Bug 2121097 - Ceph PG Autoscaler switched on for existing pools post rolling_update playbook
Summary: Ceph PG Autoscaler switched on for existing pools post rolling_update playbook
Alias: None
Product: Red Hat Ceph Storage
Classification: Red Hat Storage
Component: Ceph-Ansible
Version: 4.3
Hardware: x86_64
OS: Linux
Target Milestone: ---
: 5.3z1
Assignee: Teoman ONAY
QA Contact: Vivek Das
Depends On:
TreeView+ depends on / blocked
Reported: 2022-08-24 12:54 UTC by Raz
Modified: 2022-09-21 15:24 UTC (History)
8 users (show)

Fixed In Version:
Doc Type: If docs needed, set a value
Doc Text:
Clone Of:
Last Closed: 2022-09-21 15:24:16 UTC

Attachments (Terms of Use)

System ID Private Priority Status Summary Last Updated
Ceph Project Bug Tracker 57180 0 None None None 2022-08-24 15:50:59 UTC
Red Hat Issue Tracker RHCEPH-5141 0 None None None 2022-08-24 13:02:52 UTC

Description Raz 2022-08-24 12:54:38 UTC
Description of problem:

All pools in the cluster should have their pg autoscaler states set fact by the rolling update.yml playbook, which should also disable the pools before beginning the upgrade and re-enable pg autoscaler on all pools from the start.

All of our pools had an off/warn pg autoscaler status when the rolling update playbook was executed on our environment, however it appears that the state of all cluster's pool autoscaler changed to a "on" state.

Version-Release number of selected component (if applicable):

How reproducible:
Using the rolling update.yml playbook, upgrade the RHCEPH version 4.2z4 to 5.0z4 cluster with pg autoscaler enabled for any existing pools in the off/warn state.

Steps to Reproduce:
1. Running RHCEPH 4.2z4
2. Verifying that the pg autoscaler's existing pools are in the off/warn state.
3. implementing the infrastructure/rolling update.yml Ansible playbook to upgrade to version 5.0z4
4. Review ceph pg autoscaler status cmd for pools autoscaler status once 

upgrading is successful (switched to "on" state).
Actual results:
Review the ceph pg autoscaler status command for pools autoscaler status once the update is successful (switched to "on" state).

Expected results:
Existing pools with an off/warn state for pg autoscaler should remain in that state after running the rolling update.yml playbook.

Additional info:
The next task in the rolling update.yml playbook might be commented as a solution to this problem.

- name: re-enable pg autoscale on pools

Note You need to log in before you can comment on or make changes to this bug.