Addressed issues

The following issues have been addressed in the Mirantis Container Cloud release 2.6.0 and the Cluster release 5.13.0:

  • [11302] [LCM] Fixed the issue with inability to delete a Container Cloud project with attached MKE clusters that failed to be cleaned up properly.

  • [11967] [LCM] Added vrrp_script chk_myscript to the Keepalived configuration to prevent issues with VIP (Virtual IP) pointing to a node with broken Kubernetes API.

  • [10491] [LCM] Fixed the issue with kubelet being randomly stuck, for example, after a management cluster upgrade. The fix enables automatic restart of kubelet in case of failures.

  • [7782] [bootstrap] Renamed the SSH key used during bootstrap for every cloud provider from openstack_tmp to an accurate and clear ssh_key.

  • [11927] [StackLight] Fixed the issue with StackLight failing to integrate with an external proxy with authentication handled by a proxy server and ignoring the HTTP Authorization header for basic authentication passed by Prometheus Alertmanager.

  • [11001] [StackLight] Fixed the issue with Patroni pod failing to start and remaining in the CrashLoopBackOff status after the management cluster update.

  • [10829] [IAM] Fixed the issue with the Keycloak pods failing to start during a management cluster bootstrap with the Failed to update database exception in logs.

  • [11468] [BM] Fixed the issue with the persistent volumes (PVs) that are created using local volume provisioner (LVP) not being mounted on the dedicated disk labeled as local-volume and using the root volume instead.

  • [9875] [BM] Fixed the issue with the bootstrap.sh preflight script failing with a timeout waiting for BareMetalHost if KAAS_BM_FULL_PREFLIGHT was enabled.

  • [11633] [vSphere] Fixed the issue with the vSphere-based managed cluster projects failing to be cleaned up because of stale secret(s) related to the RHEL license object(s).