← Back to the front page
- The Admin Console started (
server.start) event in the audit log now includes the Replicated version.
- Replicated Embedded Kubernetes will now set the storageClassName of the audit log’s Persistent Volume Claim.
- Application config is now included in the support bundle at
replicated/internal/app-config.json. Password fields are redacted.
info level logging has been added when a Replicated Embedded Kubernetes application is starting up.
- The command
replicated app <id> rm has been removed.
- Replicated Embedded Kubernetes will now mark its StorageClass as the default StorageClass for the cluster by setting the storageclass.kubernetes.io/is-default-class annotation.
- Fixed an issue that prevented upgrading an application alongside an auto-update of Replicated when that spec was incompatible with the previous Replicated version due to unsupported template functions.
- Fixed an issue that could cause airgapped updates to fail due to application YAML corruption when running with the Swarm or Kubernetes schedulers.
- Fixed an issue that caused an error if a private image was being used in a Programmable Preflight Check in Swarm.
- Fixed an issue that could cause settings to revert when syncing a license or updating an application when there is a replicated.conf
ImportSettingsFrom file present. Replicated can be reverted to the previous behavior by setting
ForceUseImportedSettings to true in the replicated.conf file.
- When a custom TLS certificate/key cannot be parsed successfully, a new certificate and key will be generated automatically to avoid crashing Replicated.
- CPU performance has been improved when running with the Native scheduler.
- Replicated Embedded Kubernetes will now properly wait for StatefulSets when reporting application state.
- Fixed an issue that prevents saving the application settings form when a password field is hidden from a
when condition and a password manager is enabled in the browser.
- Config groups will no longer be visible in the left navigation when only hidden items are present.
- Fixed an issue with Kubelet on worker nodes of a Replicated Embedded Kubernetes cluster failing to detect Rook plugins. The node join script now has a
wait-for-rook flag that will block until Rook is ready and then restart the Kubelet. The Replicated console will include this flag in its generated scripts when Rook is detected.
- Fixed an issue that prevented restores with Replicated Embedded Kubernetes from completing when there were no PVCs specified in the backup.
- Fixed handling of leading slashes in shared filesystem snapshot paths specified in
- Fixed a confusing log line where Replicated Embedded Kubernetes claims to restore more PVCs than it actually does.
- The command
replicatedctl app-config set will now result in an error when attempting to set a non-existent config item.
- Fixed a misleading error message in the support bundle errors.json file, when the command os.read-file was run and the file did not exist.
- Fixed an issue that could cause an error when running Programmable Preflight Checks or Test Procedures multiple times in succession when using Replicated Embedded Kubernetes.
- Fixed an error in the browser console that prevented the user from toggling on the Docker CLI option in the node add modal when using the Swarm scheduler.
- Kubernetes resource type PodDisruptionBudget will now be removed and replaced when applying changes to the Kubernetes application spec.
- Fixed an issue that could prevent the Preflight Checks page from loading in the Admin Console when no Operators are connected when using the Native scheduler.
- Added a meaningful error message when no applications are found associated with the installed scheduler.
- Added a meaningful error message when no channels are associated with a license.
- All known CVEs with fixes have been patched in all images distributed by Replicated at the time of the release. For more information see this article.