Update of pg_num/pgp_num is silently ignored when autoscale_mode is on

[priteau]

Updated pg_num/pgp_num values for a pool are silently ignored if autoscale_mode is on:

                if details['pg_autoscale_mode'] == 'on':
                    delta.pop('pg_num', None)
                    delta.pop('pgp_num', None)

https://github.com/stackhpc/ansible-collection-cephadm/blob/1.14.0/plugins/modules/cephadm_pool.py#L557-L559

This means that autoscale needs to be disabled first, but it has to be done in a separate run:

1. run with pg_autoscale_mode: off if it is currently on
2. run again to apply updated pg_num/pgp_num values

[priteau]

A warning message is generated by the module but it is only visible when running with --verbose: https://github.com/stackhpc/ansible-collection-cephadm/blob/1.14.0/plugins/modules/cephadm_pool.py#L562

[cityofships]

Good find - although personally I wouldn't be putting PG/PGP numbers into Ansible vars if I had autoscale enabled for a pool. Creates risk of rolling back changes done by pool autoscaler when playbook is run.

Maybe this was the intention of the original authors of that module.

[priteau]

Good find - although personally I wouldn't be putting PG/PGP numbers into Ansible vars if I had autoscale enabled for a pool. Creates risk of rolling back changes done by pool autoscaler when playbook is run.

Maybe that was the intention of the original authors of the module too.

I agree, but I think there are still two issues:

• if you set pg_num/pgp_num and disable autoscale for a pool, you should be able to apply it in one run instead of two. This should be an easy fix
• the fact that pg_num/pgp_num gets ignored if pg_autoscale_mode: off should be discovered more easily. I would be OK with raising an error.