Project

General

Profile

Bug #61650

pg can't recover from LAGGY state if the pool size is 1

Added by Xuehan Xu 9 months ago. Updated 8 months ago.

Status:
New
Priority:
Low
Assignee:
-
Category:
-
Target version:
-
% Done:

0%

Source:
Tags:
Backport:
Regression:
No
Severity:
3 - minor
Reviewed:
Affected Versions:
ceph-qa-suite:
Component(RADOS):
Pull request ID:
Crash signature (v1):
Crash signature (v2):

Description

In one of our online clusters, some pool is configured to have no replicas.

There are times that some PGs keep being "laggy" when we try to add new OSDs to those pools

History

#1 Updated by Xuehan Xu 9 months ago

Our clusters' version is v17.2.6

#2 Updated by Radoslaw Zarzynski 8 months ago

  • Priority changed from Normal to Low

Lowering per "no replica".

Also available in: Atom PDF