Skip to content

Release 0.16.0

This release introduces snapshots, partition support in the console, maintenance mode, a partition count preflight check, and faster restores through segment filtering.

For new installations, see the Installation guide.

For upgrading existing installations, see the associated Upgrading to 0.16.x guide.

This feature has been requested a lot, and we are happy to finally bring it to Armory.

Snapshots let Armory capture a topic’s backup data as distinct, numbered versions over time. Each snapshot is a self-contained version of the backup, so you can keep a clear history and restore from any point you choose. This helps the backup to automatically recover from topics that get recreated.

For more information, check out the new Snapshots user guide.

The console now shows progress for each partition of a topic, not just the topic as a whole. New partition tables across the backup and restore pages let you see how far along each partition is, making it easy to spot one that is lagging behind the rest.

Maintenance mode lets platform administrators temporarily put the platform in a restricted state during planned upgrades or outages. It is configured through the Helm chart’s global values:

global:
maintenance:
enabled: true
banner:
message: "Scheduled maintenance in progress."
variant: "warning" # info, warning, error

When maintenance mode is enabled, backup workloads are scaled down to zero, the API returns 503 Service Unavailable for all GraphQL and REST endpoints, and the console redirects users to a dedicated maintenance page.

The banner is shown to users regardless of whether maintenance mode is enabled, so you can also use it to announce upcoming maintenance without pausing operations.

Restores that use time, offset, or retention filters now perform segment-level pre-filtering. Armory evaluates the backup index up front and skips any segment that falls entirely outside the configured filter window, avoiding the download and scan of data that would be discarded anyway.

When you configure a restore in the console, it now runs a preflight check that compares the number of partitions in the source backup with the target topic. When the counts differ, the console warns you before the restore starts, so you can catch a misconfigured target topic instead of discovering it after the data has been written.

Partition count preflight check warning
Partition count preflight check warning
A preflight warning when the source backup and target topic partition counts differ.
  • Console: the topic selection pages now show a warning when no topics are found, pointing you to the likely cause (an empty event hub or missing permissions) instead of just an empty list.
  • Operator: reserve label headroom for StatefulSet child names to stay within the 63-character limit.

  • Backup: prevent out-of-band errors from fanning out to all backup tasks and restarting them

For a full list of changes, see the Changelog.