.. _states-top: ============ The Top File ============ Introduction ============ Most infrastructures are made up of groups of machines, each machine in the group performing a role similar to others. Those groups of machines work in concert with each other to create an application stack. To effectively manage those groups of machines, an administrator needs to be able to create roles for those groups. For example, a group of machines that serve front-end web traffic might have roles which indicate that those machines should all have the Apache webserver package installed and that the Apache service should always be running. In Salt, the file which contains a mapping between groups of machines on a network and the configuration roles that should be applied to them is called a ``top file``. Top files are named ``top.sls`` by default and they are so-named because they always exist in the "top" of a directory hierarchy that contains state files. That directory hierarchy is called a ``state tree``. A Basic Example =============== Top files have three components: - Environment: A state tree directory containing a set of state files to configure systems. - Target: A grouping of machines which will have a set of states applied to them. - State files: A list of state files to apply to a target. Each state file describes one or more states to be configured and enforced on the targeted machines. The relationship between these three components is nested as follows: - Environments contain targets - Targets contain states Putting these concepts together, we can describe a scenario in which all minions with an ID that begins with ``web`` have an ``apache`` state applied to them: .. code-block:: yaml base: # Apply SLS files from the directory root for the 'base' environment 'web*': # All minions with a minion_id that begins with 'web' - apache # Apply the state file named 'apache.sls' .. _states-top-environments: Environments ============ Environments are directory hierarchies which contain a top files and a set of state files. Environments can be used in many ways, however there is no requirement that they be used at all. In fact, the most common way to deploy Salt is with a single environment, called ``base``. It is recommended that users only create multiple environments if they have a use case which specifically calls for multiple versions of state trees. Getting Started with Top Files ============================== Each environment is defined inside a salt master configuration variable called, :conf_master:`file_roots` . In the most common single-environment setup, only the ``base`` environment is defined in :conf_master:`file_roots` along with only one directory path for the state tree. .. code-block:: yaml file_roots: base: - /srv/salt In the above example, the top file will only have a single environment to pull from. Next is a simple single-environment top file placed in ``/srv/salt/top.sls``, illustrating that for the environment called ``base``, all minions will have the state files named ``core.sls`` and ``edit.sls`` applied to them. .. code-block:: yaml base: '*': - core - edit Assuming the ``file_roots`` configuration from above, Salt will look in the ``/srv/salt`` directory for ``core.sls`` and ``edit.sls``. Multiple Environments ===================== In some cases, teams may wish to create versioned state trees which can be used to test Salt configurations in isolated sets of systems such as a staging environment before deploying states into production. For this case, multiple environments can be used to accomplish this task. To create multiple environments, the :conf_master:`file_roots` option can be expanded: .. code-block:: yaml file_roots: dev: - /srv/salt/dev qa: - /srv/salt/qa prod: - /srv/salt/prod In the above, we declare three environments: ``dev``, ``qa`` and ``prod``. Each environment has a single directory assigned to it. Our top file references the environments: .. code-block:: yaml dev: 'webserver*': - webserver 'db*': - db qa: 'webserver*': - webserver 'db*': - db prod: 'webserver*': - webserver 'db*': - db As seen above, the top file now declares the three environments and for each, targets are defined to map globs of minion IDs to state files. For example, all minions which have an ID beginning with the string ``webserver`` will have the webserver state from the requested environment assigned to it. In this manner, a proposed change to a state could first be made in a state file in ``/srv/salt/dev`` and then be applied to development webservers before moving the state into QA by copying the state file into ``/srv/salt/qa``. Choosing an Environment to Target ================================= The top file is used to assign a minion to an environment unless overridden using the methods described below. The environment in the top file must match an environment in :conf_master:`file_roots` in order for any states to be applied to that minion. The states that will be applied to a minion in a given environment can be viewed using the :py:func:`state.show_top ` execution function. Minions may be pinned to a particular environment by setting the ``environment`` value in the minion configuration file. In doing so, a minion will only request files from the environment to which it is assigned. The environment to use may also be dynamically selected at the time that a ``salt``, ``salt-call`` or ``salt-ssh`` by passing passing a flag to the execution module being called. This is most commonly done with functions in the ``state`` module by using the ``saltenv=`` argument. For example, to run a ``highstate`` on all minions, using the state files in the ``prod`` state tree, run: ``salt '*' state.highstate saltenv=prod``. .. note:: Not all functions accept ``saltenv`` as an argument See individual function documentation to verify. Shorthand ========= If you assign only one SLS to a system, as in this example, a shorthand is also available: .. code-block:: yaml base: '*': global dev: 'webserver*': webserver 'db*': db qa: 'webserver*': webserver 'db*': db prod: 'webserver*': webserver 'db*': db Advanced Minion Targeting ========================= In addition to globs, minions can be specified in top files a few other ways. Some common ones are :doc:`compound matches ` and :doc:`node groups `. Below is a slightly more complex top file example, showing the different types of matches you can perform: .. code-block:: yaml # All files will be taken from the file path specified in the base # environment in the ``file_roots`` configuration value. base: # All minions get the following three state files applied '*': - ldap-client - networking - salt.minion # All minions which have an ID that begins with the phrase # 'salt-master' will have an SLS file applied that is named # 'master.sls' and is in the 'salt' directory, underneath # the root specified in the ``base`` environment in the # configuration value for ``file_roots``. 'salt-master*': - salt.master # Minions that have an ID matching the following regular # expression will have the state file called 'web.sls' in the # nagios/mon directory applied. Additionally, minions matching # the regular expression will also have the 'server.sls' file # in the apache/ directory applied. # NOTE! # # Take note of the 'match' directive here, which tells Salt # to treat the target string as a regex to be matched! '^(memcache|web).(qa|prod).loc$': - match: pcre - nagios.mon.web - apache.server # Minions that have a grain set indicating that they are running # the Ubuntu operating system will have the state file called # 'ubuntu.sls' in the 'repos' directory applied. # # Again take note of the 'match' directive here which tells # Salt to match against a grain instead of a minion ID. 'os:Ubuntu': - match: grain - repos.ubuntu # Minions that are either RedHat or CentOS should have the 'epel.sls' # state applied, from the 'repos/' directory. 'os:(RedHat|CentOS)': - match: grain_pcre - repos.epel # The three minions with the IDs of 'foo', 'bar' and 'baz' should # have 'database.sls' applied. 'foo,bar,baz': - match: list - database # Any minion for which the pillar key 'somekey' is set and has a value # of that key matching 'abc' will have the 'xyz.sls' state applied. 'somekey:abc': - match: pillar - xyz # All minions which begin with the strings 'nag1' or any minion with # a grain set called 'role' with the value of 'monitoring' will have # the 'server.sls' state file applied from the 'nagios/' directory. 'nag1* or G@role:monitoring': - match: compound - nagios.server How Top Files Are Compiled ========================== When using multiple environments, it is not necessary to create a top file for each environment. The most common approach, and the easiest to maintain, is to use a single top file placed in only one environment. However, some workflows do call for multiple top files. In this case, top files may be merged together to create ``high data`` for the state compiler to use as a source to compile states on a minion. For the following discussion of top file compilation, assume the following configuration: ``/etc/salt/master``: .. code-block:: yaml file_roots: first_env: - /srv/salt/first second_env: - /srv/salt/second ``/srv/salt/first/top.sls``: .. code-block:: yaml first_env: '*': - first second_env: '*': - second The astute reader will ask how the state compiler resolves which should be an obvious conflict if a minion is not pinned to a particular environment and if no environment argument is passed into a state function. Given the above, it is initially unclear whether ``first.sls`` will be applied or whether ``second.sls`` will be applied in a ``salt '*' state.highstate`` command. When conflicting keys arise, there are several configuration options which control the behaviour of salt: - ``env_order`` Setting ``env_order`` will set the order in which environments are processed by the state compiler. - ``top_file_merging_strategy`` Can be set to ``same``, which will process only the top file from the environment that the minion belongs to via the ``environment`` configuration setting or the environment that is requested via the ``saltenv`` argument supported by some functions in the ``state`` module. Can also be set to ``merge``. This is the default. When set to ``merge``, top files will be merged together. The order in which top files are merged together can be controlled with ``env_order``. - ``default_top`` If ``top_file_merging_strategy`` is set to ``same`` and an environment does not contain a top file, the top file in the environment specified by ``default_top`` will be used instead.