Loading...
Loading...
After a Veeam Kasten Disaster Recovery (KDR) restore is performed on an instance of Veeam Kasten for Kubernetes running a version earlier than 8.5.4, and that instance is later upgraded to version 8.5.4 or newer, export and backup actions for the restored applications fail with an error similar to the following: invalid repository password The restore points imported by the KDR restore are present in the catalog, but any policy or export action that references them afterward fails.
This issue occurs because Veeam Kasten for Kubernetes versions earlier than 8.5.4 import application restore points into the local catalog without correctly linking the encryption key artifact (EKA) to the corresponding repository artifact. The catalog entry appears complete, but the repository password derived from it does not match, so any later operation that requires decrypting or reconnecting to that repository fails. The binding behavior was corrected in version 8.5.4. That correction applies to restore points imported after the upgrade, so restore points that were already imported under the earlier binding remain broken in the catalog.
Starting in Veeam Kasten for Kubernetes 9.0.5, the KastenDRRestore resource supports the forceReimportAppRestorePoints option. When this option is set to true, every application restore point being imported is reprocessed and its repository artifact is rebound to a valid encryption key, including restore points that were already imported successfully by an earlier KDR restore. The repair is applied by running a new KDR restore with this option enabled. The option is a field on the KastenDRRestore resource, so the repair restore is performed using the CLI or API method, regardless of how the original KDR restore was performed. Confirm that the Veeam Kasten for Kubernetes instance is running version 9.0.5 or later. Edit or create the KastenDRRestore resource for the affected restore, and set forceReimportAppRestorePoints to true:
Note: Do not enable forceReimportAppRestorePoints for a KDR restore performed on a clean installation, or for the first KDR restore performed on a cluster. In those cases no existing bindings are present to repair, so the option is unnecessary. Veeam Kasten Documentation > Operating Veeam Kasten > Veeam Kasten Disaster Recovery Veeam Kasten Documentation > API > KastenDR Resources
Click on a version to see all relevant bugs
Veeam Integration
Learn more about where this data comes from
BugZero Plan
Streamline upgrades with automated vendor bug scrubs
BugZero Prevent
Wish you caught this bug sooner? Get proactive today.