[ansible-project] May I treat playbooks similar to database migrations?

[Jan Koprowski]

Hello,

I would like to start writing playbooks treating each playbook similar to *database
migration*.

Let say I'm releasing my roles/playbooks repository with version 1.0
In version 1.1 I would like to upgrade reviewboard to version 2.5 (from
1.5), glusterfs to version 3.6 (from 3.4) and introduce deduplication.

Is it a good idea to write playbooks like?
- 1_*upgrade_reviewboard_1.5_to_2.5.yml*
* - 2_upgrade_glusterfs_3.4_to_3.6.yml*
* - 3_turn_on_glusterfs_deduplication.yml*

Assuming that something will know how which playbooks should be run and
in which order is it a good direction to follow?

The idea behind that is to be able to share infrastructure with others as
repository.
Treat it as a product/software where people can upgrade to higher version.

Let say on each node we have deployed fact file returning version of
infrastructure so we can easily determine which version is current.

Regards,

Jan