CHANGELOG.md

k8s-backup-operator Changelog

All notable changes to this project will be documented in this file.

The format is based on Keep a Changelog, and this project adheres to Semantic Versioning.

[Unreleased]

Added

• [#123] add german docs translations

[v3.3.3] - 2026-07-10

Fixed

• [#121] Add NetworkPolicy to allow ingress traffic from prometheus

[v3.3.2] - 2026-06-11

Changed

• [#117] Use Backup retry timeout in requeue timing

[v3.3.1] - 2026-06-03

• [#113] Load backupRetryTimeLimit dynamically on each backup creation

[v3.3.0] - 2026-05-29

Changed

• [#113] Store backup retry time via config map

[v3.2.0] - 2026-04-17

Changed

• [#111] Restore now cleans up additional backup-scoped resources before recreating them from the backup
  • Workloads that depend on those resources can be labeled for temporary scale-down during restore, and their previous replica count is restored afterwards.

Fixed

• set retry time limit for backup reconciliation to default 1h (60 minutes)

[v3.1.0] - 2026-04-14

Added

• [#104] add a retry time limit for backup reconciliation, default 1h
• [#109] Backup additional resources with the k8s.cloudogu.com/backup-scope label key

[v3.0.3] - 2026-03-26

Fixed

• [#106] Add missing RBAC for maintenance handling
• [#107] Add imagePullSecrets to cronjobs created from backupschedules to fix image pull errors in airgapped environments

[v3.0.2] - 2026-03-12

Changed

• Update version constraint to match latest velero version

[v3.0.1] - 2026-03-05

Security

• [#101] Fix Go stdlib CVE-2025-68121

[v3.0.0] - 2026-02-27

Important

Breaking change! New compatible versions of k8s-service-discovery and k8s-ces-assets are required.

Changed

• [#99] Write maintenance mode to special maintenance ConfigMap
  • Previously, the maintenance mode was stored in the global config, which caused Dogus to restart unnecessarily.

[v2.2.3] - 2026-01-28

Removed

• [#97], [#89] Backup & restore of owner references
  • This is not needed anymore, because we only back up Dogu-specific data. The dogu-operator is able to restore owner references itself now.

[v2.2.2] - 2026-01-22

Changed

• [#93] Update backup documentation
• [#95] Enhance error handling when fetching resources in the backup and restore process.

[v2.2.1] - 2025-12-18

Fixed

• [#91] Make kubectl image configurable and fix image not found error in airgapped environments.

[v2.2.0] - 2025-12-12

Added

• [#87] Add metrics for backups and restores.

[v2.1.0] - 2025-11-21

Changed

• [#85] Save backup metadata in annotations also on velero backups.
  • This enables restoring backups in "new" clusters when using the Admin-UI.

Fixed

• [#85] deleting backups.
• [#85] adding finalizer to synchronized backups from velero-backups.

[v2.0.0] - 2025-11-07

Breaking Change ahead!

Backups now only contain dogu-related data & config.

Changed

• [#79] Only cleanup dogus before restoring
• [#80] Backups now only contain the following:
  • Dogus, Configmaps, Secrets and PVCs

[v1.6.0] - 2025-09-03

Changed

• [#71] Move backup CRDs into k8s-backup-lib

[v1.5.0] - 2025-07-17

Added

• [#69] add metadata mapping for logLevel

[v1.4.7] - 2025-07-11

Changed

• [#67] Enhanced documentation and integrated other sources

[v1.4.6] - 2025-06-26

Fixed

• [#65] Empty namespace when creating synced backups caused error

[v1.4.5] - 2025-06-06

Changed

• [#63] Update version constraint to match latest velero version

[v1.4.4] - 2025-06-05

Changed

• [#61] Update velero dependency
• [#61] Restore status only for resources in group k8s.cloudogu.com

[v1.4.3] - 2025-05-14

Changed

• [#57] Update k8s-io-libs to v0.33.0

[v1.4.2] - 2025-05-13

Fixed

• [#59] Restore of an old ecosystem-certificate could make the ecosystem unavailable

Changed

• [#59] Exclude ecosystem-certificate from cleanup

[v1.4.1] - 2025-05-08

Fixed

• [#54] Restore of the velero deployment disrupted the restore
  • Previously, the Velero deployment would get restored as well, which caused disruptions of the restore if the deployment is different from the one in the backup.
• [#56] Cleanup of backup components lead to errors after restore
  • This is because the component operator would detect a downgrade, which is not allowed. Or worse, an upgrade during the restore operation would cause it to fail.

Changed

• [#54] Exclude all resources with label k8s.cloudogu.com/part-of: backup from restores.
• [#56] Exclude all backup-related components from cleanup when restoring.

[v1.4.0] - 2025-05-07

Added

• [#50] Exclude resources in cleanup before restore

[v1.3.4] - 2025-04-30

Changed

• [#51] Set sensible resource requests and limits

[v1.3.3] - 2025-04-10

Fixed

• [#44] Fixed resource ownerReferences after a restore of a previous backup
  • subsequently, this fixes the failure to fully delete some resources after a restore

[v1.3.2] - 2025-04-03

Fixed

• [#46] Fixed endless loop of reconciles because the image was not set in update manager and difference comparison did not check the cron job provider correctly
• [#46] Fixed missing CronJobs apiGroup in RBACs
• [#46] Add a waiting routine to fix race condition while deleting and restoring resources

Added

• [#47] Add additional print columns and aliases to CRDs

[v1.3.1] - 2024-12-19

Fixed

• [#42] Removed unnecessary rbac proxy to fix CVE-2024-45337

[v1.3.0] - 2024-12-05

Added

• [#40] Add NetworkPolicy to deny all ingress traffic

[v1.2.0] - 2024-11-29

Changed

• [#37] Refactor rbac permissions to be more clear and better match the use cases

Removed

• [#37] Leader election and leader election rbac permissions
• [#37] Metrics rbac permissions

Fixed

• Do not abort restore when maintenance mode cannot be activated

[v1.1.1] - 2024-10-29

Fixed

• [#35] Use correct helm dependency constraint for backup-operator-crd.

[v1.1.0] - 2024-10-28

Changed

• [#33] Make imagePullSecrets configurable via helm values and use ces-container-registries as default.

[v1.0.0] - 2024-10-18

Changed

• [#31] Use cluster native config instead of the etcd.

[v0.11.0] - 2024-09-18

Changed

• [#29] Relicense to AGPL-3.0-only

[v0.10.1] - 2024-01-10

Fixed

• [#27] Added missing watch permission for statefulsets.
  • This is used when waiting for the etcd on maintenance switch.

Changed

• [#14] Updated docs for installing and configuring k8s-longhorn and k8s-velero.

[v0.10.0] - 2023-12-19

Added

• [#23] Added docs for installing the operator in an existent Cloudogu EcoSystem and on an empty cluster.

Fixed

• Fix value of label k8s.cloudogu.com/part-of in helm template do avoid deletion in cleanup process.

[v0.9.0] - 2023-12-08

Added

• [#14] Patch template for mirroring this component and its images

Changed

• [#17] Replace create-backup-script with an operator subcommand. This way, the backup schedule cron job can use the same image as the operator.
• [#15] Delete kustomize structure and hold the operator yaml files just in a helm chart.

[v0.8.0] - 2023-12-04

Added

• [#19] Sync backup list with provider on operator startup

[v0.7.0] - 2023-11-30

Added

• [#13] Sync completed (velero) backups with backup CRs after a restore has finished

[v0.6.0] - 2023-11-22

Added

• [#11] Automated deletion of Backups via various retention strategies.

[v0.5.0] - 2023-11-09

Added

• [#8] Functionality to schedule backups via a BackupSchedule Resource

[v0.4.0] - 2023-10-25

Added

• [#7] Functionality to restore a backup to the namespace where the backup-operator is deployed
  • Before the restore is applied, resources in this namespace which are irrelevant to the backup process are removed to provide a clean slate
  • Currently, only the velero provider is supported

[v0.3.0] - 2023-10-06

Added

• [#3] Functionality to create a backup from the namespace where the backup-operator is deployed
  • Velero is used as a first provider.

[v0.2.0] - 2023-10-05

Added

• [#4] Add CRD-Release to Jenkinsfile

[v0.1.0] - 2023-09-05

Initial release