v0.0.9

What's Changed

• Upgrade to latest version of kubebuilder by @dethi in #29

Full Changelog: v0.0.8...v0.0.9

v0.0.8

What's Changed

• controller: fix reconciliation log by @dethi in #27
• main.go: Add pprof server by @tomasbanet in #28
• Add workflow to build and push docker image to Github by @dethi in #30
• workflow/release: use the correct trigger by @dethi in #31

New Contributors

• @dethi made their first contribution in #27
• @tomasbanet made their first contribution in #28

Full Changelog: v0.0.7...v0.0.8

v0.0.7

Bug fix for hbase-operator restarting all managed statefulset pods when only # of replicas have changed:

• Removing unnecessary annotations from new version SHA calculation: "kubectl.kubernetes.io/last-applied-configuration" is updated with every update of the HBase CRD including just replicas count update. It should not be included into pod temlpate metadata to avoid unnesesary pods restart.

v0.0.6

QOL Improvements

• Log in operator log # of regions in transition when RSes are drained
• Improvement in logging message during draining and deletion of RSes

v0.0.5

• Remove "kubectl.kubernetes.io/last-applied-configuration" annotation from statefulset revision SHA calculation. This annotation blob contained the replica count that would force a full cluster restart whenever a replica count was updated via spinnaker and triggerd this annotation to update. Previous attempts at fixing this bug directly edited the CRD and replica count field, which is why this bug was not caught.

v0.0.4

• Remove replica count from revision hash calculation
• Operator and hbase CRD still must be deployed to the same namespace, but can now be configured to deploy to namespaces other than default "hbase"

v0.0.3

• ensureStatefulSetPods() now picks pods in reverse ordinal order to delete/restart

v0.0.2

Includes:

• Bugfix for CRD statefulsets restarting all pods when only replica count has updated. Now the expected behavior is that revision SHA does not change when replica count is updated, which in turn does not trigger all pods to restart. Behavior for downsizing replica count is an unsupported feature.

v0.0.1

Initial release