Fixed the issue wherein a deployment of a bare metal node with an LVM volume on top of a mdadm-based
raid10failed during provisioning due to insufficient cleanup of RAID devices.
 Fixed the issue wherein the
KubeContainersCPUThrottlingHighalerts for Equinix Metal and AWS deployments raised due to low default deployment limits set for Equinix Metal and AWS controller containers.
 Fixed the issue that caused StackLight endpoints to crash on start with the private key does not match public key error message.
 Fixed the issue that caused constant restarts of the
kaas-exporterpod. Increased the memory for
kaas-exporterrequests and limits.
 Improved the certificate expiration alerts by enhancing the alert severities.
 Fixed the issue wherein variables values in the PostgreSQL Grafana dashboard were not calculated.
 Fixed the issue wherein the Cluster > Health panel showed N/A in the Elasticsearch Grafana dashboard.
 Fixed the issue wherein the cordon-drain states were not removed after unsetting the maintenance mode for a machine.
 Fixed the issue wherein controller pods were killed by OOM after a successful deployment of a management or regional cluster.