You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
[OPERATOR] Support node.machine.sapcloud.io/effective-creation-timeout on MachineDeployment to override shoot configured machineControllerManager.machineCreationTimeout for use by dependency-watchdog by @elankath [#1104]
[OPERATOR] Support new metrics: machine_create_duration_seconds, machine_initialize_duration_seconds, machine_join_duration_seconds,machine_drain_duration_seconds,machine_delete_duration_seconds by @elankath [#1104]
[DEVELOPER] Introduce simulation provider to run integration tests for MCM without an actual cluster by @takoverflow [#1129]
🐛 Bug Fixes
[OPERATOR] Fix for Machine getting stuck with empty phase when finalizer addition exhausts retry budget by @elankath [#1142]
[OPERATOR] Ensure auto-preserve failed machines are ordered by their PreserveExpiryTime in case of deletion due to reduction is mcs.spec.Replica or mcs.spec.autoPreserveFailedMachineMax by @aaronfern [#1138]
[OPERATOR] Patch machine.Spec.Class.Name on machine object with the new machine class's name after it undergoes an inplace update. by @thiyyakat [#1153]
[OPERATOR] Fix preservation reconcile loop to avoid unconditionally uncordoning nodes unrelated to machine preservation, prevent spurious writes on non-preserved machines, and eliminate conflict errors caused by stale machine object comparison in the defer annotation sync. by @gardener-ci-robot [#1118]
[OPERATOR] MCM now deletes the backing node when its machine is force-deleted, and clears finalizers on nodes terminating without a backing machine.
Machine creation status is also modified to distinguish cloud provider failures from node registration failures. by @gagan16k [#1102]
[OPERATOR] Fix drain error drop in preservation flow where a failed drain on a previously-visited preserved node would not cause the reconciler to retry fast enough. by @thiyyakat [#1145]
[OPERATOR] Prevent error while evicting pod X: pod disruption budget X/X is misconfigured and requires zero voluntary evictions when pods are being evicted by @Kumm-Kai [#1127]
[DEVELOPER] Revert changes on machineTerminationFlow which caused a machine to be stuck in Terminating phase by @gagan16k [#1157]
🏃 Others
[OPERATOR] The orphan-VM safety controller no longer deletes the backing VM of a preserved machine. by @gagan16k [#1146]
[OPERATOR] De-couple cordoning and drain in the various operations that involve draining such as updates and machine preservation. by @thiyyakat [#1119]
[DEVELOPER] Added integration tests for machine preservation by @aaronfern [#1124]