Chapter 7. Asynchronous errata updates

The snap-schedule module retains a defined number of snapshots

With this release, snap-schedule module supports a new retention specification to retain a user-defined number of snapshots. For example, if you have specified 50 snapshots to retain irrespective of the snapshot creation cadence, then the snapshot is pruned after a new snapshot is created. The actual number of snapshots retained is 1 less than the maximum number specified. In this example, 49 snapshots are retained so that there is a margin of 1 snapshot that can be created on the file system on the next iteration. The retained snapshot avoids breaching the system configured limit of mds_max_snaps_per_dir.

Introduction of dump dir command

With this release, a new command dump dir is introduced to dump the directory information.

Snapshot schedule for subvolume with option --subvol is now available

With this enhancement, users can schedule snapshots for subvolumes in default as well as non-default subvolume groups using the --subvol argument.

The output of the counter dump command is now split into different sections

With this enhancement, labeled performance counters for the Ceph Object Gateway operation metrics are split into different sections in the output of counter dump command for the user operation counters and bucket operation counters.

Introduction of a new configuration variable rgw_op_counters_dump_expiration

With this enhancement, a new configuration variable rgw_op_counters_dump_expiration is introduced. This new variable controls the number of seconds a labeled performance counter is going to be emitted from the ceph counter dump command.

Perf counters for S3 operations can now be sent Prometheus

With this enhancement, you can send a new labeled counter to Prometheus by using the Ceph exporter daemon and is useful for common S3 operations, such as PUTs, which can be visualized per bucket or per user.

Large omap warnings in the archive zone environment

Presently, in the archive zone environment, it is possible that you see large omap warnings due to several versions of the same object in a bucket being written to a single bucket index shard object. The recommendation is to reduce the 'max_objs_per_shard' config option to 50,000 to account for the omap olh entries on the archive zone. This will help keep the number of omap entries per bucket index shard object in check to prevent large omap warnings.

Token import throws an error when trying to import multisite configuration from a remote cluster

The multi-site period information does not return realm name now due to this import multi-site configuration gives error on submitting the form. Workaround: Import the multi-site configuration through command line interface.

CVE-2023-49569 for grafana-container: go-git is present in the Ceph 7.0z2 release

A CVE is present in Ceph 7.0z2 because the Ceph 7.0 release uses an older version of Grafana than what is available in the Ceph 6.1 release.

Improved rbd_diff_iterate2() API performance

Previously, RBD diff-iterate was not guaranteed to execute locally if exclusive lock was available when diffing against the beginning of time (fromsnapname == NULL) in fast-diff mode (whole_object == true with fast-diff image feature enabled and valid).

rgw-restore-bucket-index tool can now restore the bucket indices for versioned buckets

With this enhancement, the rgw-restore-bucket-index tool now works as broadly as possible, with the ability to restore the bucket indices for un-versioned as well as for versioned buckets.

Some Ceph Object Gateway applications using S3 client SDKs can experience unexpected errors

Presently, some applications using S3 client SDKs could experience an unexpected 403 error when uploading a zero-length object, if an external checksum is requested.

Prometheus metrics are no longer used

This release introduces new feature-rich labeled perf counters, replacing the Object Gateway-related Prometheus metrics previously used. The new metrics are being introduced before complete removal to allow overlapping usage.