2.27.3¶
Important
For MOSK clusters, Container Cloud 2.27.3 is the first patch release of MOSK 24.2.x series using the patch Cluster release 17.2.3. For the update path of 24.1 and 24.2 series, see MOSK documentation: Cluster update scheme.
The Container Cloud patch release 2.27.3, which is based on the 2.27.0 major release, provides the following updates:
Support for the patch Cluster releases 16.2.3 and 17.2.3 that represents Mirantis OpenStack for Kubernetes (MOSK) patch release 24.2.1.
MKE:
Support for MKE 3.7.12.
Improvements in the MKE benchmark compliance (control ID 5.1.5): analyzed and fixed the majority of failed compliance checks for the following components:
Container Cloud:
iam-keycloak
in thekaas
namespace andopensearch-dashboards
in thestacklight
namespaceMOSK:
opensearch-dashboards
in thestacklight
namespace
Bare metal: update of Ubuntu mirror from ubuntu-2024-07-16-014744 to ubuntu-2024-08-06-014502 along with update of the minor kernel version from 5.15.0-116-generic to 5.15.0-117-generic.
VMware vSphere: suspension of support for cluster deployment, update, and attachment. For details, see Deprecation notes.
Security fixes for CVEs in images.
This patch release also supports the latest major Cluster releases 17.2.0 and 16.2.0. And it does not support greenfield deployments based on deprecated Cluster releases. Use the latest available Cluster release instead.
For main deliverables of the parent Container Cloud release of 2.27.3, refer to 2.27.0.
- Security notes
- Known issues
- Bare metal
- [47202] Inspection error on bare metal hosts after dnsmasq restart
- [46245] Lack of access permissions for HOC and HOCM objects
- [42386] A load balancer service does not obtain the external IP address
- [41305] DHCP responses are lost between dnsmasq and dhcp-relay pods
- [24005] Deletion of a node with ironic Pod is stuck in the Terminating state
- LCM
- [39437] Failure to replace a master node on a Container Cloud cluster
- [31186,34132] Pods get stuck during MariaDB operations
- [30294] Replacement of a master node is stuck on the calico-node Pod start
- [5782] Manager machine fails to be deployed during node replacement
- [5568] The calico-kube-controllers Pod fails to clean up resources
- Ceph
- Bare metal
- Artifacts