Week Ending January 11, 2026

Developer News

Next Tuesday is the monthly New Contributor Orientation. Join at one of two times to learn all about starting your contribution journey.

Kind will start using kubeadm v1beta4 configuration for Kubernetes 1.36, introducing breaking changes that may require updates to tools, scripts, and configs. Benjamin has asked maintainers to review the v0.31.0 release notes and update their setups accordingly.

Release Schedule

Next Deadline: Release Cycle Begins

The release team leads have been chosen and they are selecting their shadows.

Patch releases have been delayed 1 week; cherry-picks are due this Friday, with patches to come out next Wednesday.

Featured PRs

• 136086: Graduate watch_list_duration_seconds to Beta

This PR graduates the watch_list_duration_seconds metric to Beta. Previously, the metric was Alpha and not guaranteed to remain stable. With this change, the metric’s behavior and labels are now considered stable, making it safe for long-term monitoring, alerting, and observability tooling.

• 136117: Enable declarative-native validation for API strategies

This PR adds utilities that allow API strategy.go files to opt into declarative-native validation. Native declarative validations now run even when the DeclarativeValidation feature gate is disabled, ensuring critical IDL-defined rules are always enforced and preventing validation bypasses due to misconfiguration or panics.

Other Merges

• kubeadm switches tests to NewClientset
• endpoints field in discovery.k8s.io/v1 EndpointSlice is now optional
• Add auditID to watch start logs
• Add logging when WatchList requests complete their initial listing phase
• Scheduler aligns victim metrics semantics between async and sync preemption
• csiDriverClient.nodeGetCapabilities now returns NewUncertainProgressError
• Wire discovery and networking API group for declarative validation and migrate Endpoint.Addresses,EndpointSlice.AddressType and IngressClassParametersReference.Name
• ipvs and winkernel kube-proxy backends are reverted to their pre-1.34 behavior
• Sort runtime handlers list returned by CRI runtime
• Add --concurrent-resourceclaim-syncs to configure kube-controller-manager ResourceClaim reconcile concurrency
• Optimize DRA allocator

Subprojects and Dependency Updates

• Cluster-autoscaler-chart 9.54.1 continues to support scaling Kubernetes worker nodes within autoscaling groups.
• Cluster-api v1.10.10 expands Kubernetes version support up to v1.33.x for both management and workload clusters and includes a Runtime SDK fix improving client certificate and key rotation, with no dependency changes.
• CoreDNS v1.14.0 improves security and reliability with a regex length limit, Kubernetes API rate limiting, cleaner SOA handling, consolidated errors (show_first), enhanced metrics, and key plugin fixes, with additional security updates noted in the advisory.
• CRI-O v1.34.4 updates cosign-installer to v4.0.0, adopts the new cosign bundle format for release notes, fixes exec CPU affinity, improves SELinux label handling for init containers, and ignores /etc/passwd mounts.
• Prometheus v3.9.1 fixes an agent crash shortly after startup caused by an invalid object type and resolves an issue where relabel keep/drop rules were not working correctly.
• Prometheus v3.9.0 makes Native Histograms stable, introduces the scrape_native_histograms setting, adds API and TSDB improvements, enhances PromQL and scraping performance, and includes multiple UI and stability fixes.

Last Week In Kubernetes Development (LWKD) is a product of multiple contributors participating in Kubernetes SIG Contributor Experience. All original content is licensed Creative Commons Share-Alike, although linked content and images may be differently licensed. LWKD does collect some information on readers, see our privacy notice for details.

You may contribute to LWKD by submitting pull requests or issues on the LWKD github repo.