Platform Changelog

feedbefore 2026-10-08all releases, patches included

2026-10-0625 days ago

GitLab Transcend: Speed you can trust, all the way to productionannouncementGitLab announced more than a dozen features at its Transcend event, including goal-driven agent flows, custom flows, an MCP Server, Artifact Central in beta, Dependency Firewall in early access, and Secrets Manager set to reach GA this month. The vendor says GitLab Orbit context graph reduces agent retries by up to 45x and token use by up to 4.5x during its beta. Credit and usage controls add spend caps and alerts at the subscription, group, and user level.platform teamsPlatform teams evaluating GitLab can plan to govern agent actions and MCP access under existing policy, set AI spend caps, and route packages through Artifact Central and Dependency Firewall. Many capabilities are GA later or in beta or early access, so rollout timing matters.GitLab Blog ↗GitLab Jenkins GitHub Actionswritten by claude-haiku-5-5, prompt 7cbe154, checked by publish rules

2026-10-0526 days ago

Scaling Kubernetes Workloads with Node SwappostThe Kubernetes blog reports benchmarks of node swap backed by Local NVMe SSDs across a Linux kernel build, headless Chrome sandboxes under runc, gVisor and Kata, and Python sandboxes. Node swap reached General Availability in Kubernetes v1.34 and is configured through the kubelet with LimitedSwap. The post reports pod density gains of up to 3x in these tests, with the latency cost mainly attributed to CPU competition at peak density.platform teamsPlatform teams running memory-heavy CI or agent sandbox workloads can evaluate swap on Local SSD nodes to raise pod density, but must set workloads to Burstable QoS and size swap for burst rather than steady-state use. The results come from Google Cloud test nodes and the post does not give independent methodology beyond linked repositories.Kubernetes Blog ↗Amazon EKSwritten by claude-haiku-5-5, prompt 7cbe154, checked by publish rules

2026-10-0229 days ago

Amazon EKS and Amazon EKS Distro now support Kubernetes version 1.37announcementAWS announced that Amazon EKS and Amazon EKS Distro now support Kubernetes version 1.37 in all EKS regions, including the GovCloud (US) Regions. Clusters can be created on 1.37 or upgraded through the EKS console, eksctl, or infrastructure-as-code tools. The source highlights Metrics API promotion to GA, DRA device taints and tolerations reaching GA, and Horizontal Pod Autoscaler scale-to-zero entering beta and enabled by default.platform teamsPlatform teams running EKS can plan upgrades to 1.37 and can use EKS cluster insights to check for issues before upgrading. Autoscalers with minReplicas: 0 using object or external metrics will scale idle workloads to zero by default, which teams should account for in their scaling policies.What's New with AWS ↗Amazon EKSwritten by claude-haiku-5-5, prompt 7cbe154, checked by publish rules

2026-09-291 month ago

Open Policy Agentv1.21.1patch

2026-09-281 month ago

Fix pod distribution drift in Amazon EKS with the Kubernetes deschedulerpostThe AWS Containers Blog describes how pods spread across three Availability Zones with soft topology spread constraints can drift into two zones after a node-availability gap and stay skewed. In a measured run on Amazon EKS with a 1,000-pod fleet, the post reports the skew persisted for over 16 hours after capacity returned. It then shows the Kubernetes descheduler's RemovePodsViolatingTopologySpreadConstraint plugin, configured to act on ScheduleAnyway constraints, restoring balance in about 183 evictions with PDB limits applied.platform teamsPlatform teams running soft topology spread on EKS should expect silent zone drift after node disruptions and can use the descheduler with ScheduleAnyway enabled, topologyBalanceNodeFit, and PDBs to correct it. Teams needing strict zone balance should consider DoNotSchedule and alert on pending pods instead.AWS Containers Blog ↗Amazon EKSwritten by claude-haiku-5-5, prompt 7cbe154, checked by publish rules

2026-09-241 month ago

Open Policy Agentv1.21.0minor++++~~~~~~~-******

2026-09-211 month ago

Kubernetes v1.37: Tracking When a PersistentVolumeClaim Was Last Used (Beta)postKubernetes v1.37 promotes the PersistentVolumeClaimUnusedSinceTime feature gate to Beta and enables it by default. The PVC protection controller now sets an Unused condition on each PVC, with status True and reason NoPodsUsingPVC when no non-terminated pod references it. The condition's lastTransitionTime records when the PVC became idle.platform teamsPlatform teams can list idle PVCs with kubectl and jq instead of cross-referencing pods and volumes with custom scripts, which helps find orphaned storage that drives cloud costs. The feature is on by default in v1.37, and the project intends to reach GA in a later release.Kubernetes Blog ↗editorCorrection 2026-10-20: this was listed as a Amazon EKS v1.37 release by mistake. It is Kubernetes Blog coverage, not release notes from the project.written by claude-haiku-5-5, prompt 7cbe154, checked by publish rules

2026-09-171 month ago

Rate limits on GitLab.com are changingannouncementGitLab.com is changing its API rate limits so they depend on subscription tier. Free accounts and unauthenticated requests are limited starting October 19, 2026, with preview windows on October 7 and October 14, and Premium and Ultimate limits change in January 2027. Unauthenticated requests are capped at 60 per hour per IP address, and exceeding a limit returns HTTP 429 with a Retry-After header.platform teamsPlatform teams whose pipelines, bots, or agents call the GitLab.com API without credentials will hit the 60 requests per hour anonymous cap and should move calls to personal access, OAuth, or job tokens. Teams should check RateLimit-Remaining headers against their plan's published limits before the October 19 change, and self-managed and Dedicated instances are not affected.GitLab Blog ↗GitLabwritten by claude-haiku-5-5, prompt 7cbe154, checked by publish rules

See who spent your AI credits and set fair caps per teamannouncementGitLab 19.4 adds per-user management of GitLab Credits, letting admins set a flat per-user cap, add overrides for individuals, and see spend. Admins can also export per-event usage data covering up to 31 days, delivered by email as a summary, per-event file and manifest. Usage syncs periodically, so spend already in flight can cross a cap before enforcement catches up.platform teamsPlatform teams can allocate AI agent budgets per team and per developer, map namespace IDs to cost centers for showback, and pull session-level records for investigating spikes. Teams should account for the periodic sync delay when setting caps.GitLab Blog ↗GitLabwritten by claude-haiku-5-5, prompt 7cbe154, checked by publish rules

2026-09-151 month ago

Kubernetes v1.37: Pod-Level Resource Managers graduated to BetapostKubernetes v1.37 graduates the Pod-Level Resource Managers feature to Beta, disabled by default and controlled by the PodLevelResourceManagers feature gate. The Topology Manager, CPU Manager and Memory Manager can now use pod-level resource declarations when making hardware placement decisions. The v1 PodResources gRPC service also adds top-level cpu_ids and memory fields for pod-level exclusive assignments.platform teamsPlatform teams running latency-sensitive workloads can reserve exclusive NUMA-aligned CPU and memory for primary containers while placing non-Guaranteed sidecars in a pod-isolated shared pool. Teams that opt in should check monitoring and device plugins that read PodResources, since they can now query pod-level assignments without double-counting containers.Kubernetes Blog ↗editorCorrection 2026-10-20: this was listed as a Amazon EKS v1.37 release by mistake. It is Kubernetes Blog coverage, not release notes from the project.written by claude-haiku-5-5, prompt 7cbe154, checked by publish rules

2026-09-101 month ago

Kubernetes v1.37: Scheduler Preemption for In-Place Pod Resize (Alpha)postKubernetes v1.37 adds an Alpha feature behind the InPlacePodVerticalScalingSchedulerPreemption feature gate that lets the kube-scheduler preempt lower-priority Pods on the same node so that a running Pod's deferred in-place resize can be actuated. Previously a Deferred resize could remain pending indefinitely until node capacity freed up. The Kubelet delegates this preemption decision to the scheduler rather than evicting locally.platform teamsPlatform teams bin-packing lower-priority batch workloads can allow higher-priority Pods to scale up in place without restarts, but must enable the gate across kube-apiserver, kube-scheduler and kubelet on v1.37 clusters. The feature is Alpha, and the new spec.podPreemptionPolicy node field lets controllers opt specific nodes out of resize preemption.Kubernetes Blog ↗editorCorrection 2026-10-20: this was listed as a Amazon EKS v1.37 release by mistake. It is Kubernetes Blog coverage, not release notes from the project.written by claude-haiku-5-5, prompt 7cbe154, checked by publish rules

Kyvernov1.19.1patchsecurity++**************!!

2026-09-081 month ago

Kubernetes v1.37: Advancing Workload-Aware SchedulingpostKubernetes v1.37 graduates the core Workload and PodGroup APIs, which enable gang scheduling, to Beta, along with workload-aware preemption and shared DRA ResourceClaims for PodGroups. The release adds a CompositePodGroup API for multi-level hierarchical scheduling, and makes minCount mutable so controllers can resize gangs. The alpha v1alpha2 API was replaced by v1alpha3, and the disruptionMode values were renamed from PodGroup and Pod to all and single.platform teamsPlatform teams running batch or AI/ML workloads on Kubernetes should plan for the v1alpha2 to v1alpha3 breaking change and the disruptionMode rename if they use the alpha APIs. Teams can also begin evaluating native gang scheduling and hierarchical scheduling for JobSet and LeaderWorkerSet workloads.Kubernetes Blog ↗written by claude-haiku-5-5, prompt 7cbe154, checked by publish rules

2026-09-041 month ago

Kubernetes v1.37: KubeletInUserNamespace (aka Rootless mode) Graduates to BetapostKubernetes v1.37 promotes the KubeletInUserNamespace feature gate to beta, allowing kubelet, container runtimes, CNI plugins and kube-proxy to run as a non-root user inside a Linux user namespace. The feature gate is now enabled by default, but nodes only run rootless if they are set up that way outside Kubernetes. Kubelet now reports a runningInUserNamespace property on nodes.platform teamsPlatform teams can run node components without host root to limit container-breakout impact, but some CNI and CSI drivers may break, so test them first. Administrators can use the new runningInUserNamespace node property to keep root-dependent workloads off rootless nodes.Kubernetes Blog ↗editorCorrection 2026-10-20: this was listed as a Kubernetes MCP Server v1.37 release by mistake. It is Kubernetes Blog coverage, not release notes from the project.written by claude-haiku-5-5, prompt 7cbe154, checked by publish rules

2026-09-031 month ago

Open Policy Agentv1.20.2patch

2026-09-021 month ago

Kubernetes v1.37: Scale Workloads to Zero with HorizontalPodAutoscalerpostKubernetes v1.37 moves horizontal autoscaling to zero replicas to Beta and enables it by default. An HPA using an object or external metric can now scale a workload to zero and bring it back when the metric changes, a capability that previously required an add-on or the Alpha feature gate. The blog notes that Services do not buffer requests, so request-driven workloads need a separate buffering layer.platform teamsPlatform teams can scale idle queue consumers and batch workers to zero to cut costs on expensive CPU or GPU Pods, but must accept cold-start latency and make sure the external metrics adapter is reliable. Before upgrading, control plane components must both have the feature enabled, and HPAs with minReplicas 0 need an object or external metric.Kubernetes Blog ↗editorCorrection 2026-10-20: this was listed as a Amazon EKS v1.37 release by mistake. It is Kubernetes Blog coverage, not release notes from the project.written by claude-haiku-5-5, prompt 7cbe154, checked by publish rules

2026-08-282 months ago

Open Policy Agentv1.20.1patch

2026-08-272 months ago

Open Policy Agentv1.20.0minor++++++++~~-********

2026-08-212 months ago

Karpenterv1.14.1patch(LTS)

2026-08-202 months ago

Kyvernov1.19.0minorsecurity++++-*************!

CloudBees Smart Tests Helps Leading Software Company Double Release Velocity by Saving 40K Testing Hours in One Yearcase studyCloudBees published a case study about an unnamed leading software company that adopted its AI-powered Smart Tests. The vendor reports a doubling of annual release velocity, 40,000 saved test execution hours, and an 80% cut in cloud costs. The source gives no methodology or baseline for these figures.platform teamsPlatform teams evaluating test selection in CI pipelines may find this relevant, but the figures are vendor-reported and unverified. The source does not describe how the tool works or how the results were measured.CloudBees Blog ↗CloudBeeswritten by claude-haiku-5-5, prompt 7cbe154, checked by publish rules

2026-08-173 months ago

Open Policy Agentv1.19.1patch

2026-07-303 months ago

Open Policy Agentv1.19.0minor+++~~~~~**********

2026-07-174 months ago

Karpenterv1.11.3patch

Karpenterv1.10.2patch

Karpenterv1.9.2patch(LTS)