.The radosgw-admin utility supports a new bucket limitcheck command
The `radosgw-admin` utility has a new command `bucket limitcheck` to warn the administrator when a bucket needs resharding. Previously, buckets with more objects than is recommended could be unnoticed and cause performance issues. This new command reports on bucket status with respect to the configured bucket sharding recommendations ensuring that administrators can detect overloaded buckets easily.
Description of problem:
RGW should generate warnings when buckets are in danger exceeding the same maximum number of entries per index shard (appx. 100000, for buckets with indexes)
Version-Release number of selected component (if applicable):
No comparable feature exists.
We propose to address the issue by providing a new radosgw-admin sub-command which scans existing, indexed buckets and reports on those which have exceeded or have approached within a defined percentage (default 10%) of the safe maximum entry limit of 100K per index shard.
Proposed upstream PR:
Matt would you please rebase the PR at https://github.com/ceph/ceph/pull/11796 so we can get this upstream?
(In reply to Ken Dreyer (Red Hat) from comment #11)
> Matt would you please rebase the PR at
> https://github.com/ceph/ceph/pull/11796 so we can get this upstream?
@ken, rebased last week, I asked Casey to re-review.
Made the changes. Verified with 5 different buckets with varying objects.
Warning messages are displayed as per the configuration set.
rgw_override_bucket_index_max_shards = 10
rgw_safe_max_objects_per_shard = 1000
radosgw-admin bucket limit check --warnings-only
"fill_status": "WARN 100.000000%"
"fill_status": "OVER 100.000000%"
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, and where to find the updated
files, follow the link below.
If the solution does not work for you, open a new bug report.