v0.28.0

Description

Release Date: 2026-11-20

SUSE® Rancher Prime Cluster API v0.28.0 is the first version of the v0.28 series, which is shipped as part of Rancher v2.16.

A version check introduced in Cluster API v1.14, causes an issue in CAPA EKS and CAPG GKE clusters that leads to the CAPI object being marked as updating, even when it reaches a healthy, available state. We’re collaborating upstream to have this issue resolved as soon as possible. No clusters outside of EKS and GKE are affected.

Notable Changes

For the first time, SUSE® Rancher Prime Cluster API installs a core CAPI controller created downstream that addresses security vulnerabilities that affect the upstream version.

This new release includes relevant changes that require your attention:

Cluster API Add-on Provider Fleet is now unsupported

See the migration guide in the official docs.

After announcing the deprecation of this add-on provider in Rancher v2.14, support is entirely dropped with v2.16.

  • Provider installation has been removed from rancher-turtles-providers Helm chart.

  • It is no longer tested/validated to work with SUSE® Rancher Prime Cluster API >=v0.28.0.

  • Relevant features of the add-on provider are now integrated in SUSE® Rancher Prime Cluster API:

    • synchronization of labels from CAPI to Fleet clusters.

    • propagation of class variables as template values.

    • cross-namespace class references.

    • If you’re interested on implementation details:

      • feat: add Fleet Controller by @anmazzotti in #2803

      • feat: populate Fleet cluster’s .spec.templateValues by @yiannistri in #2796

Cluster API Provider RKE2 now supports in-place updates via system-agent (experimental/tech-preview)

This feature is experimental/tech-preview and prime-only. It is disabled by default and not recommended to run this in production environments before it is eventually promoted to stable.

  • Update the Kubernetes version of the cluster’s control plane in-place: no Machine is replaced during the operation.

  • system-agent performs the operation on the RKE2 workload cluster.

If you’re planning on testing this feature, you need to explicitly enable it via the rancher-turtles-providers Helm chart. You must use this chart to have the in-place updates runtime extension installed. Refer to the documentation for more information on how this feature works and follow the user guide to enable it.

Refer to the documentation here for more information.

Full Changelog:

Review the full list of changes: v0.27.4…​v0.28.0

Download

Name Created At Updated At

rancher-turtles-0.28.0.tgz

2026-11-20

2026-11-20

Information retrieved from v0.28.0 GitHub release.