[RELEASE] ScyllaDB Manager 3.7.0

Release Notes
, ,
1

The ScyllaDB team announces the release of ScyllaDB Manager 3.7.0, a production-ready minor release of the stable 3.7 branch.

ScyllaDB Manager is a centralized cluster administration and recurrent tasks automation tool.

This release focuses on the support for Alternator clusters backup and restore tasks and integration with the upcoming ScyllaDB 2025.4 release.

Below are the changes in this release.

Alternator backup and restore (#4509)

Even though Alternator can be thought of as just an alternative frontend to CQL, it comes with different tricks and considerations. Before ScyllaDB Manager 3.7.0 release, backing up and restoring Alternator clusters worked in the exact same way as for the CQL clusters. This worked in most scenarios, but it didn’t cover all corner cases where manual alterations would be needed.

ScyllaDB Manager 3.7.0 provides improvements to Alternator backup and restore tasks, so that manual alterations are no longer needed. In order for those improvements to take place, users need to add Alternator credentials for managed clusters to ScyllaDB Manager.

Alternator credentials are handled in the same way as the regular CQL credentials and can be added with: sctool cluster update --alternator-access-key-id --alternator-secret-access-key.
Both Alternator and CQL credentials can be linked to the same underlying CQL role (see authentication in Alternator for more context).
Note that both CQL and Alternator credentials are required for backup and restore tasks in Alternator clusters.

Restoring contents of Alternator Secondary Indexes has a slight limitation.
When restore task is completed, restored LSIs are available, yet they are still in the process of applying view updates. This means that querying LSIs right after restore might return partial results. When all updates are applied, LSIs will return full results.
The same applies to GSIs for ScyllaDB versions older than 2025.1.

Incremental repair for tablet keyspaces (#4591)

ScyllaDB 2025.4 adds incremental repair feature for tablet keyspaces. It allows for skipping the repair of already repaired sstables and hence decreases repair time and impact on the cluster.
By default, all repair tasks use incremental repair, but it is also possible to schedule full repair with the new sctool repair --incremental-mode flag.

Healthcheck intervals (#4445)

ScyllaDB Manager 3.7.0 increases the default healthcheck tasks interval to 1 minute to reduce healthcheck noise in bigger, multi-datacenter clusters. It also makes it possible to configure healthcheck tasks interval from scylla-manager.yaml config.

Compatibility with ScyllaDB 2025.4

These are the adjustments needed for achieving compatibility with ScyllaDB 2025.4 release:

  • Handle colocated table repair (#4565)
  • Recognize ms sstable format during restore (#4578)

Bug fixes and other improvements

  • Refresh node configurations before backup and restore tasks (#4524)
  • Display if cluster has CQL credentials added correctly (#4561)
  • Don’t always validate cluster connectivity on update (#4587)
  • Don’t panic when ScyllaDB Manager backend is unavailable (#4601)
  • Upgrade GOCQL to v1.15.3 (#4581)

Upgrade to the new release

ScyllaDB customers are encouraged to upgrade to ScyllaDB Manager 3.7.0 in coordination with the ScyllaDB support team.

The new release includes upgrades of both ScyllaDB Manager Server and Agent.

Useful Links:

Scylla Manager 3.7.0 supports the following ScyllaDB releases

  • 2025.4
  • 2025.3
  • 2025.2
  • 2025.1
  • 2024.1

You can install and run ScyllaDB Manager on Kubernetes using ScyllaDB Operator. More here.