-
Bug
-
Resolution: Not a Bug
-
Critical
-
None
-
4.15.z
-
Quality / Stability / Reliability
-
False
-
-
None
-
Critical
-
None
-
None
-
None
-
None
-
None
-
None
-
None
-
None
-
None
-
None
-
None
-
None
Description of problem:
Customer reported the nodes are getting cordon automatically in HCP cluster We're running our hcp baremetal cluster is 4.15.46. This morning we had two nodes were in cordon mode (not scheduled): xxxxocp010 xxxxocp007
Additional Details:
We observed the different currentConfig and desiredConfig on the node :machineconfiguration.openshift.io/currentConfig: 751f9f3e <----- machineconfiguration.openshift.io/desiredConfig: fddacc11 <----- machineconfiguration.openshift.io/desiredDrain: drain-fddacc11 machineconfiguration.openshift.io/lastAppliedDrain: drain-fddacc11 machineconfiguration.openshift.io/post-config-action: Rebooting machineconfiguration.openshift.io/reason: "" machineconfiguration.openshift.io/state: Working volumes.kubernetes.io/controller-managed-attach-detach: "true" creationTimestamp: "2025-04-15T12:23:34Z" This indicates that the machine configuration operator is trying to apply a new configuration to the node, but it hasn’t been applied yet, leaving the node in a mismatch state. The desiredDrain value drain-fddacc11 suggests the node is in a drain process and has been flagged for reboot post-config-action: Rebooting, but since the current and desired configs don’t match, the node isn't proceeding with the config application as expected.