Anaconda configuration files

The installer loads its default configuration from the Anaconda configuration files. The configuration can be modified by kernel arguments and cmdline options and the result is written into a runtime configuration file. The configuration is not supposed to change after that. The runtime configuration file is loaded by the Anaconda DBus modules when they are started. It means that all Anaconda processes are running with the same configuration.

Note

The software selected for the installation doesn’t change the Anaconda configuration and profiles in any way.

Note

The interactive-defaults.ks file will be replaced by the Anaconda configuration files in the future. Kickstart files should be used only for the automatic installation.

File structure

The Anaconda configuration files are written in the INI format that can be processed by configparser. The files consist of sections, options and comments. Each section is defined by a [section] header. Each option is defined by a key and optionally a value separated by the = sign. Each comment has to start on a new line prefixed by the # character.

See an example of a section:

[Storage Constraints]

# Minimal size of the total memory.
min_ram = 320 MiB

# Should we recommend to specify a swap partition?
swap_is_recommended = False

# Recommended minimal sizes of partitions.
# Specify a mount point and a size on each line.
min_partition_sizes =
    /      250 MiB
    /usr   250 MiB

# Required minimal sizes of partitions.
# Specify a mount point and a size on each line.
req_partition_sizes =

The supported sections and options are documented in the default configuration file.

Default configuration file

The default configuration file provides a full default configuration of the installer. It defines and documents all supported sections and options. The file is located at /etc/anaconda/anaconda.conf:

Profile configuration files

The profile configuration files allow to override some of the configuration options for specific profiles and products. The files are located at /etc/anaconda/profile.d/.

Note

Anaconda previously used so called install classes for the product-specific configuration. Install classes were completely removed and replaced by the profile configuration files. These configuration files used to be called product configuration files for some time.

Profile identification

Each profile has a unique profile id. It is a lower-case string with no spaces that identifies the profile. The id can be arbitrary, but the convention is to use the name of the configuration file (for example, fedora-server).

Profile detection

The profile can be specified by the inst.profile boot option or the --profile cmdline option. Based on the provided profile id, the installer will look up the right configuration file in the /etc/anaconda/profile.d/ directory.

Otherwise, the profile will be chosen based on the os-release values of the installation environment. These values are provided by the /etc/os-release or /usr/lib/os-release file containing operating system identification data. The profile can define os and variant ids that should match ID and VARIANT_ID options of the os-release files. The installer will use a profile with the best match.

File structure

Profile configuration files have one or two extra sections that describe the profile.

The [Profile] section defines a profile id of the profile. Optionally, it can specify a profile id of a base profile. For example, fedora is a base profile of fedora-server.

We support a simple inheritance of profile configurations. The installer loads configuration files of the base profiles before it loads the configuration file of the specified profile. For example, it will first load the configuration for fedora and then the configuration for fedora-server.

Note

We are not going to support multiple inheritance. It would significantly increase the complexity of the profile configuration files in an unintuitive way. You can easily compare two configuration files and verify the parts they are supposed to share. We do that in our unit tests.

The [Profile Detection] defines the operating system id and the variant id that should match os-release values of the expected installation environment. It is useful for assigning the profile to a specific product (for example, Fedora Server). This section is optional.

Note

We are not going to support wildcards in the profile detection. This used to be supported in install classes and it caused a lot of problems. Without the wildcards, we will always match at most one profile.

See an example of the profile configuration file for Fedora Server:

# Anaconda configuration file for Fedora Server.

[Profile]
# Define the profile.
profile_id = fedora-server
base_profile = fedora

[Profile Detection]
# Match os-release values.
os_id = fedora
variant_id = server

[Payload]
# Change payload-related options.
default_environment = server-product-environment

[Storage]
# Change storage-related options.
file_system_type = xfs
default_scheme = LVM

Custom configuration files

The custom configuration files allow to override some of the configuration options for specific installations. The files are located at /etc/anaconda/conf.d/.

The installer finds all files with the .conf extension in the /etc/anaconda/conf.d/ directory, sorts them by their name and loads them in this order. These files are loaded after the profile configuration files, so they have a higher priority.

For example, the initial setup installs the 10-initial-setup.conf file with a custom configuration.

Note

All configuration files have to be loaded before the installer starts to parse the kickstart file, so it is not possible to generate a configuration file in the %pre section of the kickstart file. Please, use updates.img or product.img instead.

Runtime configuration file

The runtime configuration file is a temporary file that provides a full configuration of the current installer run. It is generated by the installer and it exists only during its lifetime. The file is located at /run/anaconda/anaconda.conf.

The runtime configuration file is loaded by the Anaconda DBus modules when they are started. It allows us to run all Anaconda processes with the same configuration.

The installer makes the following steps to create the runtime configuration file. The configuration is not supposed to change after that.

Load the default configuration file from /etc/anaconda/anaconda.conf.
Load the selected profile configuration files from /etc/anaconda/profile.d/*.conf.
Load the custom configuration files from /etc/anaconda/conf.d/*.conf.
Apply the kernel arguments.
Apply the cmdline options.
Generate the runtime configuration file /run/anaconda/anaconda.conf.

Python representation

The Anaconda configuration is represented by the conf object from pyanaconda.core.configuration.anaconda. The configuration sections are represented by properties of the conf object. The configuration options are represented by properties of the section representation. All these properties are read-only.

The conf object is initialized on the first import. It loads the runtime configuration file, if it exists, otherwise it loads the default configuration file. Its main purpose is to provide access to the configuration of the current installer run.

It is safe to use the conf object in the Anaconda DBus modules and in any other Python processes that are started after a runtime configuration file has been generated.

See an example of a Python code:

from pyanaconda.core.configuration.anaconda import conf

# Is Anaconda in the debugging mode?
print(conf.anaconda.debug)

# Is the type of the installation target hardware?
print(conf.target.is_hardware)

# A path to the system root of the target.
print(conf.target.system_root)