Bug #63541: Observing client.admin crash in thread_name 'rados' on executing 'rados clearomap..' - RADOS - Ceph

Actions

Copy link

Bug #63541

open

Observing client.admin crash in thread_name 'rados' on executing 'rados clearomap..'

Added by Nitzan Mordechai 6 months ago. Updated 4 months ago.

Status:

Pending Backport

Priority:

Normal

Assignee:

Nitzan Mordechai

Category:

Target version:

Ceph - v19.0.0

% Done:

Source:

Community (dev)

Tags:

backport_processed

Backport:

pacific quincy reef

Regression:

Severity:

3 - minor

Reviewed:

Affected Versions:

ceph-qa-suite:

Component(RADOS):

Pull request ID:

54518

Crash signature (v1):

Crash signature (v2):

Description

Description of problem:

Observing client.admin crash in thread_name 'rados' on executing 'rados clearomap' for a rados pool different than the one where the object is present.

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

ceph version 18.2.0-121.el9cp

How reproducible:
Always

Steps to Reproduce:

1. Object '.dir.ae4142d0-b2e7-4ac5-977c-8507726723b5.334941.6.4' is present in the pool test_zone.rgw.buckets.index
2. Now try to perform the 'rados clearomap' over a different pool say 'test_zone.rgw.buckets.data' and we observe the crash.

sample CLI to hit the crash on node 10.8.129.101 (root/passwd)

rados clearomap -p test_zone.rgw.buckets.data .dir.ae4142d0-b2e7-4ac5-977c-8507726723b5.334941.6.4

Actual results:

observing the client.admin crash with rados clearomap

Expected results:

1. should not observe any client.admin crash
2. the cli should error out with a message that object is not present in the pool.

Related issues 3 (1 open — 2 closed)