Operator 2.1

Features and issues fixed in this release:

Customers can configure init and sidecar containers for generated pods for server instances
Services are now consistently cleaned-up on scale down or when fully stopping a domain
Resolved a timing issue between domain status updates and scaling actions

Operator 2.0.1

Features and issues fixed in this release:

Certified for use on OpenShift.
Fixed readiness probe initial period (was too short and could cause pods to be killed prematurely) and fixed readiness probe address (would not work with load balancers).
Helm archive is available so that the operator can be installed without cloning this repository.
Upgraded samples to use logstash:6.6.0.
Issue #726: Channel ports are now generated for clustered services.
Container ports are now fully populated for the Administration Server.
Operator's REST endpoint private key is now specified using a Kubernetes secret.
Several fixes for configuration override processing.

Operator 2.0

General availability version of Operator 2.0.

Operator 2.0 release candidate 2

Schema for the 2.0 domain custom resource is now final.

Please be aware that domains created with 2.0 release candidate 1 are not compatible with release candidate 2. You will need to delete any preexisting domains from release candidate 1 and regenerate their domain.yaml files.

Many other bug fixes found by our early access testers have been fixed. Thanks for your support!

Operator 2.0 release candidate 1

Operator is now installed via Helm charts, replacing the earlier scripts. The operator now supports domain home on persistent volume or in Docker image use cases, which required a redesign of the Domain schema. Customers can override domain configuration using configuration override templates. Load balancers and Ingress can now be independently configured. WebLogic logs can be directed to a persistent volume or WebLogic server console out can be directed to the pod log. Added lifecycle support for servers and significantly more configurability for generated pods. The final v2.0 release will be initial release where the operator team intends to provide backward compatability as part of future releases.

Release version 1.1

This is a minor, bug-fix release. The following changes are included:

• Improvements to documentation that describe how to scale a WebLogic cluster.
• Add documentation to list steps needed to restart domain when changes have been made to the domain properties.
• Operator resumes processing after Domain resource deleted and recreated.
• Corrected WebLogic cluster instability when startupControl is set to "ALL".
• Expose server name via javaOptions.
• Update Java Kubernetes client to 2.0.0.
• Add validation for Apache loadBalancerVolumePath.
• Upgrade Jackson databinding version to 2.9.6.
• Create headless Services per WebLogic server instance.
• Reduce number of warning messages when reading WebLogic domain configuration.
• Resolve memory continuously growing by preventing the request parameters list from growing indefinitely.
• Document recommendation to use NFS version 3.0 for running WebLogic Server on OCI Container Engine for Kubernetes.
• Add validation of legal DNS names for domainUID, adminServerName, managedServerNameBase, and clusterName.

Release version 1.0

Since release of the Technology Preview version, we have added the following functionality and support in this version of the WebLogic Server Kubernetes Operator and the WebLogic Server Kubernetes certification.

• Support for Kubernetes versions 1.7.5, 1.8.0, 1.9.0, 1.10.0
• We provide instructions in our Operator GitHub project on how to build, test, and publish the Docker image for the operator directly from Oracle Container Pipelines using the wercker.yml .
• Support for Dynamic Clusters, and auto-scaling the WebLogic Server cluster with dynamic clusters. Please read blog for details.
• Support for Apache HTTP Server and Voyager (HAPROXY-backed) Ingress Controller running within the Kubernetes cluster for load balancing HTTP requests across WebLogic Server managed servers running in clustered configurations. Integration with the Operator automates the configuration of these load balancers. Find documentation for Apache HTTP Server and Voyager Ingress Controller.
• Support for Persistent Volumes (PV) in NFS storage for multi node environments. We provide in our project a cheat sheet to configure the NFS volume on OCI, and some important notes about NFS volumes and the WebLogic Server domain in Kubernetes.
• The Delete WebLogic domain resources script, which permanently removes the Kubernetes resources for a domain(s) from a Kubernetes cluster. Please see “Removing a domain” in the README of the Operator project.
• Improved Prometheus support. See Using Prometheus to Automatically Scale WebLogic Clusters on Kubernetes.
• Integration tests posted to our WebLogic Server Kubernetes Operator GitHub project

Release version 0.2

This is a bug fix release and should still be considered a "technology preview" of the operator.

Themes of this release:

• Consistent naming of artifacts from scripts to generated Kubernetes resources
• Improved Domain resource status
• Domain cleanup script
• Plus many bug fixes and performance improvements...

Soon a Docker image matching this release will be available on the Oracle Container Registry as:

container-registry.oracle.com/middleware/weblogic-kubernetes-operator:0.2

Important: Because of the name changes, you will need to redeploy the operator and recreate and pre-existing WebLogic domains. The version 0.2 operator is not yet backwards-compatible. Our plan is to support production-quality backwards-compatibility by our version 1.0 release.