Project

General

Profile

Actions
Copy link

Bug #22721

closed

Resharding hangs with versioning-enabled buckets

Added by Martin Emrich over 6 years ago. Updated over 5 years ago.

Status:
Resolved
Priority:
High
Assignee:
Orit Wasserman
Target version:
-
% Done:

0%

Source:
Community (user)
Tags:
Backport:
jewel, luminous
Regression:
No
Severity:
2 - major
Reviewed:
Affected Versions:
Ceph - v12.2.2
ceph-qa-suite:
Pull request ID:
Crash signature (v1):
Crash signature (v2):

Description

If I enable S3 bucket versioning, resharding the bucket (either automatic or manual) hangs, and blocks access to the bucket. I found no way to restore such a bucket to good health except deleting all rgw pools and starting from scratch.

Steps to reproduce:
  • configure radosgw
  • create a bucket
  • enable S3 bucket versioning (e.g. "aws --profile ceph --endpoint-url=http://ceph-gateway:7480 s3api put-bucket-versioning --bucket testbucket --versioning-configuration Status=Enabled")
  • Upload loads of objects (I tested with >> 100k)
  • Either wait for automatic resharding to "kick in", or do manual resharding (same result), e.g. with "radosgw-admin bucket reshard --bucket=testbucket --num-shards=10"

Related issues 2 (0 open — 2 closed)

Copied to rgw - Backport #23886: luminous: Resharding hangs with versioning-enabled bucketsResolvedOrit WassermanActions
Copied to rgw - Backport #23887: jewel: Resharding hangs with versioning-enabled bucketsRejectedActions
Actions
Copy link

Also available in: Atom PDF