How to integrate in existing Ansible configuration repository?

[mueller-ma]

Hi,
I'd like to integrate this role in my already existing Ansible configuration. I created a host group (db), added this role to requirements.txt and added variables as group variables.
However using the following playbook doesn't execute the role:

- name: Deploy Postgres
  hosts: db
  roles:
    - role: vitabaks.postgresql_cluster

$ ansible-playbook playbooks/role_postgres.yml

PLAY [Deploy Postgres] ***************************************************************************************

TASK [Gathering Facts] ***************************************************************************************
ok: [db-01]

PLAY RECAP ***************************************************************************************************
db-01                  : ok=1    changed=0    unreachable=0    failed=0    skipped=0    rescued=0    ignored=0

I saw #677, but IMO it's too dangerous as your playbooks use hosts: all and when someone forgets --limit db, Ansible installs Postgres on all servers.

[vitabaks]

@mueller-ma Hi

postgresql_cluster is not a role; it is a project based on Ansible, consisting of many playbooks and about 40 roles.

Please use it according to the README.

[mueller-ma]

That means I have to duplicate my Ansible inventory or split it up into two: One for most servers, one for Postgres.

Will it be possible to use such playbook from my first post when this repo has been converted to a collection (see #272)?

[vitabaks]

Your inventory file must be compatible with the format that postgresql_cluster assumes.
https://github.com/vitabaks/postgresql_cluster/blob/master/inventory

[vitabaks]

Will it be possible to use such playbook from my first post when this repo has been converted to a collection

So far, I have no plans to change the format to the collection.

Version 2.0 will be released soon, which will provide a UI to hide all this "complexity" associated with inventory and variables files.