Project

General

Profile

Actions

Bug #17155

closed

Cache fronted erasure pool does delete all objects on a rbd rm

Added by Bryan Ballard over 7 years ago. Updated over 7 years ago.

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

0%

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

Description

So if I delete a RBD using rbd rm ec-backup/test1 (1TB) It completes and the objects are all deleted, The header does appear to be delete. The only rbd that was created in the pools was the test1 that is described

Objects before rbd creation:
4 objects in hot-cache
0 object in ec-backup

Before deletion of created rbd (was formated and some minor data put on it):
574 object in the hot-cache (ssd replicate 3)
32 objects in the ec-backup (K=3,M=2)

After deletion:
558 objects in the hot-cache
17 objects in the ec-backup

It would seem rbd rm is not handling the deletion correctly.

let me know what logs / settings to show would be helpful.

If this just a RTMF issue let me know, since I can't seem to find any reference to this behavior.

Actions #1

Updated by Bryan Ballard over 7 years ago

I noticed a typo. It should be does not in the title and body

Actions #2

Updated by Jason Dillaman over 7 years ago

  • Status changed from New to Duplicate

This is a duplicate of #14865 -- a fix has been proposed but its waiting on more test cases.

Actions #3

Updated by Bryan Ballard over 7 years ago

So is there a workaround, that would allow me to clean up the orphaned objects safely?

Actions

Also available in: Atom PDF