Description of problem: ## Issue In object storage with radosgw, accumulating large quantities of objects in the data pool that appear to be orphans. Since this is a large cluster(2PB), customer cannot execute the command "radosgw-admin orphans find". It is taking VERY long time. ## Subscription Details - As per the SKU --SVCRC12E4000, customer having ceph subscription. ## Environment * ceph version 12.2.12 (1436006594665279fe734b4c15d7e08c13ebd777) luminous (stable) ## Root cause * The issues with the `orphans find` and `bucket check --fix` issues have the same root cause ie reuploaded multiparts. * The `orphan find` command removes (or is supposed to remove) the leaked reuploaded parts from the bucket data pool, and the `bucket check --fix` command should remove the leaked reuploaded parts from the bucket index. * Both sets of functionality are required to remediate the issues caused by user reuploading parts in a multipart upload. ## Summary / Tracker /PR http://tracker.ceph.com/issues/38454 https://tracker.ceph.com/issues/44722 https://tracker.ceph.com/issues/49018 https://github.com/ceph/ceph/pull/34148 https://github.com/ceph/ceph/pull/35005 https://github.com/ceph/ceph/pull/37260 https://github.com/ceph/ceph/pull/38905 Version-Release number of selected component (if applicable): 3.3 Related tool request: https://bugzilla.redhat.com/show_bug.cgi?id=1925342
Since the problem described in this bug report should be resolved in a recent advisory, it has been closed with a resolution of ERRATA. For information on the advisory (Red Hat Ceph Storage 5.0 bug fix and enhancement), and where to find the updated files, follow the link below. If the solution does not work for you, open a new bug report. https://access.redhat.com/errata/RHBA-2021:3294