We are pleased to announce Portworx Operator 1.5.0 is now GA. Please see the major improvements and bug fixes below:
Improvements:
- Pure1 telemetry is now enabled by default, starting with Portworx Enterprise 2.8.0.
- The Operator now allows you to specify cache devices during installation.
- The Operator now supports heterogeneous node configurations for cloud storage.
- Added support for passing custom annotations to Portworx pods from StorageCluster.
- Added support for explicitly specifying the cloud provider when using cloud storage.
- The Operator can now overwrite existing Portworx volumes with user-defined custom volumes.
- Simplified the OpenShift console StorageCluster Create form.
- Upgraded CSI driver object to v1.
Bug Fixes:
- The Operator will no longer attempt to install Portworx if a DaemonSet installation already exists.
- The Operator now supports using annotation to overwrite images in “k8s.gcr.io” and other registries with custom image registry. For instance,
operator.libopenstorage.org/common-image-registries: "gcr.io,k8s.gcr.io". - The Operator no longer creates Portworx and KVDB pods repeatedly after a node is cordoned and drained.
- OpenShift no longer fails upgrade operations due to legacy immutable file attributes on config files
/etc/pwx. This issue has been fixed in Portworx 2.6.6 and above. - Portworx pods and other resources no longer randomly get deleted by the Kubernetes garbage collector.
- Changed the default port on AKS to avoid port collision.
Link to release notes here.