Project

General

Profile

Actions
Copy link

Bug #52091

closed

multisite: if mdlogs are trimmed prematurely, 'radosgw-admin sync status' stays behind

Added by Casey Bodley almost 3 years ago. Updated over 1 year ago.

Status:
Resolved
Priority:
Normal
Assignee:
-
Target version:
-
% Done:

100%

Source:
Tags:
multisite
Backport:
octopus pacific
Regression:
No
Severity:
3 - minor
Reviewed:
Affected Versions:
ceph-qa-suite:
Pull request ID:
42708
Crash signature (v1):
Crash signature (v2):

Description

'radosgw-admin sync status' compares each shard's local sync status marker with the remote's max-marker. if the remote trims their mdlog before sync can catch up, sync status shows that the shard is behind even though there's nothing there to sync

Related issues 2 (0 open — 2 closed)

Copied to rgw - Backport #52107: pacific: multisite: if mdlogs are trimmed prematurely, 'radosgw-admin sync status' stays behindResolvedCory SnyderActions
Copied to rgw - Backport #52108: octopus: multisite: if mdlogs are trimmed prematurely, 'radosgw-admin sync status' stays behindResolvedCory SnyderActions
Actions
Copy link

Also available in: Atom PDF