saltstack-node-specific-pillar-states

Allows for dynamic node specific pillar data, derived from a nodes id, without explicitly specifying the nodename in top.sls. Furthermore allows to configure the states a node should have via pillar, making the state top.sls only contain information for the specific node thus preserving privacy of node identifiers (which would otherwise be shared with all nodes, due to state top.sls being shared with all nodes).

Usecase is all environments where you do not want one salt-minion to know the ids and statenames of the others.

Examples

To configure e.g. the ssh state to be applied to node with id myhost.example.com create a file $pillar_root/myhost.sls with content:

states:
  ssh:

This will result in a top.sls for the pillar looking like this:

# salt-run pillar.show_top
base:
    ----------
    *:
        - myhost

And the pillar items:

# salt-call pillar.items
local:
    ----------
    states:
        ----------
        ssh:

The top.sls for the states will then look like this:

# salt-call state.show_top
local:
    ----------
    base:
        - ssh

and provided you have a file $file_root/ssh.sls or $file_root/ssh/init.sls the state will be applied. Also the copy of the $file_root will only contain the content of the original top.sls, e.g.:

base:
  '*':
  {% for state in pillar.get('states', {}).keys() %}
    - {{ state }}
  {% endfor %}