Bug #47555
openIn a multi-site environment, cosbench writes and deletes the master object, and the master object remains
0%
Description
In my test.cosbench writes and deletes the master object, and the master object remains. I found out that it was because master full sync was after del object。Look object.png
Here is a simple way to reproduce！！！
rgwA and rgwB are two zones of the same zonegroup
t0: rgwA and rgwB is running (set rgw_run_sync_thread=true)
t1: rgwA set rgw_run_sync_thread=false and restart it . We use s3cmd to create a bucket. And upload an object1. We use s3cmd to observe whether object1 has been synchronized in rgwB. If the synchronization has passed, proceed to the next step.
t2:rgwB set rgw_run_sync_thread=false and restart it . rgwA delete object1 .
t3:rgwA set rgw_run_sync_thread=true and restart it . LOOK radosgw-admin bucket sync status is cauht up it.
t4: rgwB set rgw_run_sync_thread=true and restart it. LOOK radosgw-admin bucket sync status is cauht up . the reslut: rgwA has object1,rgwB dosen't have object1.
Files
Updated by ren weiguo about 3 years ago
In the master branch, this problem was also reproduced.so If the bucket about the rgwA instance is
not in the incremental synchronization state, can we prohibit rgwA
from deleting object1? or Don't do it.
I also send 'In the ceph multisite master-zone, read ,write,delete objects, and the master-zone has data remaining' in ceph-devel@vger.kernel.org