github scylladb/scylla-operator v1.15.0-rc.0

latest release: v1.15.0-rc.1
pre-release2 days ago

Release notes for 1.15.0-rc.0

Container images

docker.io/scylladb/scylla-operator:1.15.0-rc.0

Changes By Kind (since 1.14.0)

Feature

  • Aggregate NodeConfig status conditions (#1918,@rzetelskik)
  • Migrate controllers to reconcile v1alpha1.ScyllaDBDatacenter (#2066,@zimnx)
  • Set up S3 credentials in e2e machinery (#2115,@rzetelskik)
  • Update dependencies (#2116,@tnozicka)
  • Propagate mount unit failures and stop reconciling preexisting units unmanaged by NodeConfig (#2134,@rzetelskik)
  • Move resolution of developerMode and additionalScyllaDBArguments to Operator (#2137,@zimnx)
  • Remove support for deprecated cpuset field (#2139,@zimnx)
  • Update ScyllaDB Monitoring to 4.8.1 and use versioned platform dashboards (#2144,@tnozicka)
  • Set open file limit for ScyllaDB processes (#2160,@zimnx)
  • Extend node tune controller with status conditions (#2165,@rzetelskik)
  • Bump ScyllaDB utils image to latest stable (#2176,@zimnx)
  • Add an E2E test for NodeConfig mount degraded condition propagation for mounts over corrupted xfs filesystems (#2178,@rzetelskik)
  • Add ClusterDomain fields into ScyllaOperatorConfig Status (#2186,@zimnx)
  • Bump monitoring stack to latest stable versions (#2203,@zimnx)
  • Update Go dependencies (#2210,@zimnx)
  • Propagate operator loglevel (#2252,@tnozicka)

Bug

  • Make 'shared' volume mount read-write for Scylla container (#2112,@rzetelskik)
  • Fix systemd skip in unit tests (#2117,@tnozicka)
  • Fix ScyllaDBDatacenter controller selector and resource naming (#2129,@zimnx)
  • Set managed hash labels in manager tasks and only update tasks when managed hash changes (#2142,@rzetelskik)
  • Fix manager tasks' status errors not being cleared on failed status updates for tasks present in manager state (#2143,@rzetelskik)
  • Don't account our Jobs into PodDisruptionBudget (#2147,@zimnx)
  • Fix storageClassName field location in monitoring example (#2151,@tnozicka)
  • Use Scylla Manager cluster labels for cluster reconciliation (#2156,@rzetelskik)
  • Fix prunning of tuning related Roles and RoleBindings (#2168,@zimnx)
  • Use a portable method to get systemd unit statuses (#2214,@rzetelskik)
  • Fix cert-manager URL in deploy release CI script (#2220,@rzetelskik)
  • Explicitly disable developerMode when it's not set in the API object (#2242,@zimnx)
  • Fix graceful termination, early ignition and related tests (#2250,@tnozicka)

Documentation

Flaky/Failing Test

  • Fix nonreliable method of validating shardawareness in e2e's (#2175,@zimnx)
  • Do not set ScyllaCluster's ManagerID to empty string when calculating status (#2177,@rzetelskik)

Other


Don't miss a new scylla-operator release

NewReleases is sending notifications on new releases.