Operator 4.3.14

Other Changes

• Added support for setting a Pod’s hostname using setHostnameAsFQDN in the serverPod specification.
• Updated the default WebLogic Monitoring Exporter version to 2.3.14.
• Removed the operator’s legacy Pod startup scheduling heuristic in favor of native Kubernetes scheduling.

Bug Fixes

• Fixed a race condition during rolling restarts of server Pods.
• Improved conflict handling for the V8 domain conversion webhook.

Full Changelog: v4.3.13...v4.3.14

Operator 4.3.13

Other Changes

• Updated jackson version to 2.22.1.
• Added support for schedulingGates in serverPod spec.
• Hardened resource shortage retry logic.

Full Changelog: v4.3.12...v4.3.13

Operator 4.3.12

Other Changes

• MII introspector job failure now includes the WDT command log in the job log.

Bug Fixes

• Fixed an invalid namespace rejection by admission webhook when applying v8 schema.

Full Changelog: v4.3.11...v4.3.12

Operator 4.3.11

Other Changes

• Updated Jackson Databind version to 2.21.4.
• Clarify maxConcurrentStartup description for domain and cluster specification.

Bug Fixes

• Fixed invalid namespace rejection by admission webhook when applying v8 schema.

Full Changelog: v4.3.10...v4.3.11

Operator 4.3.10

Other Changes

• Hardened operator container restarts logic
• Updated supported Kubernetes version to include v1.36.1

Bug Fixes

• Fixed Domain info cached during domain lifecycle decision
• Restored logging properties usage and fixed a problem with ConcurrentModificationException

Full Changelog: v4.3.9...v4.3.10

Operator 4.3.9

Other Changes

• Limited the InitialDomainOnPV creating Persistent Volume use case using hostPath volumes to developer mode only. Default is disabled, set domainOnPV.localDeveloperMode=true in helm chart value to enable it.
• Hardened operator webhook compoenent operation.
• Added troubleshooting notes for using Helm version 4 during operator upgrades with server-side apply.
• WebLogic Monitoring Exporter default version is now 2.3.11.
• Dependency updates for BouncyCastle to version 1.84 and Databind to version 2.21.1.

Bug fixes

• Resolved the missing cause for APIException in the operator webhook during lifecycle operations.

• Corrected an error in the lifecycle scalingAction script.
• Corrected port name generation conflicts and errors.

Full Changelog: v4.3.8...v4.3.9

Operator 4.3.8

Other Changes

• Updated metrics to return active managed namespaces and domains.
• Added support for the HPA-style pod deletion annotation controller.kubernetes.io/pod-deletion-cost.
• Added a Helm chart value apiServerConnectTimeoutSeconds (default: 10s).

Bug Fixes

• Corrected the operator’s pod recycling logic to minimize repeated introspector jobs.
• Fixed a misleading “missing next step” message in the operator log.

Full Changelog: v4.3.7...v4.3.8

Operator 4.3.7

Bug Fixes

• Resolved an issue where the operator was repeatedly incorrectly logging that one or more namespaces were no longer being managed.

Full Changelog: v4.3.6...v4.3.7

Operator 4.3.6

Other Changes

• The Helm chart now enforces that the Kubernetes version is 1.25 or greater.

Bug Fixes

• Resolved an issue where evicted or failed pods for WebLogic Server instances were not promptly deleted to allow for the creation of a replacement pod.

Full Changelog: v4.3.5...v4.3.6

Operator 4.3.5

Other Changes

• The default WebLogic Monitoring Exporter version is now 2.3.8.
• The pod name is now passed to the monitoring exporter sidecar container to allow this value to be used in metrics.

Bug Fixes

• Resolved an issue with the handling of logging-related failures during operator initialization.
• Resolved an issue with monitoring exporter configurations that included excludedKeyValues or includedKeyValues.
• Resolved an issue where the note value for generated Kubernetes events was not properly limited to 1024 characters.

Full Changelog: v4.3.4...v4.3.5