Module 'pg_autoscaler' has failed: division by zero

Bug #1868587 reported by Frode Nordahl
10
This bug affects 2 people
Affects Status Importance Assigned to Milestone
ceph (Ubuntu)
Fix Released
High
James Page

Bug Description

While enabling Ceph RadosGW on Bionic/Ussuri I have run into intermittent failures which appear to be caused by the PG autoscaler.

HEALTH_ERR Module 'pg_autoscaler' has failed: division by zero

It has recently been fixed upstream [0][1] and I do not think we have this in our packages.

0: https://github.com/ceph/ceph/commit/4a45b438c8921e47e329fb04535f8ecedf2a3051
1: https://github.com/ceph/ceph/pull/33420/commits/1db33f7785bc189df65aabd4b2459aa629009584

Frode Nordahl (fnordahl)
description: updated
Revision history for this message
James Page (james-page) wrote :

This fix was included in the 5.2.0 release on the 23rd March - so will be included in my upload for that to Focal development. Closing this bug as covered elsewhere.

Changed in ceph (Ubuntu):
status: New → Triaged
importance: Undecided → High
milestone: none → ubuntu-20.03
assignee: nobody → James Page (james-page)
status: Triaged → Fix Released
Revision history for this message
Thiago Martins (martinx) wrote :

I'm seeing this on the latest Ubuntu 20.04.1, fully upgraded.

Ceph version 15.2.3-0ubuntu0.20.04.2.

The `ceph status` is "HEALTH_ERR", Module 'pg_autoscaler' has failed: division by zero, 1 daemons have recently crashed.

:-(

To post a comment you must log in.
This report contains Public information  
Everyone can see this information.

Other bug subscribers

Remote bug watches

Bug watches keep track of this bug in other bug trackers.