Skip to content

Releases: libopenstorage/operator

Portworx Enterprise Operator 1.2.0

20 Feb 02:30
Compare
Choose a tag to compare

Updates

  • Added support to manage Prometheus to collect Portworx metrics
  • Added support for heterogeneous node configurations. It is now possible to group nodes using label selectors and override the cluster level configuration for groups of nodes. Storage, network, runtime options, and environment variables can now be overridden in the node-specific configuration.
  • Added an option to disable the default Portworx storage classes through annotation: portworx.io/disable-storage-class.
  • On OpenShift installs, the default Portworx starting port is now 17001 to avoid port conflicts in OpenShift.
  • Updated minimum supported Kubernetes version to 1.12.0.

Bug fixes

  • Fixed broken volume mounts for PKS.

Portworx Enterprise Operator 1.1.1

07 Feb 18:10
Compare
Choose a tag to compare

Updates

  • On fresh OpenShift installs, the default Portworx starting port is now 9011 instead of 9001 due to an internal port conflict with the OpenShift OAuth service.
  • Portworx now uses image pull secrets to pull container images for all components deployed through the Operator.

Portworx Enterprise Operator 1.1.0

14 Nov 03:55
Compare
Choose a tag to compare

Updates

  • You can now expose Portworx metrics to an external Prometheus system.
  • You can now install and manage Portworx Autopilot through the Operator.
  • CRDs created from the Operator adhere to a complete structural schema.

The Operator now does the following:

  • Automatically upgrades Portworx components based on your Portworx version.
  • Exposes its metrics to Prometheus.
  • Creates sample storage classes and makes them available to users in your deployed cluster.
  • Uses a default storage spec for the cluster if you specify none.
  • Uses a StorageNode CRD, rather than a StorageNodeStatus CRD.

Bug fixes

The Operator now does the following:

  • Correctly updates the StorageCluster status during StorageCluster deletion.
  • Continues reconciling storage node pods, even if component deployments experience errors, and logs errors as an event in the StorageCluster object.
  • Allows you to deploy only one StorageCluster in a Kubernetes cluster.

Portworx Enterprise Operator 1.0.6

01 Oct 22:31
Compare
Choose a tag to compare

Updates

  • Add a new stable channel for the Portworx operator
  • Use default images for Portworx (2.1.5), Stork (2.2.5) and Lighthouse (2.0.5) if none specified
  • CSI support for Portworx 2.2+ and Kubernetes 1.14+. Added the resizer and snapshotter CSI sidecars.
  • Update the base docker image to rhel7-atomic:7.6 for security fixes

Bug fixes

  • Add missing Kind and ApiVersion to Stork policy config map
  • On component removal for Portworx driver, delete objects that are only owned by the StorageCluster. If multiple owners present, just remove the ownership

Portworx Enterprise Operator 1.0.5

08 Aug 06:12
Compare
Choose a tag to compare

Updates

  • Added support for managing STORK deployment
spec:
  stork:
    enabled: true
    image: openstorage/stork:2.2.4
    env: 
    - name: FOO
      value: bar
  • Added support for managing Lighthouse for Portworx driver
spec:
  userInterface:
    enabled: true
    image: portworx/px-lighthouse:2.0.4
  • Added support for managing CSI side-cars for Portworx driver
spec:
  featureGates:
    CSI: "true"
  • Improved cluster status and added new CRD StorageNodeStatus to reflect node level status for storage nodes
  • Improved observability of Portworx cluster by enabling status, version and other descriptors in Openshift GUI
  • Added support for using secure KVDB for Portworx driver
spec:
  kvdb:
    authSecret: kvdb-auth-secret
  • Allow passing environment variables to STORK. This also allows enabling Portworx RBAC for STORK.
  • Add annotation (portworx.io/service-type) to overwrite Service type for Portworx and Lighthouse components

Bug fixes

  • Env variables in spec should override the ones added by the operator internally
  • Adding values to empty configurations like env, featureGates, etc was not triggering a restart

Portworx Enterprise Operator 1.0.4

01 Aug 22:45
Compare
Choose a tag to compare

Updates

  • Updated the default portworx image to registry.connect.redhat.com/portworx/openstorage-operator:1.0.4

Portworx Enterprise Operator 1.0.3

08 Aug 05:56
Compare
Choose a tag to compare

Bug fixes

  • Enable PVC controller for Openshift if Portworx is not running in kube-system namespace.