Kubernetes: Weekly Recap - Gang Scheduling Momentum and Kubelet Cleanup
This week's development centered on building out gang scheduling infrastructure for the scheduler, alongside a wave of kubelet dependency cleanup and a cluster of coordinated cherry picks fixing job controller, kubeadm, and cri-API regressions. Fifty pull requests and two additional commits landed, with SIG Scheduling and SIG Node driving the bulk of the substantive changes.
Duration: PT3M23S
Episode overview
This episode is a short developer briefing from Kubernetes.
It explains recent repository work in plain language.
- Show: Kubernetes
- Published: 2026-07-06T09:09:56Z
- Audio duration: PT3M23S
Transcript excerpt
This excerpt keeps the crawler page concise. Listen to the episode or use the RSS feed for the full update.
This is the Kubernetes Weekly Recap for June 29th through July 6th. Fifty pull request activity items and two additional commits this week.
The clearest pattern: gang scheduling is becoming real infrastructure, not just a proposal. A cluster of PRs from macsko and collaborators built out core plumbing for pod groups in the scheduler — caching and snapshotting pod group objects in PR 140077, tracking incomplete pod groups in PR 139952, and a follow-up…
Second theme: kubelet is shedding external dependencies and cleaning up long-standing debt. PR 140026 moves out-of-memory event detection in-tree, dropping the c-advisor library dependency entirely. PR 139018 removes an open containers cgroups dependency from end-to-end tests. And PR 139713 continues a multi-part…
Third theme: reliability cleanup through cherry picks. Several regressions got coordinated fixes across release branches one-three-four, one-three-five, and one-three-six: the job controller under-reporting active pods during backoff, fixed in PRs 139751, 139642, and 139752; a kubeadm etcd learner promotion bug…
Rounding out the week, DRA continued incremental hardening — a nil-guard for pending allocations in PR 140176, and…
Ne…
Nearby episodes from Kubernetes
- Chasing Down Race Conditions and Stale State
- Kubelet Reliability Sweep
- Cherry-Pick Season and DRA Hardening
- Code Quality and API Stability Push
- Weekly Recap - Validation Framework Overhaul and Stability Fixes
- API Machinery Cleanup and Watch Cache Refactoring
- Platform Reliability and Testing Improvements
- Infrastructure Hardening and Performance Optimizations