Project

General

Profile

Actions
Copy link

Bug #49369

closed

Bucket removing using radosgw-admin is not working

Added by Jiffin Tony Thottan about 3 years ago. Updated 4 months ago.

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

0%

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

Description

OBC used in the Kubernetes environment delete bucket via rook using the following command

radosgw-admin bucket rm --bucket=<name> --purge-objects

This was working fine in ceph v15.2.8 but working in the master branch(didn't try on non-kube setup)

Files

bucketdeltiondebuglog (87.3 KB) bucketdeltiondebuglog Jiffin Tony Thottan, 03/15/2021 08:41 AM
Actions
Copy link
 #1

Updated by Casey Bodley about 3 years ago

This was working fine in ceph v15.2.8 but working in the master branch(didn't try on non-kube setup)

you mean not working on master? can you try to get a debug log from the radosgw-admin command with --debug-rgw 20 --debug-ms 1?

Actions
Copy link
 #2

Updated by Mark Kogan about 3 years ago

in addition to the debug logs asked by Casey,
is it possible to please try from inside the rgw pod?
(assuming that originally it did not work from the toolbox pod)

oc exec -it -n rook-ceph $(oc -n rook-ceph get pod -l "app=rook-ceph-rgw" -o jsonpath='{.items[0].metadata.name}') -- bash

radosgw-admin --version

radosgw-admin --keyring=/etc/ceph/keyring-store/keyring --mon-host=$ROOK_CEPH_MON_HOST --name=client.radosgw.gateway bucket rm --bucket=<name> --purge-objects
Actions
Copy link
 #3

Updated by Casey Bodley about 3 years ago

  • Status changed from New to Need More Info
Actions
Copy link
 #5

Updated by Jiffin Tony Thottan about 3 years ago

Hi Casey and Mark,

Sorry for much delayed reply, I lost track for this issue
Here radosgw-admin command was running from the rook-operator pod which has built on ceph octopus image
  1. radosgw-admin --version(operator pod)
    ceph version 15.2.9 (357616cbf726abb779ca75a551e8d02568e15b17) octopus (stable)

And ceph cluster was running on master
#radosgw-admin --version (rgw pod)
ceph version 17.0.0-1983-g6a19e303 (6a19e303187c2defceb9c785284ca401a4309c47) quincy (dev)

Rook operator was removing the bucket as part of the OBC delete request which will be executed from the rook-operator pod. That deletion was failing. If I delete from rgw pod it is working fine.

Is the version difference the issue here? Even it that case it should return some error right?

Attaching the debug log for "radosgw-admin bucket rm" command from operator pod run

Actions
Copy link
 #6

Updated by Jiffin Tony Thottan 4 months ago

  • Status changed from Need More Info to Closed

Not seen in new versions of radosgw-admin cmd

Actions
Copy link

Also available in: Atom PDF