Addressed issues

The following issues have been addressed in the MOSK 25.1 release:

OpenStack

  • [43058] Resolved the issue with the Cronjob for MariaDB that prevented it from transitioning to the APPLYING state after changing the OpenStackDeployment custom resource.

  • [47269] Resolved the issue that prevented instances from live-migrating.

  • [48890] Resolved the issue that caused an extremely high load on the gateway nodes.

Bare metal

  • [49678] Resolved the issue that caused the flapping status (Configure → Ready → Configure → Ready) of machines where any HostOSConfiguration object was targeted and migration to containerd was applied.

Update

  • [49078] Resolved the issue that caused migration to containerd to get stuck due to orphaned Docker containers.

StackLight

  • [49340] Resolved the issue that caused failure of tag-based log filtering using the tag_include parameter for logging.externalOutputs when output_kind: audit is selected.

  • [45215] Resolved the performance issue in the OpenStack PortProber Grafana dashboard when handling large amounts of metrics that caused time ranges to exceed one hour.

    Implemented recording rules and updated the dashboard to leverage them, resulting in significant performance improvements. But be aware that the updated dashboard will only display data collected after cluster update.

    To access older data, use the OpenStack PortProber [Deprecated] dashboard that will be removed in one of the following releases due to being unreliable when querying extended time ranges in high-load clusters.

  • [42660] Resolved the issue that caused the Nova - Hypervisor Overview Grafana dashboard to display the load average (per vCPU), allocated memory, and allocated disk (allocated by VMs) instead of real CPU, memory, and disk utilization with data collected from node-exporter and OpenStack Nova.

  • [39368] Resolved the issue that caused the DockerSwarmNodeFlapping to be firing during cluster update.

    It is expected to see the DockerSwarmNodeFlapping and DockerSwarmServiceReplicasFlapping alerts firing during cluster update to Container Cloud 2.29.0 but only before the StackLight component is updated.

  • [39077] Resolved the issue that caused the TelegrafGatherErrors for telegraf-docker-swarm to be firing during cluster update.

    Reworked the TelegrafGatherErrors alert and replaced it with TelegrafSMARTGatherErrors and TelegrafDockerSwarmGatherErrors alerts.

    It is expected to see the TelegrafGatherErrors alert firing during cluster update to Container Cloud 2.29.0 but only before the StackLight component is updated.