs=sshd sshd-jsmith
Example 15.3. Adding a Service Group to the Rule
While a single service can be added to a rule, it is also possible to add an entire service group. Like a single service, this uses the hbacrule-add-service command, only with the --hbacsvcgroups option that specifies the group name.
$ ipa hbacrule-add-service --hbacsvcgroups=login loginRule
Table 15.1. Host-Based Access Control Command and Options
|
Command
|
Description
|
Arguments
|
Source or Target Entry
|
|---|
|
hbacrule-add
|
Adds a new host-based access control rule.
|
--usercat=all, which applies the rule to every user
--hostcat=all, which sets every host as an allowed target server
--servicecat=all, which sets every configured service as an allowed target service
ruleName, which is the required unique identifier for the new rule
|
|
|
hbacrule-add-host
|
Adds a target host to the access control rule. A target host can be accessed by other servers and users in the domain.
|
--hosts, which adds an individual server or command-separated list of servers as an allowed target server
--hostgroups, which adds a host group to the rule and every host within the host group is an allowed target server
ruleName, which is the rule to which to add the target server
|
Target
|
|
hbacrule-add-service
|
Adds a service type to the rule.
|
--hbacsvcs, which adds an individual service type or a comma-separated list of service type as an allowed target service
--hbacsvcgroups, which adds a service group to the rule and every service within the service group is an allowed target service
ruleName, which is the rule to which to add the target service
|
Target
|
|
hbacrule-add-user
|
Adds a user to the access control rule. The user is then able to access any allowed target host or service within the domain.
|
--users, which adds an individual user or command-separated list of users to the rule
--groups, which adds a user group to the rule and, thus, every user within the group
ruleName, which is the rule to which to add the user
|
Source
|
|
hbacrule-disable | hbacrule-enable
|
Disables or enables a host-based access control rule. Rules can be disabled if their behavior needs to be evaluated (for troubleshooting or to test a new rule).
|
ruleName, which is the rule to disable or enable
|
|
15.4. Testing Host-Based Access Control Rules
Implementing host-based access controls effectively can be tricky because it requires that all of the hosts be properly configured and the access is properly applied to users and services.
The hbactest command can test different host-based access control scenarios to make sure that the rules are working as expected.
The hbactest command does not work with trusted Active Directory users. Active Directory user/group associations are determined dynamically, as a user logs in, and those data are not stored in the FreeIPA LDAP directory. The hbactest command, then, is unable to resolve the group memberships to check how access control rules will be applied.
15.4.1. The Limits of Host-Based Access Control Configuration
The access control configuration should always be tested before it is implemented to prevent authorization failures.
Host-based access control rules depend on a lot of interactions — between hosts, services, DNS lookups, and users. If any element is misconfigured, then the rule can behave in unexpected ways.
FreeIPA includes a testing tool to verify that access control rules are behaving in the expected way by testing the access in a defined scenario. There are several situations where this testing is useful:
A new rule needs to be tested before it is implemented.
There are problems with the existing rules, and the testing tool can identify what rule is behaving badly.
A subset of existing rules can be tested to see how they are performing.
15.4.2. Test Scenarios for Host-Based Access Control (CLI-Based)
The hbactest command does not work with trusted Active Directory users. Active Directory user/group associations are determined dynamically, as a user logs in, and those data are not stored in the FreeIPA LDAP directory. The hbactest command, then, is unable to resolve the group memberships to check how access control rules will be applied.
The hbactest command tests configured host-based access control rules in very specific situations. A test run defines:
The user to run the operation as to test the rule performance for that user (--user).
Using the login client Y (--service).
To target host Z (--host).
The rule to test (--rules); if this is not used, then all enabled rules are tested.
Optional The hbactest returns detailed information about which rules were matched, not matched, or invalid. This detailed rule output can be disabled using --nodetail, so the test simply runs and returns whether access was granted.
The hbactest script does not actually connect to the target host. Instead, it uses the rules within the FreeIPA database to simulate how those rules would be applied in a specific situation as if an SSSD client were connecting to the FreeIPA server.
More briefly, it performs a simulated test run based on the given information and configuration, but it does not actually attempt a service request against the target host.
Example 15.4. Testing All Active Rules
The most basic command checks all active rules. It requires a specific connection scenario, so the user, login service and target host have to be given, and the testing tool checks the connection.
$ ipa hbactest --user=jsmith --host=target.example.com --service=ssh
--------------------
Access granted: True
--------------------
notmatched: my-second-rule
notmatched: my-third-rule
matched: myrule
matched: allow_all
Example 15.5. Testing a Specific Rule
It is possible to check a specific rule (or several rules).
$ ipa hbactest --user=jsmith --host=target.example.com --service=ssh --rules=myrule
---------------------
Access granted: True
---------------------
notmatched: myrule
Example 15.6. Testing Specific Rules Plus All Enabled
The --rules option lists specific rules to test, but it may be useful to test the specified rules against all of the enabled rules in the domain. This can be done by adding the --enabled option, which includes the (unspecified) enabled rules along with the specified rules.
$ ipa hbactest --user=jsmith --host=target.example.com --service=ssh --rules=myrule --enabled
--------------------
Access granted: True
--------------------
matched: my-second-rule
notmatched: my-third-rule
matched: myrule
matched: allow_all
It is possible to run a similar comparison against disabled rules by using the --disable option. With the --rules option, the specified rule plus all of the disabled rules are checked. With the --disabled option, all disabled rules are checked.
15.4.3. Testing Host-Based Access Control Rules in the UI
Testing host-based access control can help confirm that the rule performs as expected before it is deployed or to troubleshoot a rule once it is already active.
The hbactest command does not work with trusted Active Directory users. Active Directory user/group associations are determined dynamically, as a user logs in, and those data are not stored in the FreeIPA LDAP directory. The hbactest command, then, is unable to resolve the group memberships to check how access control rules will be applied.
By the nature of host-based access control rules, a test must define and verify a very specific set of criteria, A test run defines:
The user to run the operation as to test the rule performance for that user (Who).
To target host Z (Accessing).
Using the login client Y (Via Service).
The rule to test; if this is not used, then all enabled rules are tested (Rules).
The test environment is defined on the HBAC TEST page in the Host Based Access Control tab under Policy. A series of tabs is set up for each configuration step.
Once the environment is defined, then the test is run simply by clicking a button on the Run Test page. The results show clearly whether access was granted or denied to the users, and then runs through the rules which matched the given parameters.
To change some of the parameters and check for other results, click the New Test button at the bottom of the test results page. If that button is not selected, the form is not reset, so a new test will not run, even if test settings are changed.
Chapter 16. Policy: Defining SELinux User Maps
Security-enhanced Linux (SELinux) sets rules over what system users can access processes, files, directories, and system settings. Both the system administrator and applications themselves can define security contexts that restrict or allow user access and even access from other applications.
As part of defining centralized security policies in the FreeIPA domain, FreeIPA provides a way to map FreeIPA users to SELinux users and automatically grant or restrict access to clients and services within the FreeIPA domain, per host, based on the defined SELinux policies.
16.1. About FreeIPA, SELinux, and Mapping Users
Security-enhanced Linux defines kernel-level, mandatory access controls for how users, processes, and applications can interact with other resources on a system. These rules for interactions, called contexts, look at the data and behavior characteristics of different objects on the system and then set rules, called policies, which create contexts based on the security implications of each specific object. This is in contrast to higher-level discretionary access controls which are concerned primarily with file ownership and user identity, without accounting for data criticality or applciation behavior.
System users are associated with an SELinux role. The role is assigned both a multi-layer security context (MLS) a multi-category security context (MCS). The MLS/MCS contexts confine users to what processes, files, and operations they can access on the system.
SELinux users and policies function at the system level, not the network level. This means that SELinux users are configured independently on each system. While this is acceptable in many situations — SELinux has common defined system users and SELinux-aware services define their own policies — it has some issues when dealing with remote users and systems that access local resources. Remote users and services can get shuffled into a default guest context without a lot of intelligence about what their actual SELinux user and role should be.
This is how FreeIPA can cleanly integrate an identity domain with local SELinux services. FreeIPA can map FreeIPA users to configured SELinux roles per host. Mapping SELinux and FreeIPA users improves user administration:
Remote users can be granted appropriate SELinux user contexts based on their FreeIPA group assignments. This also allows administrators to consistently apply the same policies to the same users without having to create local accounts or reconfigure SELinux.
SELinux users are automatically updated as hosts are added to the IT environment or as users are added, removed, or changed, without having to edit local systems.
SELinux policies can be planned and related to domain-wide security policies through settings like FreeIPA host-based access control rules.
Administrators gain environment-wide visibility and control over how users and systems are assigned in SELinux.
SELinux user maps are comprised of three parts: the SELinux user for the system, an FreeIPA user, and an FreeIPA host. These define two separate relationships. First, it defines a map for the SELinux user on a specific host (the local or target system). Second, it defines a map for the SELinux user and the FreeIPA user.
This arrangement allows administrators to set different SELinux users for the same FreeIPA users, depending on which host they are accessing.
SELinux user maps work with the Systerm Security Services Daemon (SSSD) and the pam_selinux module. When a remote user attempts to log into a machine, SSSD checks its FreeIPA identity provider to collect the user information, including any SELinux maps. The PAM module then processes the user and assigns it the appropriate SELinux user context.
The core of an SElinux mapping rule is the SELinux system user. Each map is associated with the SELinux user first. The SELinux users which are available for mapping are configured in the FreeIPA server, so there is a central and universal list. These are SELinux users which are configured on every host in the FreeIPA domain. By default, there are five common SELinux users defined:
In the FreeIPA server configuration, each SELinux user is configured with both its username and its MLS/MCS range, SELinux_username:MLS[:MCS], and this format is used to identify the SELinux user when configuring maps.
The FreeIPA user and host configuration is very flexible. Users and hosts can be explicitly and individually assigned to an SELinux user map individually, or user groups or host groups can be explicitly assigned to the map.
An extra layer of security is possible by using host-based access control rules. As long as the host-based access control rule defines a user and a host, it can be used for an SELinux user map. Host-based access control rules (described in
Chapter 15, Policy: Configuring Host-Based Access Control) help integrate SELinux user maps with other access controls in FreeIPA and can help limit or allow host-based user access for remote users, as well as defining local security contexts.
If a host-based access control rule is associated with an SELinux user map, the host-based access control rule cannot be deleted until it is removed from the SELinux user map configuration.
16.2. Configuring SELinux Users in FreeIPA
SELinux user maps, as the name implies, creates an association between an SELinux user and an FreeIPA user. Before that association can be established, the FreeIPA server has to be aware of what SELinux users are configured on the systems it manages.
The available SELinux users are part of the FreeIPA server configuration. This is a list, in order from most to least confined, of the SELinux users. The SELinux user entry itself has this format:
SELinux_username:MLS[:MCS]
The individual user entries are separated with a dollar sign ($).
Since there is no requirement on user entries to have an SELinux map, many entries may be unmapped. The FreeIPA server configuration can also set a default SELinux user (which is part of the larger SELinux map list) to use for otherwise unmapped FreeIPA user entries.
In the top menu, click the IPA Server main tab and the Configuration subtab.
Scroll to the bottom of the list of server configuration areas, to SELINUX OPTIONS.
Set the SELinux user configuration.
There are two areas that can be edited: the prioritized list of SELinux users and the default SELinux user to use for unmapped FreeIPA users.
The SELinux user map order gives the list of SELinux users, defined on the local Linux system , which are available for configuring mapping rules. This is a prioritized list, from most to least confined. Each SELinux user has the format SELinux_user:MLS.
The Default SELinux user field sets the SELinux user to use for unmapped FreeIPA users.
Click the Update link at the top of the page to save the changes.
Before SELinux mapping rules can be created, there has to be a defined and universal list of SELinux users which are available to be mapped. This is set in the FreeIPA server configuration:
[jsmith@server ~]$ ipa config-show
...
SELinux user map order: unconfined_u:s0-s0:c0.c1023$guest_u:s0$xguest_u:s0$user_u:s0-s0:c0.c1023$staff_u:s0-s0:c0.c1023
Default SELinux user: unconfined_u:s0-s0:c0.c1023
The SELinux user settings can be edited using the config-mod command.
Example 16.1. List of SELinux Users
The complete list of SELinux users is passed in the --ipaselinuxusermaporder option. This list sets a priority order, from most to least confined users.
The SELinux user entry itself has this format:
SELinux_user:MLS:MCS
The individual user entries are separated with a dollar sign ($).
For example:
[jsmith@server ~]$ ipa config-mod --ipaselinuxusermaporder="unconfined_u:s0-s0:c0.c1023$guest_u:s0$xguest_u:s0$user_u:s0-s0:c0.c1023$staff_u:s0-s0:c0.c1023"
The default SELinux user, used for unmapped entries, must be included in the user map list or the edit operation fails. Likewise, if the default is edited, it must be changed to a user in the SELinux map list or the map list must be updated first.
Example 16.2. Default SELinux User
FreeIPA users are not required to have a specific SELinux user mapped to their account. However, the local system still checks the FreeIPA entry for an SELinux user to use for the FreeIPA user account. The default SELinux user sets the fallback user to use for unmapped FreeIPA user entries; this is, by default, the default SELinux user for system users on Fedora, unconfined_u.
This default user can be changed with the --ipaselinuxusermapdefault. For example:
[jsmith@server ~]$ ipa config-mod --ipaselinuxusermapdefault="guest_u:s0"
16.3. Mapping SELinux Users and FreeIPA Users
An SELinux map associates an SELinux user with an FreeIPA user (or users). However, SELinux settings are local to each host system, so a map not only needs to map the SELinux user with an FreeIPA user but also with a host system.
The rule definition primarily identifies the SELinux user; the SELinux user is the basis of the rule.
The other half of the map is comprised of defined FreeIPA users and defined FreeIPA hosts. (There can be one single user or host or multiple users and hosts or user and host groups in the map.) The users and hosts can be defined either by explicitly listing users and hosts or by referencing a host-based access control rule.
The host-based access control rule must contain users and hosts, not just services.
In the top menu, click the Policy main tab and the SELinux User Mappings subtab.
In the list of mappings, click the Add button to create a new map.
Enter the name for the map and the SELinux user exactly as it appears in the FreeIPA server configuration. SElinux users have the format SELinux_username:MLS[:MCS].
Click Add and Edit to add the FreeIPA user information.
As described in the introduction, an SELinux map has three parts: the SELinux user and an FreeIPA user/host pairing. That FreeIPA user/host pair can be defined in one of two ways: it can be set for explicit users on explicit hosts, or it can be defined using a host-based access control rule.
To set a host-based access control rule, select the rule from the drop-down menu in the General area of the configuration. Using a host-based access control rule also introduces access controls on what hosts a remote user can use to access a target machine. Only one host-based access control rule can be set.
Alternatively, scroll down the Users and Hosts areas, and click the Add link to assign users, user groups, hosts, or host groups to the SELinux map.
Select the users (or hosts or groups) on the left, click the right arrows button (>>) to move them to the Prospective column, and click the Add button to add them to the rule.
Either a host-based access control rule can be given or the users and hosts can be set manually. Both options cannot be used at the same time.
Click the Update link at the top to save the changes to the SELinux user map.
An SELinux map rule has three fundamental parts:
The SELinux user (--selinuxuser)
The user or user groups which are associated with the SELinux user (--users or --groups)
The host or host groups which are associated with the SELinux user (--hosts or --hostgroups)
Alternatively, a host-based access control rule which specifies both hosts and users in it (--hbacrule)
A rule can be created with all information at once using the selinuxusermap-add command. Users and hosts can be added to a rule after it is created by using the selinuxusermap-add-user and selinuxusermap-add-host commands, respectively.
Example 16.3. Creating a New SELinux Map
The --selinuxuser value must be the SELinux user name exactly as it appears in the FreeIPA server configuration. SElinux users have the format SELinux_username:MLS[:MCS].
Both a user and a host (or appropriate groups) must be specified for the SELinux mapping to be valid. Users, hosts, or groups can be specified in comma-separated lists.
[jsmith@server ~]$ ipa selinuxusermap-add --users=jsmith,bjensen,jrockford --hosts=server.example.com,test.example.com --selinuxuser="xguest_u:s0" selinux1
Example 16.4. Creating an SELinux Map with a Host-Based Access Control Rule
The --hbacrule value identifies the host-based access control rule to use for mapping. Using a host-based access control rule introduces access controls on what hosts a remote user can use to access a target machine, along with applying SELinux contexts after the remote user as logged into the target machine.
The access control rule must specify both users and hosts appropriately so that the SELinux map can construct the SELinux user, FreeIPA user, and host triple.
Only one host-based access control rule can be specified.
[jsmith@server ~]$ ipa selinuxusermap-add --hbacrule=webserver --selinuxuser="xguest_u:s0" selinux1
Example 16.5. Adding a User to an SELinux Map
While all of the users and hosts can be added to a map when it is created, users and hosts can also be added after the rule is created. This is done using a specific command, either selinuxusermap-add-user or selinuxusermap-add-host.
[jsmith@server ~]$ ipa selinuxusermap-add-user --users=jsmith selinux1
It is not necessary to use a separate command to add a host-based access control rule after the rule is configured because there can only be one. If the selinuxusermap-mod command is used with the --hbacrule option, it adds the host-based access control rule or overwrites the previous one.
A specific user or host can be removed from an SELinux map by using either the selinuxusermap-remove-host or selinuxusermap-remove-user command.
Example 16.6. Removing a User from an SELinux Map
[jsmith@server ~]$ ipa selinuxusermap-remove-user --users=jsmith selinux1
16.4. Troubleshooting SELinux Login Problems
SELinux maps only work for remote users, not for users with a local account.
When a remote user logs in, authenticating against the FreeIPA server, then the PAM SElinux modules create a file for that user in /etc/selinux/policy_name/logins/login.
If that file does not exist, then it means that SSSD is not properly configured to use the FreeIPA server as one of its identity providers. This is required for SELinux mapping to work. Configuring SSSD is covered in the
Red Hat 6 Deployment Guide.
If the file exists but the remote user was given the wrong SELinux context, then the pam_selinux module may not be properly configured in the PAM stack. This is the module that reads the SELinux information and sets the user context. If the module is missing, then nothing processes the SELinux map and the user is defined a default context on the system.
Chapter 17. Policy: Defining Automatic Group Membership for Users and Hosts
Most of the policies and configuration within the FreeIPA domain are based on groups. Settings from sudo rules to automount to access control are defined for groups, and then those settings are applied to group members.
Managing group membership is an important factor in managing users and hosts. Creating automember groups defines rules to add users and hosts to specified groups automatically, as soon as a new entry is added.
17.1. About Automembership
One of the most critical tasks for managing policies, identities, and security is managing group membership in FreeIPA. Groups are the core of most policy configuration.
By default, hosts do not belong to any group when they are created; users are added to the catchall ipausers group. Even if custom groups are configured and all policy configuration is in place, users and hosts cannot take advantage of those policies until they are joined to groups. Of course, this can be done manually, but it is both more efficient and more consistent if group membership can be assigned automatically.
This is done with automembership groups.
Automembership is essentially an automatic, global entry filter that organizes entries, at least in part, based on specific criteria. An automember rule, then, is the way that that filter is specified.
For example, there can be a lot of different, repeatable ways that to categorize identities within the IT and organizational environment:
Adding all hosts or all users to a single global group.
Adding employees to specific groups based on their employee type, ID number, manager, or physical location.
Dividing hosts based on their IP address or subnet.
Automembers provide a way to pre-sort those entries. That makes it easier to configure the actual behavior that you want to configure — like granting different sudo rules to different user types or machines on different subnets or have different automount settings for different users.
Automembership only applies to new users or groups. Changing the configuration on an existing user or group does not affect group membership, either by adding or removing the user/host in the group.
Automembership is a flag or a target set on an existing user group or host group. An automembership rule is created as a policy. This is a sister entry to the actual group entry and it signals that the given group is used for automatic group membership.
Once the rule is created — once the group is identified as being a target — then the next step is to define automember conditions. Conditions are regular expression filters that are used to identify group members. Conditions can be inclusive or exclusive, meaning that matching entries can be added or ignored based on those conditions.
There can be multiple conditions in a single rule. A user or host entry can match multiple rules and be added to multiple groups.
Automembership is a way of imposing reliable order on user and host group entries as they are created.
The key to using automember groups effectively is to plan your overall FreeIPA structure — the access control policies, sudo rules, host/service management rules, host groups, and user groups.
Once the structure is in place, then several things are clear:
What groups will be used in the FreeIPA
What specific groups different types of users and hosts need to belong to to perform their designated functions
What delineating attributes can be used to filter users and hosts into the appropriate groups
17.2. Defining Automembership Rules (Basic Procedure)
Open the Policy tab, and select the Automembers subtab.
In the top of the Automembers area, select the type of autogroup to create, either USER GROUP RULES or HOST GROUP RULES.
In the drop-down menu, select the group for which to create the automember rule.
Click the Add and Edit button.
In the edit page for the rule, click the + Add by the type of condition to create to identify entries.
Select the attribute to use as the basis for the search and then set the regular expression to use to match the attribute value.
Conditions can look for entries either to
include in the group or to explicitly
exclude from the group. The format of a condition is a Perl-compatible regular expression (PCRE). For more information on PCRE patterns, see
the pcresyntax(3) man page.
Exclude conditions are evaluated first and take precedence over include conditions.
Click Add and Add Another to add another condition. A single rule can have multiple include and exclude conditions. When all conditions have been configured, click the Add button to save the last condition and close the dialog window.
There are two commands used to define an automember policy:
A command to flag the group as an automember group, automember-add
A command to add regular expression conditions to identify group members, automember-add-condition
For example:
Create the automember policy entry for the group. Use the --type to identify whether the target group is a user group (group) or a host group (hostgroup). This command has the format:
ipa automember-add --type=group|hostgroup groupName
For example:
[jsmith@server ~]$ ipa automember-add --type=group exampleGroup
Create the conditions for the rule. To set multiple patterns, either give a comma-separated list of patterns in the --inclusive-regex|--exclusive-regex options or run the command multiple times.
This command has the format:
ipa automember-add-condition --type=group|hostgroup --key=attribute --inclusive-regex=regex | --exclusive-regex=regex groupName
As with the automember rule, the condition must specify the type of group (--type) and the name of the target group (groupName).
The condition must also specify the attribute (the key) and any patterns for the attribute value. The --key is the attribute name that is the focus of the condition. Then, there is a regular expression pattern to identify matching values; matching entries can either be included (--inclusive-regex) or excluded (--exclusive-regex) from the group. Exclusion rules take precedence.
For example, to include all employees with Barbara Jensen as a manager, but excluding the temporary employees:
[jsmith@server ~]$ ipa automember-add-condition --type=group --key=manager --inclusive-regex=^uid=bjensen$ exampleGroup
[jsmith@server ~]$ ipa automember-add-condition --type=group --key=employeetype --exclusive-regex=^temp exampleGroup
The regular expression can match any port of the string. Using a caret (^) means that it must match at the beginning, and using a dollar sign ($) means that it must match at the end. Wrapping the pattern in ^ and $ means that it must be an exact match.
To remove a condition for a rule, pass the full condition information, both the key and the regular expression:
[jsmith@server ~]$ ipa automember-remove-condition --key=fqdn --type=hostgroup --inclusive-regex=^web[1-9]+\.example\.com webservers
To remove the entire rule, simply run the automember-del command.
17.3. Examples of Using Automember Groups
These examples are shown using the CLI; the same configuration can be performed in the web UI.
All entries can be added to a single, global group regardless of what other groups they are also added to.
Entries can be added to specific automember groups. If the new entry does not match any autogroup, then it is added to a default or fallback group.
These strategies are mutually exclusive. If an entry matches a global group, then it does match an automember group and would, therefore, not be added to the fallback group.
17.3.1. Setting an All Users/Hosts Rule
To add all users or all hosts to a single group, use an inclusive regular expression for some attribute (such as cn or fqdn) which all entries will contain.
A regular expression to match all entries is simply .*. For example, to add all hosts to the same host group:
[jsmith@server ~]$ ipa automember-add-condition --type=hostgroup allhosts --inclusive-regex=.* --key=fqdn
--------------------------------
Added condition(s) to "allhosts"
--------------------------------
Automember Rule: allhosts
Inclusive Regex: fqdn=.*
----------------------------
Number of conditions added 1
----------------------------
Every host added after that is automatically added to the allhosts group:
[jsmith@server ~]$ ipa host-add test.example.com
-----------------------------
Added host "test.example.com"
-----------------------------
Host name: test.example.com
Principal name: host/test.example.com@EXAMPLE.COM
Password: False
Keytab: False
Managed by: test.example.com
[jsmith@server ~]$ ipa hostgroup-show allhosts
Host-group: allhosts
Description: Default hostgroup
Member hosts: test.example.com
17.3.2. Defining Default Automembership Groups
There is a special command to set a default group, automember-default-group-set. This sets the group name (--default-group) and group type(--type), similar to an automember rule, but there is no condition to match. By definition, default group members are unmatched entries.
For example:
[jsmith@server ~]$ ipa automember-default-group-set --default-group=ipaclients --type=hostgroup
[jsmith@server ~]$ ipa automember-default-group-set --default-group=ipausers --type=group
A default group rule can be removed using the automember-default-group-remove command. Since there is only one default group for a group type, it is only necessary to give the group type, not the group name:
[jsmith@server ~]$ ipa automember-default-group-remove --type=hostgroup
17.3.3. Using Automembership Groups with Windows Users
When a user is created in FreeIPA, that user is automatically added as a member to the ipausers group (which is the default group for all new users, apart from any automember group). However, when a Windows user is synced over from Active Directory, that user is not automatically added to the ipausers group.
New Windows users can be added to the ipausers group, as with users created in FreeIPA, by using an automember group. Every Windows user is added with the ntUser object class; that object class can be used as an inclusive filter to identify new Windows users to add to the automember group.
First, define the ipausers group as an automember group:
[jsmith@server ~]$ ipa automember-add --type=group ipausers
Then, use the ntUser object class as a condition to add users:
[jsmith@server ~]$ ipa automember-add-condition ipausers --key=objectclass --type=group --inclusive-regex=ntUser
Chapter 18. Configuration: Defining Access Control within FreeIPA
Access control is a security system which defines who can access certain resources — from machines to services to entries — and what kinds of operations they are allowed to perform. FreeIPA provides several access control areas to make it very clear what kind of access is being granted and to whom it is granted. As part of this, FreeIPA draws a distinction between access controls to resources within the domain and access control to the FreeIPA configuration itself.
This chapter details the different internal access control mechanisms that are available for users within FreeIPA to the FreeIPA server and other FreeIPA users.
18.1. About Access Controls for FreeIPA Entries
Access control defines the rights or permissions users have been granted to perform operations on other users or objects.
18.1.1. A Brief Look at Access Control Concepts
The FreeIPA access control structure is based on standard LDAP access controls. Access within the FreeIPA server is based on the FreeIPA users (who are stored in the backend Directory Server instance) who are allowed to access other FreeIPA entities (which are also stored as LDAP entries in the Directory Server instance).
An access control rule has three parts:
Who can perform the operation. This is the entity who is being granted permission to do something; this is the actor. In LDAP access control models, this is called the bind rule because it defines who the user is (based on their bind information) and can optionally require other limits on the bind attempt, such as restricting attempts to a certain time of day or a certain machine.
What can be accessed. This defines the entry which the actor is allowed to perform operations on. This is the target of the access control rule.
What type of operation can be performed. The last part is determining what kinds of actions the user is allowed to perform. The most common operations are add, delete, write, read, and search. In FreeIPA, all users are implicitly granted read and search rights to all entries in the FreeIPA domain, with restrictions only for sensitive attributes like passwords and Kerberos keys. (Anonymous users are restricted from seeing security-related configuration, like sudo rules and host-based access control.)
The only rights which can be granted are add, delete, and write — the permissions required to modify an entry.
FreeIPA does not provide a way to grant read access explicitly, and this is an important distinction from standard LDAP access control rules. In LDAP, all operations, including read, are implicitly denied and must be explicitly granted. In FreeIPA, read and search access are implicitly granted to any authenticated user.
Because read access is already granted, there is no way through the UI to grant read access. However, there is an option in the CLI tools to grant read access for special cases where there may be a broad deny rule set but read access should be granted to specific attributes. For example, read access is blocked to password attributes, but could be allowed by a special read permission.
When any operation is attempted, the first thing that the FreeIPA client does is send user credentials as part of the bind operation. The backend Directory Server checks those user credentials and then checks the user account to see if the user has permission to perform the requested operation.
18.1.2. Access Control Methods in FreeIPA
To make access control rules simple and clear to implement, FreeIPA divides access control definitions into three categories:
Self-service rules, which define what operations a user can perform on his own personal entry. The access control type only allows write permissions to attributes within the entry; it does not allow add or delete operations for the entry itself.
Delegation rules, which allow a specific user group to perform write (edit) operations on specific attributes for users in another user group. Like self-service rules, this form of access control rule is limited to editing the values of specific attributes; it does not grant the ability to add or remove whole entries or control over unspecified attributes.
Role-based access control, which creates special access control groups which are then granted much broader authority over all types of entities in the FreeIPA domain. Roles can be granted edit, add, and delete rights, meaning they can be granted complete control over entire entries, not just selected attributes.
Some roles are already created and available within FreeIPA. Special roles can be created to manage any type of entry in specific ways, such as hosts, automount configuration, netgroups, DNS settings, and FreeIPA configuration.
18.2. Defining Self-Service Settings
Self-service access control rules define the operations that an entity can perform on itself. These rules define only what attributes a user (or other FreeIPA entity) can edit on their personal entries.
Two self-service rules exist by default:
A rule for editing some general attributes in the personal entry, including given name and surname, phone numbers, and addresses.
A rule to edit user passwords, including two Samba passwords, the Kerberos password, and the general user password.
18.2.1. Creating Self-Service Rules from the Web UI
Open the IPA Server tab in the top menu, and select the Self Service Permissions subtab.
Click the Add link at the top of the list of self-service ACIs.
Enter the name of the rule in the pop-up window. Spaces are allowed.
Select the checkboxes by the attributes which this ACI will permit users to edit.
Click the Add button to save the new self-service ACI.
18.2.2. Creating Self-Service Rules from the Command Line
A new self-service rule can be added using the selfservice-add command. There are two required options, --permissions to set whether the ACI grants write, add, or delete permission and --attrs to give the full list of attributes which this ACI grants permission to.
$ ipa selfservice-add "Users can manage their own name details" --permissions=write --attrs=givenname,displayname,title,initials
-----------------------------------------------------------
Added selfservice "Users can manage their own name details"
-----------------------------------------------------------
Self-service name: Users can manage their own name details
Permissions: write
Attributes: givenname, displayname, title, initials18.2.3. Editing Self-Service Rules
In the self-service entry in the web UI, the only element that can be edited is the list of attributes that are included in the ACI. The checkboxes can be selected or deselected.
With the command line, self-service rules are edited using the ipa selfservice-mod command. The --attrs option overwrites whatever the previous list of supported attributes was, so always include the complete list of attributes along with any new attributes.
$ ipa selfservice-mod "Users can manage their own name details" --attrs=givenname,displayname,title,initials,surname
--------------------------------------------------------------
Modified selfservice "Users can manage their own name details"
--------------------------------------------------------------
Self-service name: Users can manage their own name details
Permissions: write
Attributes: givenname, displayname, title, initials
Include all of the attributes when modifying a self-service rule, including existing ones.
18.3. Delegating Permissions over Users
Delegation is very similar to roles in that one group of users is assigned permission to manage the entries for another group of users. However, the delegated authority is much more similar to self-service rules in that complete access is granted but only to specific user attributes, not to the entire entry. Also, the groups in delegated authority are existing FreeIPA user groups instead of roles specifically created for access controls.
18.3.1. Delegating Access to User Groups in the Web UI
Open the IPA Server tab in the top menu, and select the Delegations subtab.
Click the Add link at the top of the list of delegation ACIs.
Name the new delegation ACI.
In the User group drop-down menu, select the group whose entries can be edited by members of the delegation group.
In the Member user group drop-down menu, select the group who is being granted permissions to the entries of users in the user group.
In the attributes box, select the checkboxes by the attributes to which the member user group is being granted permission.
Click the Add button to save the new delegation ACI.
18.3.2. Delegating Access to User Groups in the Command Line
A new delegation access control rule is added using the delegation-add command. There are three required arguments:
--group, the group who is being granted permissions to the entries of users in the user group.
--membergroup, the group whose entries can be edited by members of the delegation group.
--attrs, the attributes which users in the member group are allowed to edit.
For example:
$ ipa delegation-add "basic manager attrs" --attrs=manager,title,employeetype,employeenumber --group=engineering_managers --membergroup=engineering
--------------------------------------
Added delegation "basic manager attrs"
--------------------------------------
Delegation name: basic manager attrs
Permissions: write
Attributes: manager, title, employeetype, employeenumber
Member user group: engineering
User group: engineering_managers
Delegation rules are edited using the delegation-mod command. The --attrs option overwrites whatever the previous list of supported attributes was, so always include the complete list of attributes along with any new attributes.
$ ipa delegation-mod "basic manager attrs" --attrs=manager,title,employeetype,employeenumber,displayname
-----------------------------------------
Modified delegation "basic manager attrs"
-----------------------------------------
Delegation name: basic manager attrs
Permissions: write
Attributes: manager, title, employeetype, employeenumber, displayname
Member user group: engineering
User group: engineering_managers
Include all of the attributes when modifying a delegation rule, including existing ones.
18.4. Defining Role-Based Access Controls
Role-based access control grants a very different kind of authority to users compared to self-service and delegation access controls. Role-based access controls are fundamentally administrative, with the potential to add, delete, and significantly modify entries.
There are three parts to role-based access controls:
The permission. The permission defines a specific operation or set of operations (write, add, or delete) and the target entries within the FreeIPA LDAP directory to which those operations apply. Permissions are building blocks; they can be assigned to multiple privileges as needed.
The privileges available to a role. A privilege is essentially a group of permissions. Permissions are not applied directly to a role. Permissions are added to a privilege so that the privilege creates a coherent and complete picture of a set of access control rules. For example, a permission can be created to add, edit, and delete automount locations. Then that permission can be combined with another permission relating to managing FTP services, and they can be used to create a single privilege that relates to managing filesystems.
The role. This is the list of FreeIPA users who are able to perform the actions defined in the privileges.
It is possible to create entirely new permissions, as well as to create new privileges based on existing permissions or new permissions. A list of the default privileges and their associated permissions are in
Table 18.1, “Privileges and Permissions in FreeIPA”.
FreeIPA does not provide a way to grant read access explicitly, and this is an important distinction from standard LDAP access control rules. In LDAP, all operations, including read, are implicitly denied and must be explicitly granted. In FreeIPA, read and search access are implicitly granted to any authenticated user.
Because read access is already granted, there is no way through the UI to grant read access. However, there is an option in the CLI tools to grant read access for special cases where there may be a broad deny rule set but read access should be granted to specific attributes. For example, read access is blocked to password attributes, but could be allowed by a special read permission.
Table 18.1. Privileges and Permissions in FreeIPA
|
Privilege
|
Associated Permissions
|
|---|
|
Automount Administrators
|
| Add_Automount_maps | | Remove_Automount_maps | | Add_Automount_keys | | Remove_Automount_keys |
|
|
Certificate Administrators
|
| Retrieve_Certificates_from_the_CA | | Request_Certificate | | Request_Certificates_from_a_different_hos | | Get_Certificates_status_from_the_CA | | Revoke_Certificate | | Certificate_Remove_Hold |
|
|
Delegation Administrator
|
| Add_Roles | | Remove_Roles | | Modify_Roles | | Modify_Role_membership | | Modify_privilege_membership |
|
|
DNS Administrators (for users)
|
| add_dns_entries | | remove_dns_entries | | update_dns_entries |
|
|
DNS Servers (for machines)
|
| add_dns_entries | | remove_dns_entries | | update_dns_entries |
|
|
Group Administrators
|
| Add_Groups | | Remove_Groups | | Modify_Groups | | Modify_Group_membership |
|
|
HBAC Administrator
|
| Add_HBAC_rule | | Delete_HBAC_rule | | Modify_HBAC_rule | | Manage_HBAC_rule_membership | | Add_HBAC_services | | Delete_HBAC_services | | Add_HBAC_service_groups | | Delete_HBAC_service_groups | | Manage_HBAC_service_group_membership |
|
|
Host Administrators
|
| Add_Hosts | | Remove_Hosts | | Modify_Hosts | | Manage_host_keytab | | Enroll_a_host | | Add_krbPrincipalName_to_a_host |
|
|
Host Enrollment
|
| Manage_host_keytab | | Enroll_a_host | | Add_krbPrincipalName_to_a_host |
|
|
Host Group Administrators
|
| Add_Hostgroups | | Remove_Hostgroups | | Modify_Hostgroups | | Modify_Hostgroup_membership |
|
|
Modify Users and Reset Passwords
|
|
|
Netgroups Administrators
|
| Add_netgroups | | Remove_netgroups | | Modify_netgroups | | Modify_netgroup_membership |
|
|
Password Policy Administrator
|
| Add_Group_Password_Policy_costemplate | | Delete_Group_Password_Policy_costemplate | | Modify_Group_Password_Policy_costemplate | | Add_Group_Password_Policy | | Delete_Group_Password_Policy | | Modify_Group_Password_Policy |
|
|
Replication Administrators
|
| Add_Replication_Agreements | | Remove_Replication_Agreements | | Modify_Replication_Agreements |
|
|
Service Administrators
|
| Add_Services | | Remove_Services | | Modify_Services | | Manage_service_keytab |
|
|
Sudo Administrator
|
| Add_Sudo_rule | | Delete_Sudo_rule | | Modify_Sudo_rule | | Add_Sudo_command | | Delete_Sudo_command | | Modify_Sudo_command | | Add_Sudo_command_group | | Delete_Sudo_command_group | | Manage_Sudo_command_group_membership |
|
|
User Administrators
|
| Change_a_user_password | | Add_user_to_default_group | | Unlock_user_accounts | | Remove_Users | | Modify_Users | | Add_Users |
|
|
Write IPA Configuration
|
|
18.4.1.1. Creating Roles in the Web UI
Open the IPA Server tab in the top menu, and select the Role Based Access Control subtab.
Click the Add link at the top of the list of role-based ACIs.
Enter the role name and a description.
Click the Add and Edit button to save the new role and go to the configuration page.
Open the Privileges tab in the role configuration page.
Click the Add link at the top of the list of privileges to add a new privilege.
Enter the role name and a description.
18.4.1.2. Creating Roles in the Command Line
Add the new role:
# ipa role-add --desc="User Administrator" useradmin
------------------------
Added role "useradmin"
------------------------
Role name: useradmin
Description: User Administrator
Add the required privileges to the role:
# ipa role-add-privilege --privileges="User Administrators" useradmin
Role name: useradmin
Description: User Administrator
Privileges: user administrators
----------------------------
Number of privileges added 1
----------------------------
Add the required groups to the role. In this case, we are adding only a single group, useradmin, which already exists.
# ipa role-add-member --groups=useradmins useradmin
Role name: useradmin
Description: User Administrator
Member groups: useradmins
Privileges: user administrators
-------------------------
Number of members added 1
-------------------------
18.4.2. Creating New Permissions
FreeIPA does not provide a way to grant read access explicitly, and this is an important distinction from standard LDAP access control rules. In LDAP, all operations, including read, are implicitly denied and must be explicitly granted. In FreeIPA, read and search access are implicitly granted to any authenticated user.
Because read access is already granted, there is no way through the UI to grant read access. However, there is an option in the CLI tools to grant read access for special cases where there may be a broad deny rule set but read access should be granted to specific attributes. For example, read access is blocked to password attributes, but could be allowed by a special read permission.
18.4.2.1. Creating New Permissions from the Web UI
Open the IPA Server tab in the top menu, and select the Role Based Access Control subtab.
Select the Permissions task link.
Click the Add link at the top of the list of permissions.
Enter the name of the new permission.
Select the checkboxes next to the allowed operations for this permission.
Select the method to use to identify the target entries from the Target drop-down menu. There are four different methods:
Type looks for an entry type like user, host, or service and then provides a list of all possible attributes for that entry type. The attributes which will be accessible through this ACI are selected from the list.
Filter uses an LDAP filter to identify which entries the permission applies to.
Subtree targets every entry beneath the specified subtree entry. All attributes within the matching entries can be modified.
Target group specifies a user group, and all the user entries within that group are available through the ACI. All attributes within the matching entries can be modified.
For Filter, Subtree, and Target group methods, no attributes are set in the UI. These must be added later using ipa permission-mod --attrs. If no attributes are set for the permission then, by default, all attributes are excluded.
Fill in the required information to identify the target entries, depending on the selected type.
Click the Add button to save the permission.
For Filter, Subtree, and Target group methods, set the attributes for the ACI to include. This must be done from the command line.
For example:
[jsmith@ipaserver ~]$ ipa permission-mod "manage Windows groups" --attrs=description,member
18.4.2.2. Creating New Permissions from the Command Line
A new permission is added using the permission-add command. All permissions require a list of attributes over which permission is granted (--attr), a list of allowed actions (--permissions), and the target entries for the ACI. There are four methods to identify the target entries:
--type looks for an entry type like user, host, or service and then provides a list of all possible attributes for that entry type.
--filter uses an LDAP filter to identify which entries the permission applies to.
--subtree targets every entry beneath the specified subtree entry.
--targetgroup specifies a user group, and all the user entries within that group are available through the ACI.
Example 18.1. Adding a Permission with a Filter
A filter can be any valid LDAP filter.
$ ipa permission-add "manage Windows groups" --filter="(!(objectclass=posixgroup))" --permissions=write --attrs=description
The permission-add command does not validate the given LDAP filter. Verify that the filter returns the expected results before configuring the permission.
Example 18.2. Adding a Permission for a Subtree
All a subtree filter requires is a DN within the directory. Since FreeIPA uses a simplified, flat directory tree structure, this can be used to target some types of entries, like automount locations, which are containers or parent entries for other configuration.
$ ipa permission-add "manage automount locations" --subtree="ldap://ldap.example.com:389/cn=automount,dc=example,dc=com" --permissions=write --attrs=automountmapname,automountkey,automountInformation
Example 18.3. Adding a Permission Based on Object Type
There seven object types that can be used to form a permission:
user
group
host
service
hostgroup
netgroup
dnsrecord
Each type has its own set of allowed attributes, in a comma-separated list.
$ ipa permission-add "manage service" --permissions=all --type=service --attrs=krbprincipalkey,krbprincipalname,managedby
The attributes (--attrs) must exist and be allowed attributes for the given object type, or the permission operation fails with schema syntax errors.
18.4.3. Creating New Privileges
18.4.3.1. Creating New Privileges from the Web UI
Open the IPA Server tab in the top menu, and select the Role Based Access Control subtab.
Select the Privileges task link.
Click the Add link at the top of the list of privileges.
Enter the name and a description of the privilege.
Click the Add and Edit button to go to the privilege configuration page to add permissions.
Select the Permissions tab.
Click the Add link at the top of the list of permissions to add permission to the privilege.
Click the checkbox by the names of the permissions to add, and click the right arrows button, >>, to move the permissions to the selection box.
Click the Add button.
18.4.3.2. Creating New Privileges from the Command Line
Privilege entries are created using the privilege-add command, and then permissions are added to the privilege group using the privilege-add-permission command.
Create the privilege entry.
$ ipa privilege-add "managing filesystems" --desc="for filesystems"
Assign the desired permissions. For example:
$ ipa privilege-add-permission "managing filesystems" --permissions="managing automount","managing ftp services"
Chapter 19. Configuration: Configuring the FreeIPA Server
The FreeIPA servers and backend services are configured with default settings that are applicable in most environments.
There are some configuration areas where the FreeIPA server configuration can be tweaked to improve security or performance in certain situations.
This chapter covers information about the FreeIPA configuration, including files and logs used by the FreeIPA server, and procedures for updating the FreeIPA server configuration itself.
19.1. FreeIPA Files and Logs
FreeIPA is a unifying framework that combines disparate Linux services into a single management context. However, the underlying technologies — such as Kerberos, DNS, 389 Directory Server, and Dogtag Certificate System — retain their own configuration files and log files. FreeIPA directly manages each of these elements through their own configuration files and tools.
This section covers the directories, files, and logs used specifically by FreeIPA. For more information about the configuration files or logs for a specific server used within FreeIPA, see the product documentation.
19.1.1. A Reference of FreeIPA Server Configuration Files and Directories
Table 19.1. FreeIPA Server Configuration Files and Directories
|
Directory or File
|
Description
|
|---|
|
Server Configuration
|
|
/etc/ipa
|
The main FreeIPA configuration directory.
|
|
/etc/ipa/default.conf
|
The primary configuration file for FreeIPA.
|
|
/etc/ipa/ca.crt
|
The CA certificate issued by the FreeIPA server's CA.
|
|
~/.ipa/
|
A user-specific FreeIPA directory that is created on the local system in the system user's home directory the first time the user runs a FreeIPA command.
|
|
FreeIPA Logs
|
|
~/.ipa/log/cli.log
|
The log file for all XML-RPC calls and responses by the FreeIPA command-line tools. This is created in the home directory for the system user who runs the tools, who may have a different name than the FreeIPA user.
|
|
/var/log/ipaclient-install.log
|
The installation log for the client service.
|
|
/var/log/ipaserver-install.log
|
The installation log for the FreeIPA server.
|
|
System Services
|
|
/etc/rc.d/init.d/ipa
|
The FreeIPA server init script.
|
|
Web UI
|
|
/etc/ipa/html
|
A symlink directory in the main configuration directory for the HTML files used by the FreeIPA web UI.
|
| /etc/httpd/conf.d/ipa.conf | | /etc/httpd/conf.d/ipa-rewrite.conf |
|
The configuration files used by the Apache host for the web UI application.
|
|
/etc/httpd/conf/ipa.keytab
|
The keytab file used by the web UI service.
|
|
/usr/share/ipa
|
The main directory for all of the HTML files, scripts, and stylesheets used by the web UI.
|
| /usr/share/ipa/ipa-rewrite.conf | | /usr/share/ipa/ipa.conf |
|
The configuration files used by the Apache host for the web UI application.
|
|
/usr/share/ipa/updates
|
Contains any updated files, schema, and other elements for FreeIPA.
|
|
/usr/share/ipa/html
|
Contains the HTML files, JavaScript files, and stylesheets used by the web UI.
|
|
/usr/share/ipa/ipaclient
|
Contains the JavaScript files used to access Firefox's autoconfiguration feature and set up the Firefox browser to work in the FreeIPA Kerberos realm.
|
|
/usr/share/ipa/migration
|
Contains HTML pages, stylesheets, and Python scripts used for running the FreeIPA server in migration mode.
|
|
/usr/share/ipa/ui
|
Contains all of the scripts used by the UI to perform FreeIPA operations.
|
|
/var/log/httpd
|
The log files for the Apache web server.
|
|
Kerberos
|
|
/etc/krb5.conf
|
The Kerberos service configuration file.
|
|
SSSD
|
|
/etc/sssd/sssd.api.d/sssd-ipa.conf
|
The configuration file used to identify the FreeIPA server, FreeIPA Directory Server, and other FreeIPA services used by SSSD.
|
|
/var/log/sssd
|
The log files for SSSD.
|
|
389 Directory Server
|
|
/var/lib/dirsrv/slapd-REALM_NAME
|
All of the schema, configuration, and database files associated with the Directory Server instance used by the FreeIPA server.
|
|
/var/log/dirsrv/slapd-REALM_NAME
|
Log files associated with the Directory Server instance used by the FreeIPA server.
|
|
Dogtag Certificate System
|
|
/etc/pki-ca
|
The main directory for the FreeIPA CA instance.
|
|
/etc/pki-ca/conf/CS.cfg
|
The main configuration file for the FreeIPA CA instance.
|
|
/var/lib/dirsrv/slapd-PKI-IPA/
|
All of the schema, configuration, and database files associated with the Directory Server instance used by the FreeIPA CA.
|
|
/var/log/dirsrv/slapd-PKI-IPA/
|
Log files associated with the Directory Server instance used by the FreeIPA CA.
|
|
Cache Files
|
|
/var/cache/ipa
|
Cache files for the FreeIPA server and the FreeIPA Kerberos password daemon.
|
|
System Backups
|
|
/var/lib/ipa/sysrestore
|
Contains backups of all of the system files and scripts that were reconfigured when the FreeIPA server was installed. These include the original .conf files for NSS, Kerberos (both krb5.conf and kdc.conf), and NTP.
|
|
/var/lib/ipa-client/sysrestore
|
Contains backups of all of the system files and scripts that were reconfigured when the FreeIPA client was installed. Commonly, this is the sssd.conf file for SSSD authentication services.
|
19.1.2. About default.conf and Context Configuration Files
Certain global defaults — like the realm information, the LDAP configuration, and the CA settings — are stored in the default.conf file. This configuration file is referenced when the FreeIPA client and servers start and every time the ipa command is run to supply information as operations are performed.
The parameters in the default.conf file are simple attribute=value pairs. The attributes are case-insensitive and order-insensitive.
[global]
basedn=dc=example,dc=com
realm=EXAMPLE.COM
domain=example.com
xmlrpc_uri=https://server.example.com/ipa/xml
ldap_uri=ldapi://%2fvar%2frun%2fslapd-EXAMPLE-COM.socket
enable_ra=True
ra_plugin=dogtag
mode=production
When adding more configuration attributes or overriding the global values, users can create additional context configuration files. A server.conf and cli.conf file can be created to create different options when the FreeIPA server is started or when the ipa command is run, respectively. The FreeIPA server checks the server.conf and cli.conf files first, and then checks the default.conf file.
Any configuration files in the /etc/ipa directory apply to all users for the system. Users can set individual overrides by creating default.conf, server.conf, or cli.conf files in their local FreeIPA directory, ~/.ipa/. This optional file is merged with default.conf and used by the local FreeIPA services.
19.1.3. Checking FreeIPA Server Logs
FreeIPA unifies several different Linux services, so it relies on those services' native logs for tracking and debugging those services.
The other services (Apache, 389 Directory Server, and Dogtag Certificate System) all have detailed logs and log levels. See the specific server documentation for more information on return codes, log formats, and log levels.
Table 19.2. FreeIPA Log Files
|
Service
|
Log File
|
Description
|
Additional Information
|
|---|
|
FreeIPA server
|
/var/log/ipaserver-install.log
|
Server installation log
|
|
|
FreeIPA server
|
~/.ipa/log/cli.log
|
Command-line tool log
|
|
|
FreeIPA client
|
/var/log/ipaclient-install.log
|
Client installation log
|
|
|
Apache server
|
| /var/log/httpd/access | | /var/log/httpd/error |
|
These are standard access and error logs for Apache servers. Both the web UI and the XML-RPC command-line interface use Apache, so some FreeIPA-specific messages will be recorded in the error log along with the Apache messages.
|
Apache log chapter
|
|
Dogtag Certificate System
|
/var/log/pki-ca-install.log
|
The installation log for the FreeIPA CA.
|
|
|
Dogtag Certificate System
|
| /var/log/pki-ca/debug | | /var/log/pki-ca/system | | /var/log/pki-ca/transactions | | /var/log/pki-ca/signedAudit |
|
These logs mainly relate to certificate operations. In FreeIPA, this is used for service principals, hosts, and other entities which use certificates.
|
Logging chapter
|
|
389 Directory Server
|
| /var/log/dirsrv/slapd-REALM/access | | /var/log/dirsrv/slapd-REALM/audit | | /var/log/dirsrv/slapd-REALM/errors |
|
The access and error logs both contain detailed information about attempted access and operations for the domain Directory Server instance. The error log setting can be changed to provide very detailed output.
|
The access log is buffered, so the server only writes to the log every 30 seconds, by default.
|
|
389 Directory Server
|
| /var/log/dirsrv/slapd-REALM/access | | /var/log/dirsrv/slapd-REALM/audit | | /var/log/dirsrv/slapd-REALM/errors |
|
This directory server instance is used by the FreeIPA CA to store certificate information. Most operational data here will be related to server-replica interactions.
|
The access log is buffered, so the server only writes to the log every 30 seconds, by default.
|
|
Kerberos
|
/var/log/krb5libs.log
|
This is the primary log file for Kerberos connections.
|
This location is configured in the krb5.conf file, so it could be different on some systems.
|
|
Kerberos
|
/var/log/krb5kdc.log
|
This is the primary log file for the Kerberos KDC server.
|
This location is configured in the krb5.conf file, so it could be different on some systems.
|
|
Kerberos
|
/var/log/kadmind.log
|
This is the primary log file for the Kerberos administration server.
|
This location is configured in the krb5.conf file, so it could be different on some systems.
|
|
DNS
|
/var/log/messages
|
DNS error messages are included with other system messages.
|
DNS logging is not enabled by default. DNS logging is enabled by running the querylog command:
/usr/sbin/rndc querylog
This begins writing log messages to the system's /var/log/messages file. To turn off logging, run the querylog command again.
|
19.1.3.1. Enabling Server Debug Logging
Debug logging for the FreeIPA server is set in the server.conf file.
Editing the defaults.conf configuration file affects all FreeIPA components, not only the FreeIPA server.
Edit or create the server.conf file.
vim server.conf
Add the debug line and set its value to true.
[global]
debug=True
Restart the Apache daemon to load the changes.
service httpd restart
19.1.3.2. Debugging Command-Line Operations
Any command-line operation with the ipa command can return debug information by using the -v option. For example:
$ ipa -v user-show admin
ipa: INFO: trying https://ipaserver.example.com/ipa/xml
First name: John
Last name: Smythe
User login [jsmythe]:
ipa: INFO: Forwarding 'user_add' to server u'https://ipaserver.example.com/ipa/xml'
--------------------
Added user "jsmythe"
--------------------
User login: jsmythe
First name: John
Last name: Smythe
Full name: John Smythe
Display name: John Smythe
Initials: JS
Home directory: /home/jsmythe
GECOS field: John Smythe
Login shell: /bin/sh
Kerberos principal: jsmythe@EXAMPLE.COM
UID: 1966800003
GID: 1966800003
Keytab: False
Password: False
Using the option twice,
-vv, displays the XML-RPC exchange:
$ ipa -vv user-add
ipa: INFO: trying https://ipaserver.example.com/ipa/xml
First name: Jane
Last name: Russell
User login [jrussell]:
ipa: INFO: Forwarding 'user_add' to server u'https://ipaserver.example.com/ipa/xml'
send: u'POST /ipa/xml HTTP/1.0\r\nHost: ipaserver.example.com\r\nAccept-Language: en-us\r\nAuthorization: negotiate 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\r\nUser-Agent: xmlrpclib.py/1.0.1 (by www.pythonware.com)\r\nContent-Type: text/xml\r\nContent-Length: 1240\r\n\r\n'
send: "<?xml version='1.0' encoding='UTF-8'?>\n<methodCall>\n<methodName>user_add</methodName>\n<params>\n<param>\n<value><array><data>\n<value><string>jrussell</string></value>\n</data></array></value>\n</param>\n<param>\n<value><struct>\n<member>\n<name>all</name>\n<value><boolean>0</boolean></value>\n</member>\n<member>\n<name>displayname</name>\n<value><string>Jane Russell</string></value>\n</member>\n<member>\n<name>cn</name>\n<value><string>Jane Russell</string></value>\n</member>\n<member>\n<name>noprivate</name>\n<value><boolean>0</boolean></value>\n</member>\n<member>\n<name>uidnumber</name>\n<value><int>999</int></value>\n</member>\n<member>\n<name>raw</name>\n<value><boolean>0</boolean></value>\n</member>\n<member>\n<name>version</name>\n<value><string>2.11</string></value>\n</member>\n<member>\n<name>gecos</name>\n<value><string>Jane Russell</string></value>\n</member>\n<member>\n<name>sn</name>\n<value><string>Russell</string></value>\n</member>\n<member>\n<name>krbprincipalname</name>\n<value><string>jrussell@EXAMPLE.COM</string></value>\n</member>\n<member>\n<name>givenname</name>\n<value><string>Jane</string></value>\n</member>\n<member>\n<name>initials</name>\n<value><string>JR</string></value>\n</member>\n</struct></value>\n</param>\n</params>\n</methodCall>\n"
reply: 'HTTP/1.1 200 OK\r\n'
header: Date: Thu, 15 Sep 2011 00:50:39 GMT
header: Server: Apache/2.2.15 (Red Hat)
header: WWW-Authenticate: Negotiate YIGZBgkqhkiG9xIBAgICAG+BiTCBhqADAgEFoQMCAQ+iejB4oAMCARKicQRvVl5x6Zt9PbWNzvPEWkdu+3PTCq/ZVKjGHM+1zDBz81GL/f+/Pr75zTuveLYn9de0C3k27vz96fn2HQsy9qVH7sfqn0RWGQWzl+kDkuD6bJ/Dp/mpJvicW5gSkCSH6/UCNuE4I0xqwabLIz8MM/5o
header: Connection: close
header: Content-Type: text/xml; charset=utf-8
body: "<?xml version='1.0' encoding='UTF-8'?>\n<methodResponse>\n<params>\n<param>\n<value><struct>\n<member>\n<name>result</name>\n<value><struct>\n<member>\n<name>dn</name>\n<value><string>uid=jrussell,cn=users,cn=accounts,dc=example,dc=com</string></value>\n</member>\n<member>\n<name>has_keytab</name>\n<value><boolean>0</boolean></value>\n</member>\n<member>\n<name>displayname</name>\n<value><array><data>\n<value><string>Jane Russell</string></value>\n</data></array></value>\n</member>\n<member>\n<name>uid</name>\n<value><array><data>\n<value><string>jrussell</string></value>\n</data></array></value>\n</member>\n<member>\n<name>objectclass</name>\n<value><array><data>\n<value><string>top</string></value>\n<value><string>person</string></value>\n<value><string>organizationalperson</string></value>\n<value><string>inetorgperson</string></value>\n<value><string>inetuser</string></value>\n<value><string>posixaccount</string></value>\n<value><string>krbprincipalaux</string></value>\n<value><string>krbticketpolicyaux</string></value>\n<"
body: 'value><string>ipaobject</string></value>\n</data></array></value>\n</member>\n<member>\n<name>loginshell</name>\n<value><array><data>\n<value><string>/bin/sh</string></value>\n</data></array></value>\n</member>\n<member>\n<name>uidnumber</name>\n<value><array><data>\n<value><string>1966800004</string></value>\n</data></array></value>\n</member>\n<member>\n<name>initials</name>\n<value><array><data>\n<value><string>JR</string></value>\n</data></array></value>\n</member>\n<member>\n<name>gidnumber</name>\n<value><array><data>\n<value><string>1966800004</string></value>\n</data></array></value>\n</member>\n<member>\n<name>gecos</name>\n<value><array><data>\n<value><string>Jane Russell</string></value>\n</data></array></value>\n</member>\n<member>\n<name>sn</name>\n<value><array><data>\n<value><string>Russell</string></value>\n</data></array></value>\n</member>\n<member>\n<name>homedirectory</name>\n<value><array><data>\n<value><string>/home/jrussell</string></value>\n</data></array></value>\n</member>\n<member>\n<name>has_password</name>\n<value><boolean>0</'
body: 'boolean></value>\n</member>\n<member>\n<name>krbprincipalname</name>\n<value><array><data>\n<value><string>jrussell@EXAMPLE.COM</string></value>\n</data></array></value>\n</member>\n<member>\n<name>givenname</name>\n<value><array><data>\n<value><string>Jane</string></value>\n</data></array></value>\n</member>\n<member>\n<name>cn</name>\n<value><array><data>\n<value><string>Jane Russell</string></value>\n</data></array></value>\n</member>\n<member>\n<name>ipauniqueid</name>\n<value><array><data>\n<value><string>bba27e6e-df34-11e0-a5f4-001143d2c060</string></value>\n</data></array></value>\n</member>\n</struct></value>\n</member>\n<member>\n<name>value</name>\n<value><string>jrussell</string></value>\n</member>\n<member>\n<name>summary</name>\n<value><string>Added user "jrussell"</string></value>\n</member>\n</struct></value>\n</param>\n</params>\n</methodResponse>\n'
---------------------
Added user "jrussell"
---------------------
User login: jrussell
First name: Jane
Last name: Russell
Full name: Jane Russell
Display name: Jane Russell
Initials: JR
Home directory: /home/jrussell
GECOS field: Jane Russell
Login shell: /bin/sh
Kerberos principal: jrussell@EXAMPLE.COM
UID: 1966800004
GID: 1966800004
Keytab: False
Password: False
The -v and -vv options are global options and must be used before the subcommand when running ipa.
19.2. Disabling Anonymous Binds
Accessing domain resources and running client tools always require Kerberos authentication. However, the backend LDAP directory used by the FreeIPA server allows anonymous binds by default. This potentially opens up all of the domain configuration to unauthorized users, including information about users, machines, groups, services, netgroups, and DNS configuration.
It is possible to disable anonymous binds on the 389 Directory Server instance by using LDAP tools to reset the nsslapd-allow-anonymous-access attribute.
Change the nsslapd-allow-anonymous-access attribute to rootdse.
ldapmodify -x -D "cn=Directory Manager" -w secret -h server.example.com -p 389
Enter LDAP Password:
dn: cn=config
changetype: modify
replace: nsslapd-allow-anonymous-access
nsslapd-allow-anonymous-access: rootdse
Anonymous access can be completely allowed (on) or completely blocked (off). However, completely blocking anonymous access also blocks external clients from checking the server configuration. LDAP and web clients are not necessarily domain clients, so they connect anonymously to read the root DSE file to get connection information.
The rootdse allows access to the root DSE and server configuration without any access to the directory data.
Restart the 389 Directory Server instance to load the new setting.
service dirsrv restart
19.3. Configuring Alternate Certificate Authorities
FreeIPA creates a Dogtag Certificate System certificate authority (CA) during the server installation process. To use an external CA, it is possible to create the required server certificates and then import them into the 389 Directory Server and the HTTP server, which require FreeIPA server certificates.
Save an ASCII copy of the CA certificate as /usr/share/ipa/html/ca.crt. This allows users to download the correct certificate when they configure their browsers.
Use the ipa-server-certinstall command to install the certificate.
# /usr/sbin/ipa-server-certinstall -d /path/to/pkcs12.p12
To keep using browser autoconfiguration in Firefox, regenerate the /usr/share/ipa/html/configure.jar file.
Create a directory, and then create the new security databases in that directory.
# mkdir /tmp/signdb
# certutil -N -d /tmp/signdb
Import the PKCS #12 file for the signing certificate into that directory.
# pk12util -i /path/to/pkcs12.p12 -d /tmp/signdb
Make a temporary signing directory, and copy the FreeIPA JavaScript file to that directory.
# mkdir /tmp/sign
# cp /usr/share/ipa/html/preferences.html /tmp/sign
Use the object signing certificate to sign the JavaScript file and to regenerate the
configure.jar file.
# signtool -d /tmp/signdb -k Signing_cert_nickname -Z /usr/share/ipa/html/configure.jar -e .html /tmp/sign
19.4. Configuring CRLs and OCSP Responders
A certificate is created with a validity period, meaning it has a point where it expires and is no longer valid. The expiration date is contained in the certificate itself, so a client always checks the validity period in the certificate to see if the certificate is still valid.
However, a certificate can also be revoked before its validity period is up, but this information is not contained in the certificate. A CA publishes a certificate revocation list (CRL), which contains a complete list of every certificate that was issued by that CA and subsequently revoked. A client can check the CRL to see if a certificate within its validity period has been revoked and is, therefore, invalid.
Validity checks are performed using the online certificate status protocol (OCSP), which sends a request to an OCSP responder. Each CA integrated with the FreeIPA server uses an internal OCSP responder, and any client which runs a validity check can check the FreeIPA CA's internal OCSP responder.
Every certificate issued by the FreeIPA CA puts its OCSP responder service URL in the certificate. For example:
http://ipaserver.example.com:9180/ca/ocsp
For the FreeIPA OCSP responder to be available, port 9180 needs to be open in the firewall.
19.4.1. Using an OSCP Responder with SELinux
Clients can use the FreeIPA OCSP responder to check certificate validity or to retrieve CRLs. A client can be a number of different services, but is most frequently an Apache server and the mod_revocator module (which handles CRL and OCSP operations).
The FreeIPA CA has an OCSP responder listening over port 9180, which is also the port available for CRL retrieval. This port is protected by default SELinux policies to prevent unauthorized access. If an Apache server attempts to connect to the OCSP port, then it may be denied access by SELinux.
The Apache server, on the local machine, must be granted access to port 9180 for it to be able to connect to the FreeIPA OCSP responder. There are two ways to work around this by changing the SELinux policies:
Edit the SELinux policy to allow Apache servers using the mod_revocator module to connect to port 9180:
semodule -i revoker.pp
Generate a new SELinux policy to allow access based on the SELinux error logs for the mod_revocator connection attempt.
audit2allow -a -M revoker
19.4.2. Changing the CRL Update Interval
The CRL file is automatically generated by the Dogtag Certificate System CA every four hours. This interval can be changed by editing the Dogtag Certificate System configuration.
Stop the CA server.
service pki-ca stop
Open the CS.cfg file.
vim /etc/pki-ca/CS.cfg
Change the ca.crl.MasterCRL.autoUpdateInterval to the new interval setting.
Restart the CA server.
service pki-ca start
19.4.3. Changing the OCSP Responder Location
Each FreeIPA server generates its own CRL. Likewise, each FreeIPA server uses its own OCSP responder, with its own OCSP responder URL in the certificates it issues.
A DNS CNAME can be used by FreeIPA clients, and then from there be redirected to the appropriate FreeIPA server OCSP responder.
Open the certificate profile.
vim /var/lib/pki-ca/profiles/ca/caIPAserviceCert.cfg
Change the policyset.serverCertSet.9.default.params.crlDistPointsPointName_0 parameter to the DNS CNAME hostname.
Restart the CA server.
service pki-ca restart
That change must be made on every FreeIPA server, with the crlDistPointsPointName_0 parameter set to the same hostname.
19.5. Setting DNS Entries for Multi-Homed Servers
Some server machines may support multiple network interface cards (NICs). Multi-homed machines typically have multiple IPs, all assigned to the same hostname. This works fine in FreeIPA most of the time because it listens on all available interfaces, except localhost. For a server to be available through any NIC, edit the DNS zone file and add entries for each IP address. For example:
ipaserver IN A 192.168.1.100
ipaserver IN A 192.168.1.101
ipaserver IN A 192.168.1.102
19.6. Managing Replication Agreements Between FreeIPA Servers
Information is shared between the FreeIPA servers and replicas using multi-master replication. What this means is that servers and replicas all receive updates and, therefore, are data masters. The domain information is copied between the servers and replicas using replication.
As replicas are added to the domain, mutual replication agreements are automatically created between the replica and the server it is based on. Additional replication agreements can be created between other replicas and servers or the configuration of the replication agreement can be changed using the ipa-replica-manage command.
When a replica is created, the replica install script creates two replication agreements: one going from the master server to the replica and one going from the replica to the master server.
As more replicas and servers are added to the domain, there can be replicas and servers that have replication agreements to other servers and replicas but not between each other. For example, the first FreeIPA server is Server A. Then, the admin creates Replica B, and the install script creates a Server A => Replica B replication agreement and a Replica B => Server A replication agreement. Next, the admin creates Replica C based on Server A. The install script creates a Server A => Replica C replication agreement and a Replica C => Server A replication agreement. Replica B and Replica C both have replication agreements with Server A — but they do not have agreements with each other. For data availability, consistency, failover tolerance, and performance, it can be beneficial to create a pair of replication agreements between Replica B and Replica C, even though their data will eventually be replicated over to each other through replication with Server A.
19.6.1. Listing Replication Agreements
The ipa-replica-manage command can list all of the servers and replicas in the replication topology, using the list command:
# ipa-replica-manage list
srv1.example.com
srv2.example.com
srv3.example.com
srv4.example.com
After getting the server/replica list, then it is possible to list the replication agreements for the server. These are the other servers/replicas to which the specified server sends updates.
# ipa-replica-manage list srv1.example.com
srv2.example.com
srv3.example.com
19.6.2. Creating and Removing Replication Agreements
Replication agreements are created by connecting one server to another server.
ipa-replica-manage server1 server2
If only one server is given, the replication agreements are created between the local host and the specified server.
For example:
# ipa-replica-manage connect srv2.example.com srv4.example.com
Replication occurs over standard LDAP; to enable SSL, then include the CA certificate for the local host (or the specified server1). The CA certificate is then installed in the remote server's certificate database to enable TLS/SSL connections. For example:
# ipa-replica-manage connect --cacert=/etc/ipa/ca.crt srv2.example.com srv4.example.com
To remove a replication agreement between specific servers/replicas, use the disconnect command:
# ipa-replica-manage disconnect srv2.example.com srv4.example.com
Using the disconnect command removes that one replication agreement but leaves both the server/replica instances in the overall replication topology. To remove a server entirely from the FreeIPA replication topology, with all its data, (and, functionally, removing it from the FreeIPA domain as a server), use the del server:
# ipa-replica-manage del srv2.example.com
19.6.3. Forcing Replication
Replication between servers and replicas occurs on a schedule. Although replication is frequent, there can be times when it is necessary to initiate the replication operation manually. For example, if a server is being taken offline for maintenance, it is necessary to flush all of the queued replication changes out of its changelog before taking it down.
To initiate a replication update manually, use the force-sync command. The server which receives the update is the local server; the server which sends the updates is specified in the --from option.
# ipa-replica-manage force-sync --from srv1.example.com
19.6.4. Reinitializing FreeIPA Servers
When a replica is first created, the database of the master server is copied, completely, over to the replica database. This process is called initialization. If a server/replica is offline for a long period of time or there is some kind of corruption in its database, then the server can be re-initialized, with a fresh and updated set of data.
This is done using the re-initialize command. The target server being initialized is the local host. The server or replica from which to pull the data to initialize the local database is specified in the --from option:
# ipa-replica-manage re-initialize --from srv1.example.com
19.6.5. Resolving Replication Conflicts
Changes — both for FreeIPA domain data and for certificate and key data — are replicated between FreeIPA servers and replicas (and, in similar paths, between FreeIPA and Active Directory servers).
Even though replication operations are run continuously, there is a chance that changes can be made on one FreeIPA server at the same time different changes are made to the same entry on a different FreeIPA server. When replication begins to process those entries, the changes collide — this is a replication conflict.
Every single directory modify operation is assigned a server-specific change state number (CSN) to track how those modifications are propagated during replication. The CSN also contains a modify timestamp. When there is a replication conflict, the timestamp is checked and the last change wins.
Simply accepting the most recent change is effective for resolving conflicts with attribute values. That method is too blunt for some types of operations, however, which affect the directory tree. Some operations, like modrdn, DN changes, or adding or removing parent and child entries, require administrator review before the conflict is resolved.
Replication conflicts are resolved by editing the entries directory in the LDAP database.
When there is a replication conflict, both entries are added to the directory and are assigned a nsds5ReplConflict attribute. This makes it easy to search for entries with a conflict:
ldapsearch -x -D "cn=directory manager" -w password -b "dc=example,dc=com" "nsds5ReplConflict=*" \* nsds5ReplConflict
19.6.5.1. Solving Naming Conflicts
When two entries are added to the FreeIPA domain with the same DN, both entries are added to the directory, but they are renamed to use the nsuniqueid attribute as a naming attribute. For example:
nsuniqueid=0a950601-435311e0-86a2f5bd-3cd26022+uid=jsmith,cn=users,cn=accounts,dc=example,dc=com
Those entries can be searched for and displayed in the FreeIPA CLI, but they cannot be edited or deleted until the conflict is resolved and the DN is updated.
To resolve the conflict:
Rename the entry using a different naming attribute, and keep the old RDN. For example:
ldapmodify -x -D "cn=directory manager" -w secret -h ipaserver.example.com -p 389
dn: nsuniqueid=66446001-1dd211b2+uid=jsmith,cn=users,cn=accounts,dc=example,dc=com
changetype: modrdn
newrdn: cn=TempValue
deleteoldrdn: 0
Remove the old RDN value of the naming attribute and the conflict marker attribute. For example:
ldapmodify -x -D "cn=directory manager" -w secret -h ipaserver.example.com -p 389
dn: cn=TempValue,cn=users,cn=accounts,dc=example,dc=com
changetype: modify
delete: uid
dc: jsmith
-
delete: nsds5ReplConflict
-
The unique identifier attribute nsuniqueid cannot be deleted.
Rename the entry with the intended attribute-value pair. For example:
ldapmodify -x -D "cn=directory manager" -w secret -h ipaserver.example.com -p 389
dn: cn=TempValue,dc=example,dc=com
changetype: modrdn
newrdn: uid=jsmith
deleteoldrdn: 1
Setting the value of the deleteoldrdn attribute to 1 deletes the temporary attribute-value pair cn=TempValue. To keep this attribute, set the value of the deleteoldrdn attribute to 0.
19.6.5.2. Solving Orphan Entry Conflicts
When a delete operation is replicated and the consumer server finds that the entry to be deleted has child entries, the conflict resolution procedure creates a glue entry to avoid having orphaned entries in the directory.
In the same way, when an add operation is replicated and the consumer server cannot find the parent entry, the conflict resolution procedure creates a glue entry representing the parent so that the new entry is not an orphan entry.
Glue entries are temporary entries that include the object classes
glue and
extensibleObject. Glue entries can be created in several ways:
If the conflict resolution procedure finds a deleted entry with a matching unique identifier, the glue entry is a resurrection of that entry, with the addition of the glue object class and the nsds5ReplConflict attribute.
In such cases, either modify the glue entry to remove the glue object class and the nsds5ReplConflict attribute to keep the entry as a normal entry or delete the glue entry and its child entries.
The server creates a minimalistic entry with the glue and extensibleObject object classes.
In such cases, modify the entry to turn it into a meaningful entry or delete it and all of its child entries.
Deleting or demoting a replica removes the FreeIPA replica from the server/replica topology so that it no longer processes FreeIPA requests and it also removes the host machine itself from the FreeIPA domain.
On an FreeIPA server, obtain a Kerberos ticket before running FreeIPA tools.
[root@replica ~]#kinit admin
List all of the configured replication agreements for the FreeIPA domain.
[root@replica ~]# ipa-replica-manage list
Directory Manager password:
ipaserver.example.com: master
ipaserver2.example.com: master
replica.example.com: master
replica2.example.com: master
Removing the replica from the topology involves deleting all the agreements between the replica and the other servers in the FreeIPA domain and all of the data about the replica in the domain configuration.
[root@replica ~]# ipa-replica-manage del replica.example.com
If the replica was configured with its own CA, then also use the ipa-csreplica-manage command to remove all of the replication agreements between the certificate databases for the replica.
This is required if the replica itself was configured with a Dogtag Certificate System CA. It is not required if only the master server or other replicas were configured with a CA.
[root@replica ~]# ipa-csreplica-manage del replica.example.com
On the replica, uninstall the replica packages.
[root@replica ~]# ipa-server-install --uninstall -U
19.8.1. Starting FreeIPA with Expired Certificates
If FreeIPA administrative server certificates expire, then most FreeIPA services will be inaccessible, including administrative services. The underlying Apache and 389 Directory Server services can be configured to allow SSL access to those services, even if the certificates are expired.
Allowing limited access with expired certificates permits Apache, Kerberos, DNS, and 389 Directory Server services to continue working. With those services active, users are able to log into the domain.
Client services such as sudo that require SSL for access will still fail because of the expired server certificates.
Change the mod_nss configuration for the Apache server to not enforce valid certificates, in the NSSEnforceValidCerts parameter. If this parameter is not already in the file, then add it.
Set the value to off.
[root@ipaserver ~]# vim /etc/httpd/conf.d/nss.conf
NSSEnforceValidCerts off
Restart Apache.
[root@ipaserver ~]# service httpd restart
Change the nsslapd-validate-cert attribute in the 389 Directory Server configuration to warn instead of true to disable validity checks.
[root@ipaserver ~]# ldapmodify -D "cn=directory manager" -w secret -p 389 -h ipaserver.example.com
dn: cn=config
changetype: modify
replace: nsslapd-validate-cert
nsslapd-validate-cert: warn
Restart 389 Directory Server.
[root@ipaserver ~]# service dirsrv restart
19.8.2. There are SASL, GSS-API, and Kerberos errors in the 389 Directory Server logs when the replica starts.
When the replica starts, there can be a series of SASL bind errors recorded in the 389 Directory Server logs stating that the GSS-API connection failed because it could not find a credentials cache:
slapd_ldap_sasl_interactive_bind - Error: could not perform interactive bind for id [] mech [GSSAPI]: error -2 (Local error) (SASL(-1): generic failure: GSSAPI Error: Unspecified GSS failure. Minor code may provide more information (Credentials cache file '/tmp/krb5cc_496' not found)) ...
The replica is looking for a credentials cache in /tmp/krb5cc_496 (where 496 is the 389 Directory Server user ID) and cannot find it.
There may also be messages that the server could not obtain Kerberos credentials for the host principal:
set_krb5_creds - Could not get initial credentials for principal [ldap/ replica1.example.com] in keytab [WRFILE:/etc/dirsrv/ds.keytab]: -1765328324 (Generic error)
These errors are both related to how and when the 389 Directory Server instance loads its Kerberos credentials cache.
While 389 Directory Server itself supports multiple different authentication mechanisms, FreeIPA only uses GSS-API for Kerberos connections. The 389 Directory Server instance for FreeIPA keeps its Kerberos credentials cache in memory. When the 389 Directory Server process ends — like when the FreeIPA replica is stopped — the credentials cache is destroyed.
Also, the 389 Directory Server is used as the backend storage for the principal information for the KDC.
When the replica then restarts, the 389 Directory Server instance starts first, since it supplies information for the KDC, and then the KDC server starts. This start order is what causes the GSS-API and Kerberos connection errors.
The 389 Directory Server attempts to open a GSS-API connection, but since there is no credentials cache yet and the KDC is not started, the GSS connection fails. Likewise, any attempt to obtain the host credentials also fails.
These errors are transient. The 389 Directory Server re-attempts the GSS-API connection after the KDC starts and it has a credentials cache. The 389 Directory Server logs then record a bind resumed message.
These startup GSS-API connection failures can be ignored as long as that connection is successfully established.
Chapter 20. Migrating from an LDAP Directory to FreeIPA
When an infrastructure has previously deployed an LDAP server for authentication and identity lookups, it is possible to migrate the user data, including passwords, to a new FreeIPA instance, without losing user or password data.
FreeIPA has migration tools to help move directory data and only requires minimal updates to clients. However, the migration process assumes a very simple deployment scenario (one LDAP directory namespace to one FreeIPA namespace).
20.1. An Overview of LDAP to FreeIPA Migration
The actual migration part of moving from an LDAP server to FreeIPA — the process of moving the data from one server to the other — is fairly straightforward. The process is simple: move data, move passwords, and move clients.
The crucial part of migration is not data migration; it is deciding how clients are going to be configured to use FreeIPA. For each client in the infrastructure, you need to decide what services (such as Kerberos and SSSD) are being used and what services can be used in the final, FreeIPA deployment.
A secondary, but significant, consideration is planning how to migrate passwords. FreeIPA requires Kerberos hashes for every user account in addition to passwords. Some of the considerations and migration paths for passwords are covered in
Section 20.1.2, “Planning Password Migration”.
20.1.1. Planning the Client Configuration
FreeIPA can support a number of different client configurations, with varying degrees of functionality, flexibility, and security. Decide which configuration is best for each individual client based on its operating system, functional area (such as development machines, production servers, or user laptops), and your IT maintenance priorities.
The different client configurations are not mutually exclusive. Most environments will have a mix of different ways that clients use to connect to the FreeIPA domain. Administrators must decide which scenario is best for each individual client.
20.1.1.1. Initial Client Configuration (Pre-Migration)
Before deciding where you want to go with the client configuration in FreeIPA, first establish where you are before the migration.
The initial state for almost all LDAP deployments that will be migrated is that there is an LDAP service providing identity and authentication services.
Linux and Unix clients use PAM_LDAP and NSS_LDAP libraries to connect directly to the LDAP services. These libraries allow clients to retrieve user information from the LDAP directory as if the data were stored in /etc/passwd or /etc/shadow. (In real life, the infrastructure may be more complex if a client uses LDAP for identity lookups and Kerberos for authentication or other configurations.)
There are structural differences between an LDAP directory and a FreeIPA server, particularly in schema support and the structure of the directory tree. (For more background on those differences, see
Section 1.1, “FreeIPA v. LDAP: A More Focused Type of Service”.) While those differences may impact data (especially with the directory tree, which affects entry names), they have little impact on the
client configuration, so it really has little impact on migrating clients to FreeIPA.
20.1.1.2. Recommended Configuration for Fedora Clients
Fedora has a service called the System Security Services Daemon (SSSD). SSSD uses special PAM and NSS libraries (pam_sss and nss_sss, respectively) which allow SSSD to be integrated very closely with FreeIPA and leverage the full authentication and identity features in FreeIPA. SSSD has a number of useful features, like caching identity information so that users can log in even if the connection is lost to the central server; these are described in the Fedora Deployment Guide.
Unlike generic LDAP directory services (using pam_ldap and nss_ldap), SSSD establishes relationships between identity and authentication information by defining domains. A domain in SSSD defines four backend functions: authentication, identity lookups, access, and password changes. The SSSD domain is then configured to use a provider to supply the information for any one (or all) of those four functions. An identity provider is always required in the domain configuration. The other three providers are optional; if an authentication, access, or password provider is not defined, then the identity provider is used for that function.
SSSD can use FreeIPA for all of its backend functions. This is the ideal configuration because it provides the full range of FreeIPA functionality, unlike generic LDAP identity providers or Kerberos authentication. For example, during daily operation, SSSD enforces host-based access control rules and security features in FreeIPA.
During the migration process from an LDAP directory to FreeIPA, SSSD can seamlessly migrate user passwords without additional user interaction.
The ipa-client-install script automatically configured SSSD to use FreeIPA for all four of its backend services, so Fedora clients are set up with the recommended configuration by default.
20.1.1.3. Alternative Supported Configuration
Unix and Linux systems such as Mac, Solaris, HP-UX, AIX, and Scientific Linux support all of the services that FreeIPA manages but do not use SSSD. Likewise, older Fedora versions (15) support SSSD but have an older version, which does not support FreeIPA as an identity provider.
When it is not possible to use a modern version of SSSD on a system, then clients can be configured to connect to the FreeIPA server as if it were an LDAP directory service for identity lookups (using nss_ldap) and to FreeIPA as if it were a regular Kerberos KDC (using pam_krb5).
If a Fedora client is using an older version of SSSD, SSSD can still be configured to use the FreeIPA server as its identity provider and its Kerberos authentication domain; this is described in the SSSD configuration section of the Fedora Deployment Guide.
Any FreeIPA domain client can be configured to use nss_ldap and pam_krb5 to connect to the FreeIPA server. For some maintenance situations and IT structures, a scenario that fits the lowest common denominator may be required, using LDAP for both identity and authentication (nss_ldap and pam_ldap). However, it is generally best practice to use the most secure configuration possible for a client (meaning SSSD and Kerberos or LDAP and Kerberos).
20.1.2. Planning Password Migration
Probably the most visible issue that can impact LDAP-to-FreeIPA migration is migrating user passwords.
FreeIPA (by default) uses Kerberos for authentication and requires that each user have Kerberos hashes stored in the FreeIPA Directory Server in addition to the standard user passwords. To generate these hashes, the user password needs to be available to the FreeIPA server in cleartext. This is the case when the user is created in FreeIPA. However, when the user is migrated from an LDAP directory, the associated user password is already hashed, so the corresponding Kerberos key cannot be generated.
Users cannot authenticate to the FreeIPA domain or access FreeIPA resources until they have Kerberos hashes.
If a user does not have a Kerberos hash
, that user cannot log into the FreeIPA domain even if he has a user account. There are three options for migrating passwords: forcing a password change, using a web page, and using SSSD.
Migrating users from an existing system provides a smoother transition but also requires parallel management of LDAP directory and FreeIPA during the migration and transition process. If you do not preserve passwords, the migration can be performed more quickly but it requires more manual work by administrators and users.
20.1.2.1. Method 1: Using Temporary Passwords and Requiring a Change
When passwords are changed in FreeIPA, they will be created with the appropriate Kerberos hashes. So one alternative for administrators is to force users to change their passwords by resetting all user passwords when user accounts are migrated. (This can also be done simply by re-creating the LDAP directory accounts in FreeIPA, which automatically creates accounts with the appropriate keys.) The new users are assigned a temporary password which they change at the first login. No passwords are migrated.
20.1.2.2. Method 2: Using the Migration Web Page
When it is running in migration mode, FreeIPA has a special web page in its web UI that will capture a cleartext password and create the appropriate Kerberos hash.
https://ipaserver.example.com/ipa/migration
Administrators could tell users to authenticate once to this web page, which would properly update their user accounts with their password and corresponding Kerberos hash, without requiring password changes.
20.1.2.3. Method 3: Using SSSD (Recommended)
SSSD can work with FreeIPA to mitigate the user impact on migrating by generating the required user keys. For deployments with a lot of users or where users shouldn't be burdened with password changes, this is the best scenario.
A user tries to log into a machine with SSSD.
SSSD attempts to perform Kerberos authentication against the FreeIPA server.
Even though the user exists in the system, the authentication will fail with the error key type is not supported because the Kerberos hashes do not yet exist.
SSSD the performs a plaintext LDAP bind over a secure connection.
FreeIPA intercepts this bind request. If the user has a Kerberos principal but no Kerberos hashes, then the FreeIPA identity provider generates the hashes and stores them in the user entry.
If authentication is successful, SSSD disconnects from FreeIPA and tries Kerberos authentication again. This time, the request succeeds because the hash exists in the entry.
That entire process is entirely transparent to the user; as far as users known, they simply log into a client service and it works as normal.
20.1.2.4. Migrating Cleartext LDAP Passwords
Although in most deployments LDAP passwords are stored encrypted, there may be some users or some environments that use cleartext passwords for user entries.
When users are migrated from the LDAP server to the FreeIPA server, their cleartext passwords are not migrated over. FreeIPA does not allow cleartext passwords. Instead, a Kerberos principle is created for the user, the keytab is set to true, and the password is set as expired. This means that FreeIPA requires the user to reset the password at the next login.
20.1.2.5. Automatically Resetting Passwords That Do Not Meet Requirements
If user passwords in the original directory do not meet the password policies defined in FreeIPA, then the passwords must be reset after migration.
Password resets are done automatically the first time the users attempts to kinit into the FreeIPA domain.
[jsmith@server ~]$ kinit
Password for jsmith@EXAMPLE.COM:
Password expired. You must change it now.
Enter new password:
Enter it again:
20.1.3. Migration Considerations and Requirements
As you are planning migrating from an LDAP server to FreeIPA, make sure that your LDAP environment is able to work with the FreeIPA migration script.
20.1.3.1. LDAP Servers Supported for Migration
The migration process from an LDAP server to FreeIPA uses a special script, ipa migrate-ds, to perform the migration. This script has certain expectations about the structure of the LDAP directory and LDAP entries in order to work. Migration is supported only for LDAPv3-compliant directory services, which include several common directories:
SunONE Directory Server
Apache Directory Server
OpenLDAP
Migration from an LDAP server to FreeIPA has been tested with 389 Directory Server.
Migration using the migration script is not supported for Microsoft Active Directory because it is not an LDAPv3-compliant directory.
20.1.3.2. Migration Environment Requirements
There are many different possible configuration scenarios for both 389 Directory Server and FreeIPA, and any of those scenarios may affect the migration process. For the example migration procedures in this chapter, these are the assumptions about the environment:
A single LDAP directory domain is being migrated to one FreeIPA realm. No consolidation is involved.
User passwords are stored as a hash in the LDAP directory that the FreeIPA Directory Server can support.
The LDAP directory instance is both the identity store and the authentication method. Client machines are configured to use pam_ldap or nss_ldap to connect to the LDAP server.
Entries use only standard LDAP schema. Custom attributes will not be migrated to FreeIPA.
20.1.3.3. Migration Tools
FreeIPA uses a specific command, ipa migrate-ds, to drive the migration process so that LDAP directory data are properly formatted and imported cleanly into the FreeIPA server.
The FreeIPA server must be configured to run in migration mode, and then the migration script can be used.
20.1.3.4. Migration Sequence
There are four major steps when migrating to FreeIPA, but the order varies slightly depending on whether you want to migrate the server first or the clients first.
With a client-based migration, SSSD is used to change the client configuration while a FreeIPA server is configured:
Deploy SSSD.
Reconfigure clients to connect to the current LDAP server and then fail over to FreeIPA.
Install the FreeIPA server.
Migrate the user data using the FreeIPA ipa-migrate-ds script. This exports the data from the LDAP directory, formats for the FreeIPA schema, and then imports it into FreeIPA.
Take the LDAP server offline and allow clients to fail over to FreeIPA transparently.
With a server migration, the LDAP to FreeIPA migration comes first:
Install the FreeIPA server.
Migrate the user data using the FreeIPA ipa-migrate-ds script. This exports the data from the LDAP directory, formats it for the FreeIPA schema, and then imports it into FreeIPA.
Optional. Deploy SSSD.
Reconfigure clients to connect to FreeIPA. It is not possible to simply replace the LDAP server. The FreeIPA directory tree — and therefore user entry DNs — is different than the previous directory tree.
While it is required that clients be reconfigured, clients do not need to be reconfigured immediately. Updated clients can point to the FreeIPA server while other clients point to the old LDAP directory, allowing a reasonable testing and transition phase after the data are migrated.
Do not run both an LDAP directory service and the FreeIPA server for very long in parallel. This introduces the risk of user data being inconsistent between the two services.
Both processes provide a general migration procedure, but it may not work in every environment. Set up a test LDAP environment and test the migration process before attempting to migrate the real LDAP environment.
20.2. Examples for Using migrate-ds
The data migration is performed with the ipa migrate-ds command. At its simplest, the command takes the LDAP URL of the directory to migrate and exports the data based on common default settings.
ipa migrate-ds ldap://ldap.example.com:389
It is possible to customize how the migrate-ds commands identifies and exports data. This is useful if the original directory tree has a unique structure or if some entries or attributes within entries should be excluded from migration.
20.2.1. Migrating Specific Subtrees
The default directory structure places person entries in the ou=People subtree and group entries in the ou=Groups subtree. These subtrees are container entries for those different types of directory data. If no options are passed with the migrate-ds command, then the utility assumes that the given LDAP directory uses the ou=People and ou=Groups structure.
Many deployments may have an entirely different directory structure (or may only want to export certain parts of the directory tree). There are two options which allow administrators to give the RDN of a different user or group subtree:
--user-container
--group-container
In both cases, the subtree must be the RDN only and must be relative to the base DN. For example, the ou=Employees,dc=example,dc=com subtree can be migrated using --user-container=ou=Employees, but ou=Employees,ou=People,dc=example,dc=com cannot be migrated with that option because ou=Employees is not a direct child of the base DN.
For example:
[root@ipaserver ~]# ipa migrate-ds --user-container=ou=employees --group-container="ou=employee groups" ldap://ldap.example.com:389
There is a third option that allows administrators to set a base DN for migration: --base-dn. With this option, it is possible to change the target for container subtrees. For example:
[root@ipaserver ~]# ipa migrate-ds --user-container=ou=employees --base-dn="ou=people,dc=example,dc=com" ldap://ldap.example.com:389
Now, the ou=Employees user subtree can be migrated from within the larger ou=People subtree without migrating every people-related subtree.
20.2.2. Specifically Including or Excluding Entries
By default, the migrate-ds script exports every user entry with the person object class and every group entry within the given user and group subtrees.
In some migration paths, only specific types of users and groups may need to be exported, or, conversely, specific users and groups may need to be excluded.
On option is to set positively which types of users and groups to include. This is done by setting which object classes to search for when looking for user or group entries.
This is a really useful option when there are custom object classes used in an environment for different user types. For example, this migrates only users with the custom fullTimeEmployee object class:
[root@ipaserver ~]# ipa migrate-ds --user-objectclass=fullTimeEmployee ldap://ldap.example.com:389
Because of the different types of groups, this is also very useful for migrating only certain types of groups (such as user groups) while excluding other types of groups, like certificate groups. For example:
[root@ipaserver ~]# ipa migrate-ds --group-objectclass=groupOfNames,groupOfUniqueNames ldap://ldap.example.com:389
Positively specifying user and groups to migrate based on object class implicitly excludes all other users and groups from migration.
Alternatively, it can be useful to migrate all user and group entries except for just a small handful of entries. Specific user or group accounts can be excluded while all others of that type are migrated. For example, this excludes a hobbies group and two users:
[root@ipaserver ~]# ipa migrate-ds --exclude-groups="Golfers Group" --exclude-users=jsmith,bjensen ldap://ldap.example.com:389
Specifying an object class to migrate can be used together with excluding specific entries. For example, this specifically includes users with the fullTimeEmployee object class, yet excludes three managers:
[root@ipaserver ~]# ipa migrate-ds --user-objectclass=fullTimeEmployee --exclude-users=jsmith,bjensen,mreynolds ldap://ldap.example.com:389
20.2.3. Excluding Entry Attributes
By default, every attribute and object class for a user or group entry is migrated. There are some cases where that may not be realistic, either because of bandwidth and network constraints or because the attribute data are no longer relevant. For example, if users are going to be assigned new user certificates as they join the FreeIPA domain, then there is no reason to migrate the userCertificate attribute.
Specific object classes and attributes can be ignored by the migrate-ds by using any of several different options:
For example, to exclude the userCertificate attribute and strongAuthenticationUser object class for users and the groupOfCertificates object class for groups:
[root@ipaserver ~]# ipa migrate-ds --user-ignore-attribute=userCertificate --user-ignore-objectclass=strongAuthenticationUser --group-ignore-objectclass=groupOfCertificates ldap://ldap.example.com:389
Make sure not to ignore any required attributes. Also, when excluding object classes, make sure to exclude any attributes which are only supported by that object class.
20.2.4. Setting the Schema to Use
By default, FreeIPA uses
RFC2307bis schema to define user, host, hostgroup, and other network identities. This schema option can be reset to use
RFC2307 schema instead:
[root@ipaserver ~]# ipa migrate-ds --schema=RFC2307 ldap://ldap.example.com:389
20.3. Scenario 1: Using SSSD as Part of Migration
This is a general migration procedure, but it may not work in every environment.
It is strongly recommended that you set up a test LDAP environment and test the migration process before attempting to migrate the real LDAP environment.
Set up SSSD. Using SSSD allows the required Kerberos keys and server certificates to be delivered to the clients.
Install SSSD on every client machine:
# yum install sssd
Configure an LDAP identity provider in SSSD to use the existing Directory Server for all functions (authentication, identity lookups, access, and password changes). This ensures every client works properly with the existing directory service.
Install FreeIPA, including any custom LDAP directory schema
, on a different machine from the existing LDAP directory.
Enable the FreeIPA server to allow migration:
# ipa config-mod --enable-migration=TRUE
Disable the compat plug-in.
# ipa-compat-manage disable
Restart the FreeIPA Directory Server instance.
# service dirsrv restart
Run the FreeIPA migration script, ipa migrate-ds. At its most basic, this requires only the LDAP URL of the LDAP directory instance to migrate:
# ipa migrate-ds ldap://ldap.example.com:389
Simply passing the LDAP URL migrates all of the directory data using common default settings. The user and group data can be selectively migrated by specifying other options, as covered in
Section 20.2, “Examples for Using migrate-ds”.
Once the information is exported, the script adds all required FreeIPA object classes and attributes and converts DNs in attributes to match the FreeIPA directory tree.
Re-enable the compat plug-in.
# ipa-compat-manage enable
Restart the FreeIPA Directory Server instance.
# service dirsrv restart
Move clients that have SSSD installed from the LDAP backend to the FreeIPA backend and enroll them as client with FreeIPA. This downloads the required keys and certificates.
On Fedora clients, this can be done using the ipa-client-install command. For example:
Have users log into a machine with SSSD and FreeIPA backend. This generates the required Kerberos keys for the user.
To monitor the user migration process, query the existing LDAP directory to see which user accounts have a password but do not yet have a Kerberos principal key.
$ ldapsearch -LL -x -D 'cn=Directory Manager' -w secret -b 'ou=people,dc=example,dc=com' '(&(!(krbprincipalkey=*))(userpassword=*))' uid
Include the quotes around the filter so that it is not interpreted by the shell.
Once users have been migrated over, configure non-SSSD clients to use the FreeIPA domain, as required.
When the migration of all clients and users is complete, decommission the LDAP directory.
20.4. Scenario 2: Migrating an LDAP Server Directly to FreeIPA
This is a general migration procedure, but it may not work in every environment.
It is strongly recommended that you set up a test LDAP environment and test the migration process before attempting to migrate the real LDAP environment.
Install the FreeIPA server, including any custom LDAP directory schema
, on a different machine from the existing LDAP directory.
Disable the compat plug-in.
# ipa-compat-manage disable
Restart the FreeIPA Directory Server instance.
# service dirsrv restart
Enable the FreeIPA server to allow migration:
# ipa config-mod --enable-migration=TRUE
Run the FreeIPA migration script, ipa migrate-ds. At its most basic, this requires only the LDAP URL of the LDAP directory instance to migrate:
# ipa migrate-ds ldap://ldap.example.com:389
Simply passing the LDAP URL migrates all of the directory data using common default settings. The user and group data can be selectively migrated by specifying other options, as covered in
Section 20.2, “Examples for Using migrate-ds”.
Once the information is exported, the script adds all required FreeIPA object classes and attributes and converts DNs in attributes to match the FreeIPA directory tree.
Re-enable the compat plug-in.
# ipa-compat-manage enable
Restart the FreeIPA Directory Server instance.
# service dirsrv restart
Update the client configuration to use PAM_LDAP and NSS_LDAP to connect to FreeIPA instead of connecting to an LDAP directory, NIS, or local files.
Install SSSD on every client machine:
# yum install sssd
Run the ipa-client-install to configure SSSD and related services to use the FreeIPA server for identity and Kerberos authentication.
Instruct users to log into FreeIPA using either SSSD client or the migration web page if SSSD is not available on the client. Both methods automatically migrate the user password into FreeIPA.
https://ipaserver.example.com/ipa/migration
Optional. Reconfigure non-SSSD clients to use Kerberos authentication (pam_krb5) instead of LDAP authentication (pam_ldap).
Use PAM_LDAP modules until all of the users have been migrated; then it is possible to use PAM_KRB5.
When the migration of all clients and users is complete, decommission the LDAP directory.
Frequently Asked Questions
- Q:
Is it possible to change the IP address of the master server?
- Q:
Why are there restrictions on the length of user and group names? How can I change this?
- Q:
What is the difference between a replica and a master server?
- Q:
Why does the ipa-client-install script fail to find the IPA server on a network that uses Active Directory DNS?
- Q:
Can an administrator who is connected to "Server B" revoke a certificate issued by "Server A"?
Q:
Is it possible to change the IP address of the master server?
A:
Yes. If you are only changing the IP address, it is sufficient to update the /etc/hosts file, the system configuration, and the DNS entry.
Q:
Why are there restrictions on the length of user and group names? How can I change this?
A:
User and group name lengths are specified in the policy. The default maximum username length is 32 characters. The maximum configurable length for user or group names is 255 characters. This complements some supported client operating systems which limit the length of usernames.
The default settings can be changed in the FreeIPA UI or using the ipa config-mod command. For example:
4 ipa config-mod --maxusername=50
Q:
What is the difference between a replica and a master server?
A:
A master server maintains a certificate authority. A replica server has its certificate issued by the master CA.
Q:
Why does the ipa-client-install script fail to find the IPA server on a network that uses Active Directory DNS?
A:
Active Directory has its own SRV records for Kerberos and LDAP. The ipa-client-install script can retrieve those records instead of any that have been added for the FreeIPA domain.
When running ipa-client-install, manually enter the server information to ensure that the script uses the FreeIPA SRV records instead of Active Directory records. The ipa-client-install options are listed in the ipa-client-install manpage.
Q:
Can an administrator who is connected to "Server B" revoke a certificate issued by "Server A"?
A:
Yes, assuming that Servers A and B contain non-cloned CAs which have their database information replicated to share revocation information only.
Part of managing machine authentication is managing machine certificates. On clients, FreeIPA manages the certificate lifecycle with the certmonger service, which works together with the certificate authority (CA) provided by FreeIPA.
The certmonger daemon and its command-line clients simplify the process of generating public/private key pairs, creating certificate requests, and submitting requests to the CA for signing. As part of managing certificates, the certmonger daemon monitors certificates for expiration and can renew certificates that are about to expire. The certificates that certmonger monitors are tracked in files stored in a configurable directory. The default location is /var/lib/certmonger/requests.
certmonger uses the FreeIPA
getcert command to manage all certificates. As covered in
Section 2.3.3.2, “Using Different CA Configurations”, an FreeIPA server can be configured to use different types of certificate authorities. The most common (and recommended) configuration is to use a full CA server, but it is also possible to use a much more limited, self-signed CA. The exact
getcert command used by
certmonger to communicate with the FreeIPA backend depends on which type of CA is used. The
ipa-getcert command is used with a full CA, while the
selfsign-getcert command is used with a self-signed CA.
Because of general security issues, self-signed certificates are not typically used in production, but can be used for development and testing.
B.1. Requesting a Certificate with certmonger
With the FreeIPA CA, certmonger uses the ipa-getcert command.
Certificates and keys are stored locally in plaintext files (.pem) or in an NSS database, identified by the certificate nickname. When requesting a certificate, then, the request should identify the location where the certificate will be stored and the nickname of the certificate. For example:
# ipa-getcert request -d /etc/pki/nssdb -n Server-Cert
The /etc/pki/nssdb file is the global NSS database, and Server-Cert is the nickname of this certificate. The certificate nickname must be unique within this database.
When requesting a certificate to be used with a FreeIPA service, the -K option is required to specify the service principal. Otherwise, certmonger assumes the certificate is for a host. The -N option must specify the certificate subject DN, and the subject base DN must match the base DN for the FreeIPA server, or the request is rejected.
$ ipa-getcert request -d /etc/httpd/alias -n Server-Cert -K HTTP/client1.example.com -N 'CN=client1.example.com,O=EXAMPLE.COM'
Example B.1. Using certmonger for a Service
$ ipa-getcert request -r -f /etc/httpd/conf/ssl.crt/server.crt -k /etc/httpd/conf/ssl.key/server.key -N CN=`hostname --fqdn` -D `hostname` -U id-kp-serverAuth
The options vary depending on whether you are using a self-signed certificate (
selfsign-getcert) and the desired configuration for the final certificate, as well as other settings. In
Example B.1, “Using certmonger for a Service”, these are common options:
The -r option will automatically renew the certificate if the key pair already exists. This is used by default.
The -f option stores the certificate in the given file.
The -k option either stores the key in the given file or, if the key file already exists, uses the key in the file.
The -N option gives the subject name.
The -D option gives the DNS domain name.
The -U option sets the extended key usage flag.
B.2. Storing Certificates in NSS Databases
By default, certmonger uses plaintext files to store the key and the certificate, but these keys and certificates can also be stored in NSS databases. This is done using the -d option to set the security database location and -n to give the certificate nickname which is used for the certificate in the database. These options are used instead of the PEM files given in the -f and -k options.
For example:
# ipa-getcert request -d /export/alias -n ServerCert ...
B.3. Tracking Certificates with certmonger
certmonger can manage the entire certificate lifecycle. Along with generating requests, certmonger can track a certificate and automatically renew it when it expires at the end of its validity period.
This is done using the start-tracking command with the getcert command. The -I option creates the tracking entry, along with pointers to the key and certificate files, either in an NSS database (-d and -n) or in the PEM file (-f and -k). The -r option tells certmonger to renew the certificate.
# ipa-getcert start-tracking -I cert1-tracker -d /export/alias -n ServerCert -r
The
-r option can be passed with the
request command, in
Example B.1, “Using certmonger for a Service”. In that case, the requested certificate is automatically tracked and renewed by
certmonger. Then, it is not necessary to configure tracking manually.
A certificate can be untracked by certmonger by using the stop-tracking command.
A
- access control instruction
See ACI.
- access control list
See ACL.
- access rights
In the context of access control, specify the level of access granted or denied. Access rights are related to the type of operation that can be performed on the directory. The following rights can be granted or denied: read, write, add, delete, search, compare, selfwrite, proxy and all.
- account inactivation
Disables a user account, group of accounts, or an entire domain so that all authentication attempts are automatically rejected.
- ACI
An instruction that grants or denies permissions to entries in the directory.
See Also access control instruction.
- ACL
The mechanism for controlling access to your directory.
See Also access control list.
- All IDs Threshold
Replaced with the ID list scan limit in Directory Server version 7.1. A size limit which is globally applied to every index key managed by the server. When the size of an individual ID list reaches this limit, the server replaces that ID list with an All IDs token.
See Also ID list scan limit.
- All IDs token
A mechanism which causes the server to assume that all directory entries match the index key. In effect, the All IDs token causes the server to behave as if no index was available for the search request.
- anonymous access
When granted, allows anyone to access directory information without providing credentials, and regardless of the conditions of the bind.
- approximate index
Allows for efficient approximate or "sounds-like" searches.
- attribute
Holds descriptive information about an entry. Attributes have a label and a value. Each attribute also follows a standard syntax for the type of information that can be stored as the attribute value.
- attribute list
A list of required and optional attributes for a given entry type or object class.
- authenticating directory server
In pass-through authentication (PTA), the authenticating Directory Server is the Directory Server that contains the authentication credentials of the requesting client. The PTA-enabled host sends PTA requests it receives from clients to the host.
- authentication
(1) Process of proving the identity of the client user to the Directory Server. Users must provide a bind DN and either the corresponding password or certificate in order to be granted access to the directory. Directory Server allows the user to perform functions or access files and directories based on the permissions granted to that user by the directory administrator.
(2) Allows a
client to make sure they are connected to a secure server, preventing another computer from impersonating the server or attempting to appear secure when it is not.
- authentication certificate
Digital file that is not transferable and not forgeable and is issued by a third party. Authentication certificates are sent from server to client or client to server in order to verify and authenticate the other party.
B
- base distinguished name
See base DN.
- base DN
Base distinguished name. A search operation is performed on the base DN, the DN of the entry and all entries below it in the directory tree.
- bind distinguished name
See bind DN.
- bind DN
Distinguished name used to authenticate to Directory Server when performing an operation.
- bind rule
In the context of access control, the bind rule specifies the credentials and conditions that a particular user or client must satisfy in order to get access to directory information.
- branch entry
An entry that represents the top of a subtree in the directory.
- browser
Software, such as Mozilla Firefox, used to request and view World Wide Web material stored as HTML files. The browser uses the HTTP protocol to communicate with the host server.
- browsing index
Speeds up the display of entries in the Directory Server Console. Browsing indexes can be created on any branch point in the directory tree to improve display performance.
See Also virtual list view index .
C
- CA
See Certificate Authority.
- cascading replication
In a cascading replication scenario, one server, often called the hub supplier, acts both as a consumer and a supplier for a particular replica. It holds a read-only replica and maintains a changelog. It receives updates from the supplier server that holds the master copy of the data and in turn supplies those updates to the consumer.
- certificate
A collection of data that associates the public keys of a network user with their DN in the directory. The certificate is stored in the directory as user object attributes.
- Certificate Authority
Company or organization that sells and issues authentication certificates. You may purchase an authentication certificate from a Certification Authority that you trust. Also known as a
CA.
- CGI
Common Gateway Interface. An interface for external programs to communicate with the HTTP server. Programs written to use CGI are called CGI programs or CGI scripts and can be written in many of the common programming languages. CGI programs handle forms or perform output parsing that is not done by the server itself.
- chaining
A method for relaying requests to another server. Results for the request are collected, compiled, and then returned to the client.
- changelog
A changelog is a record that describes the modifications that have occurred on a replica. The supplier server then replays these modifications on the replicas stored on replica servers or on other masters, in the case of multi-master replication.
- character type
Distinguishes alphabetic characters from numeric or other characters and the mapping of upper-case to lower-case letters.
- ciphertext
Encrypted information that cannot be read by anyone without the proper key to decrypt the information.
- class definition
Specifies the information needed to create an instance of a particular object and determines how the object works in relation to other objects in the directory.
- class of service
See CoS.
- classic CoS
A classic CoS identifies the template entry by both its DN and the value of one of the target entry's attributes.
- client
See LDAP client.
- code page
An internal table used by a locale in the context of the internationalization plug-in that the operating system uses to relate keyboard keys to character font screen displays.
- collation order
Provides language and cultural-specific information about how the characters of a given language are to be sorted. This information might include the sequence of letters in the alphabet or how to compare letters with accents to letters without accents.
- consumer
Server containing replicated directory trees or subtrees from a supplier server.
- consumer server
In the context of replication, a server that holds a replica that is copied from a different server is called a consumer for that replica.
- CoS
A method for sharing attributes between entries in a way that is invisible to applications.
- CoS definition entry
Identifies the type of CoS you are using. It is stored as an LDAP subentry below the branch it affects.
- CoS template entry
Contains a list of the shared attribute values.
See Also template entry.
D
- daemon
A background process on a Unix machine that is responsible for a particular system task. Daemon processes do not need human intervention to continue functioning.
- DAP
Directory Access Protocol. The ISO X.500 standard protocol that provides client access to the directory.
- data master
The server that is the master source of a particular piece of data.
- database link
An implementation of chaining. The database link behaves like a database but has no persistent storage. Instead, it points to data stored remotely.
- default index
One of a set of default indexes created per database instance. Default indexes can be modified, although care should be taken before removing them, as certain plug-ins may depend on them.
- definition entry
See CoS definition entry.
- Directory Access Protocol
See DAP.
- Directory Manager
The privileged database administrator, comparable to the root user in UNIX. Access control does not apply to the Directory Manager.
- directory service
A database application designed to manage descriptive, attribute-based information about people and resources within an organization.
- directory tree
The logical representation of the information stored in the directory. It mirrors the tree model used by most filesystems, with the tree's root point appearing at the top of the hierarchy. Also known as
DIT.
- distinguished name
String representation of an entry's name and location in an LDAP directory.
- DIT
See directory tree.
- DM
See Directory Manager.
- DN
See distinguished name.
- DNS
Domain Name System. The system used by machines on a network to associate standard IP addresses (such as 198.93.93.10) with hostnames (such as www.example.com). Machines normally get the IP address for a hostname from a DNS server, or they look it up in tables maintained on their systems.
- DNS alias
A DNS alias is a hostname that the DNS server knows points to a different hostspecifically a DNS CNAME record. Machines always have one real name, but they can have one or more aliases. For example, an alias such as www.yourdomain.domain might point to a real machine called realthing.yourdomain.domain where the server currently exists.
E
- entry
A group of lines in the LDIF file that contains information about an object.
- entry distribution
Method of distributing directory entries across more than one server in order to scale to support large numbers of entries.
- entry ID list
Each index that the directory uses is composed of a table of index keys and matching entry ID lists. The entry ID list is used by the directory to build a list of candidate entries that may match the client application's search request.
- equality index
Allows you to search efficiently for entries containing a specific attribute value.
F
- file extension
The section of a filename after the period or dot (.) that typically defines the type of file (for example, .GIF and .HTML). In the filename index.html the file extension is html.
- file type
The format of a given file. For example, graphics files are often saved in GIF format, while a text file is usually saved as ASCII text format. File types are usually identified by the file extension (for example, .GIF or .HTML).
- filter
A constraint applied to a directory query that restricts the information returned.
- filtered role
Allows you to assign entries to the role depending upon the attribute contained by each entry. You do this by specifying an LDAP filter. Entries that match the filter are said to possess the role.
G
- general access
When granted, indicates that all authenticated users can access directory information.
- GSS-API
Generic Security Services. The generic access protocol that is the native way for UNIX-based systems to access and authenticate Kerberos services; also supports session encryption.
H
- hostname
A name for a machine in the form machine.domain.dom, which is translated into an IP address. For example, www.example.com is the machine www in the subdomain example and com domain.
- HTML
Hypertext Markup Language. The formatting language used for documents on the World Wide Web. HTML files are plain text files with formatting codes that tell browsers such as the Mozilla Firefox how to display text, position graphics, and form items and to display links to other pages.
- HTTP
Hypertext Transfer Protocol. The method for exchanging information between HTTP servers and clients.
- HTTPD
An abbreviation for the HTTP daemon or service, a program that serves information using the HTTP protocol. The daemon or service is often called an httpd.
- HTTPS
A secure version of HTTP, implemented using the Secure Sockets Layer, SSL.
- hub
In the context of replication, a server that holds a replica that is copied from a different server, and, in turn, replicates it to a third server.
See Also cascading replication.
I
- ID list scan limit
A size limit which is globally applied to any indexed search operation. When the size of an individual ID list reaches this limit, the server replaces that ID list with an all IDs token.
- index key
Each index that the directory uses is composed of a table of index keys and matching entry ID lists.
- indirect CoS
An indirect CoS identifies the template entry using the value of one of the target entry's attributes.
- international index
Speeds up searches for information in international directories.
- International Standards Organization
See ISO.
- IP address
Also Internet Protocol address. A set of numbers, separated by dots, that specifies the actual location of a machine on the Internet (for example, 198.93.93.10). Directory Server supports both IPv4 and IPv6 IP addresses.
- ISO
International Standards Organization.
K
- knowledge reference
Pointers to directory information stored in different databases.
L
- LDAP
Lightweight Directory Access Protocol. Directory service protocol designed to run over TCP/IP and across multiple platforms.
- LDAP client
Software used to request and view LDAP entries from an LDAP Directory Server.
See Also browser.
- LDAP Data Interchange Format
See LDAP Data Interchange Format.
- LDAP URL
Provides the means of locating Directory Servers using DNS and then completing the query via LDAP. A sample LDAP URL is ldap://ldap.example.com.
- LDAPv3
Version 3 of the LDAP protocol, upon which Directory Server bases its schema format.
- LDBM database
A high-performance, disk-based database consisting of a set of large files that contain all of the data assigned to it. The primary data store in Directory Server.
- LDIF
LDAP Data Interchange Format. Format used to represent Directory Server entries in text form.
- leaf entry
An entry under which there are no other entries. A leaf entry cannot be a branch point in a directory tree.
- Lightweight Directory Access Protocol
See LDAP.
- locale
Identifies the collation order, character type, monetary format and time / date format used to present data for users of a specific region, culture, and/or custom. This includes information on how data of a given language is interpreted, stored, or collated. The locale also indicates which code page should be used to represent a given language.
M
- managed object
A standard value which the SNMP agent can access and send to the NMS. Each managed object is identified with an official name and a numeric identifier expressed in dot-notation.
- managed role
Allows creation of an explicit enumerated list of members.
- management information base
See MIB.
- mapping tree
A data structure that associates the names of suffixes (subtrees) with databases.
- master
See supplier.
- master agent
See SNMP master agent.
- matching rule
Provides guidelines for how the server compares strings during a search operation. In an international search, the matching rule tells the server what collation order and operator to use.
- MD5
A message digest algorithm by RSA Data Security, Inc., which can be used to produce a short digest of data that is unique with high probability and is mathematically extremely hard to produce; a piece of data that will produce the same message digest.
- MD5 signature
A message digest produced by the MD5 algorithm.
- MIB
Management Information Base. All data, or any portion thereof, associated with the SNMP network. We can think of the MIB as a database which contains the definitions of all SNMP managed objects. The MIB has a tree-like hierarchy, where the top level contains the most general information about the network and lower levels deal with specific, separate network areas.
- MIB namespace
Management Information Base namespace. The means for directory data to be named and referenced. Also called the
directory tree.
- monetary format
Specifies the monetary symbol used by specific region, whether the symbol goes before or after its value, and how monetary units are represented.
- multi-master replication
An advanced replication scenario in which two servers each hold a copy of the same read-write replica. Each server maintains a changelog for the replica. Modifications made on one server are automatically replicated to the other server. In case of conflict, a time stamp is used to determine which server holds the most recent version.
- multiplexor
The server containing the database link that communicates with the remote server.
N
- n + 1 directory problem
The problem of managing multiple instances of the same information in different directories, resulting in increased hardware and personnel costs.
- name collisions
Multiple entries with the same distinguished name.
- nested role
Allows the creation of roles that contain other roles.
- network management application
Network Management Station component that graphically displays information about SNMP managed devices, such as which device is up or down and which and how many error messages were received.
- network management station
See NMS.
- NIS
Network Information Service. A system of programs and data files that Unix machines use to collect, collate, and share specific information about machines, users, filesystems, and network parameters throughout a network of computers.
- NMS
- ns-slapd
Red Hat's LDAP Directory Server daemon or service that is responsible for all actions of the Directory Server.
See Also slapd.
O
- object class
Defines an entry type in the directory by defining which attributes are contained in the entry.
- object identifier
A string, usually of decimal numbers, that uniquely identifies a schema element, such as an object class or an attribute, in an object-oriented system. Object identifiers are assigned by ANSI, IETF or similar organizations.
See Also OID.
- OID
See object identifier.
- operational attribute
Contains information used internally by the directory to keep track of modifications and subtree properties. Operational attributes are not returned in response to a search unless explicitly requested.
P
- parent access
When granted, indicates that users have access to entries below their own in the directory tree if the bind DN is the parent of the targeted entry.
- pass-through authentication
See PTA.
- pass-through subtree
- password file
A file on Unix machines that stores Unix user login names, passwords, and user ID numbers. It is also known as /etc/passwd because of where it is kept.
- password policy
A set of rules that governs how passwords are used in a given directory.
- PDU
Encoded messages which form the basis of data exchanges between SNMP devices. Also
protocol data unit.
- permission
In the context of access control, permission states whether access to the directory information is granted or denied and the level of access that is granted or denied.
See Also access rights.
- pointer CoS
A pointer CoS identifies the template entry using the template DN only.
- presence index
Allows searches for entries that contain a specific indexed attribute.
- protocol
A set of rules that describes how devices on a network exchange information.
- protocol data unit
See PDU.
- proxy authentication
A special form of authentication where the user requesting access to the directory does not bind with its own DN but with a proxy DN.
- proxy DN
Used with proxied authorization. The proxy DN is the DN of an entry that has access permissions to the target on which the client-application is attempting to perform an operation.
- PTA
- PTA directory server
In pass-through authentication (
PTA), the PTA Directory Server is the server that sends (passes through) bind requests it receives to the
authenticating directory server.
- PTA LDAP URL
R
- RAM
Random access memory. The physical semiconductor-based memory in a computer. Information stored in RAM is lost when the computer is shut down.
- rc.local
A file on Unix machines that describes programs that are run when the machine starts. It is also called /etc/rc.local because of its location.
- RDN
The name of the actual entry itself, before the entry's ancestors have been appended to the string to form the full distinguished name. Also
relative distinguished name.
- read-only replica
A replica that refers all update operations to read-write replicas. A server can hold any number of read-only replicas.
- read-write replica
A replica that contains a master copy of directory information and can be updated. A server can hold any number of read-write replicas.
- referential integrity
Mechanism that ensures that relationships between related entries are maintained within the directory.
- referral
(1) When a server receives a search or update request from an LDAP client that it cannot process, it usually sends back to the client a pointer to the LDAP sever that can process the request.
(2) In the context of replication, when a read-only replica receives an update request, it forwards it to the server that holds the corresponding read-write replica. This forwarding process is called a referral.
- relative distinguished name
See RDN.
- replica
A database that participates in replication.
- replica-initiated replication
Replication configuration where replica servers, either hub or consumer servers, pull directory data from supplier servers. This method is available only for legacy replication.
- replication
Act of copying directory trees or subtrees from supplier servers to replica servers.
- replication agreement
Set of configuration parameters that are stored on the supplier server and identify the databases to replicate, the replica servers to which the data is pushed, the times during which replication can occur, the DN and credentials used by the supplier to bind to the consumer, and how the connection is secured.
- RFC
Request for Comments. Procedures or standards documents submitted to the Internet community. People can send comments on the technologies before they become accepted standards.
- role
An entry grouping mechanism. Each role has members, which are the entries that possess the role.
- role-based attributes
Attributes that appear on an entry because it possesses a particular role within an associated CoS template.
- root
The most privileged user available on Unix machines. The root user has complete access privileges to all files on the machine.
- root suffix
The parent of one or more sub suffixes. A directory tree can contain more than one root suffix.
S
- SASL
- schema
Definitions describing what types of information can be stored as entries in the directory. When information that does not match the schema is stored in the directory, clients attempting to access the directory may be unable to display the proper results.
- schema checking
Ensures that entries added or modified in the directory conform to the defined schema. Schema checking is on by default, and users will receive an error if they try to save an entry that does not conform to the schema.
- Secure Sockets Layer
See SSL.
- self access
When granted, indicates that users have access to their own entries if the bind DN matches the targeted entry.
- Server Console
Java-based application that allows you to perform administrative management of your Directory Server from a GUI.
- server daemon
The server daemon is a process that, once running, listens for and accepts requests from clients.
- Server Selector
Interface that allows you select and configure servers using a browser.
- server service
A process on Windows that, once running, listens for and accepts requests from clients. It is the SMB server on Windows NT.
- service
A background process on a Windows machine that is responsible for a particular system task. Service processes do not need human intervention to continue functioning.
- SIE
Server Instance Entry. The ID assigned to an instance of Directory Server during installation.
- Simple Authentication and Security Layer
See SASL.
- Simple Network Management Protocol
See SNMP.
- single-master replication
The most basic replication scenario in which multiple servers, up to four, each hold a copy of the same read-write replicas to replica servers. In a single-master replication scenario, the supplier server maintains a changelog.
- SIR
See supplier-initiated replication.
- slapd
LDAP Directory Server daemon or service that is responsible for most functions of a directory except replication.
See Also ns-slapd.
- SNMP
- SNMP master agent
Software that exchanges information between the various subagents and the NMS.
- SNMP subagent
Software that gathers information about the managed device and passes the information to the master agent. Also called a
subagent.
- SOA
See start of authority (SOA).
- SSL
A software library establishing a secure connection between two parties (client and server) used to implement HTTPS, the secure version of HTTP. Also called
Secure Sockets Layer.
- standard index
An index maintained by default.
- start of authority (SOA)
A record which contains the core information about a DNS zone.
- sub suffix
A branch underneath a root suffix.
- subagent
See SNMP subagent.
- substring index
Allows for efficient searching against substrings within entries. Substring indexes are limited to a minimum of two characters for each entry.
- suffix
The name of the entry at the top of the directory tree, below which data is stored. Multiple suffixes are possible within the same directory. Each database only has one suffix.
- superuser
The most privileged user available on Unix machines. The superuser has complete access privileges to all files on the machine. Also called
root.
- supplier
Server containing the master copy of directory trees or subtrees that are replicated to replica servers.
- supplier server
In the context of replication, a server that holds a replica that is copied to a different server is called a supplier for that replica.
- supplier-initiated replication
Replication configuration where
supplier servers replicate directory data to any replica servers.
- symmetric encryption
Encryption that uses the same key for both encrypting and decrypting. DES is an example of a symmetric encryption algorithm.
- system index
Cannot be deleted or modified as it is essential to Directory Server operations.
T
- target
In the context of access control, the target identifies the directory information to which a particular ACI applies.
- target entry
The entries within the scope of a CoS.
- TCP/IP
Transmission Control Protocol/Internet Protocol. The main network protocol for the Internet and for enterprise (company) networks.
- template entry
See CoS template entry.
- time/date format
Indicates the customary formatting for times and dates in a specific region.
- TLS
- topology
The way a directory tree is divided among physical servers and how these servers link with one another.
- Transport Layer Security
See TLS.
U
- uid
A unique number associated with each user on a Unix system.
- URL
Uniform Resource Locater. The addressing system used by the server and the client to request documents. It is often called a location. The format of a URL is protocol://machine:port/document. The port number is necessary only on selected servers, and it is often assigned by the server, freeing the user of having to place it in the URL.
V
- virtual list view index
Speeds up the display of entries in the Directory Server Console. Virtual list view indexes can be created on any branch point in the directory tree to improve display performance.
See Also browsing index.
X
- X.500 standard
The set of ISO/ITU-T documents outlining the recommended information model, object classes and attributes used by directory server implementation.
D
- DHCP, Adding Host Entries from the Command Line
- DNS
- adding zone records, Adding Records to DNS Zones
- adding zones, Adding DNS Zones
- bind-dyndb-ldap and Directory Server, About DNS in FreeIPA
- disabling zones, Enabling and Disabling Zones
- dynamic updates, Enabling Dynamic DNS Updates
- hosts with DHCP, Adding Host Entries from the Command Line
- DNS zone records, Adding Records to DNS Zones
- deleting, Deleting Records from DNS Zones
- format for adding, About the Commands to Add DNS Records
- IPv4 example, Examples of Adding DNS Resource Records
- IPv6 example, Examples of Adding DNS Resource Records
- PTR example, Examples of Adding DNS Resource Records
- SRV example, Examples of Adding DNS Resource Records
- types of records, Adding Records to DNS Zones
direct-membership.png)
