0.3.1

Changelog

• Add the ability to configure Bookkeeper options, including metrics, via the new options block (#122)
• Create a new PVC for Bookkeeper Index directory (#135)
• Set BK_useHostNameAsBookieID to false for compatibility with Pravega v0.4. It can be overridden via options when using Pravega >= 0.5 (#135)
• Set default resource requests and limits to pods (#129)
• Expose resource requests and limits to the API to allow custom configurations (#129)
• Automatically set JVM heap size on all components to values based on pod's resource limits (#129)

0.3.0

Changelog

• Add a test mode flag to disable restrictions (#119)
• Temporarily use IP address as Bookie ID (#117)
• Setup liveness and readiness probes for BookKeeper, Segment Store, and Controller pods (#102)
• Make Segment Store wait on startup until all bookies are ready (#102)
• Start Segment Store pods sequentially, one after another, for cleaner startup processes (#102)
• Create Pod Disruption Budget for BookKeeper, Segment Store, and Controller pods (#92)
• Create Pod Anti-Affinity policies to preferably schedule pods of the same component to different nodes (#92)
• Update cluster status with health information (#123)
• Cleanup Pravega metadata when a cluster is deleted (#110)
• Delete PVCs from terminated pods when scaling down the cluster (#107)

0.3.0-rc3

Changelog

• Update cluster status with health information (#123)
• Cleanup Pravega metadata when a cluster is deleted (#110)
• Delete PVCs from terminated pods when scaling down the cluster (#107)

0.3.0-rc2

Changelog

• Add a test mode flag to disable restrictions (#119)
• Temporarily use IP address as Bookie ID (#117)

0.3.0-rc1

Changelog

• Setup liveness and readiness probes for BookKeeper, Segment Store, and Controller pods (#102)
• Make Segment Store wait on startup until all bookies are ready (#102)
• Start Segment Store pods sequentially, one after another, for cleaner startup processes (#102)
• Create Pod Disruption Budget for BookKeeper, Segment Store, and Controller pods (#92)
• Create Pod Anti-Affinity policies to preferably schedule pods of the same component to different nodes (#92)

0.2.2

Changelog

• Set default values when parameters are not present in the CR specification (#89)

0.2.1

Changelog

• Bug fix: Clean up persistent volume claims (PVC) when deleting a PravegaCluster resource (#103)

0.2.0

Changelog

• Update Operator SDK dependency from v0.0.5 to v0.2.0 (#98, #105)
• Set up end-to-end testing framework (based on operator SDK command operator-sdk test local)
• Run end-to-end tests on Travis based on MiniKube

0.1.2

Changelog

• Support external connectivity via LoadBalancer and NodePort services (#77)
• Ability to scale Pravega Controller (#99)

0.1.1

Changelog

• Add operator Helm Chart (#66)
• Add ability to configure non-default service accounts for Controller, Segmentstore, and BookKeeper (#80)
• Fix error handling (#76)