Taming the Beast: Advanced Resource Management With Kubernetes
Talkat KubeCon + CloudNativeCon Europe 2025 · on Friday, 2025-04-04· at 13:45· Level 1 | Hall Entrance N10 | Room F·
summary
Kubernetes 1.30–1.32 adds concrete controls: pod-level resource limits let multiple containers share a single CPU/memory budget (reducing overprovisioning), in-place pod resizing lets CPU be increased live and supports a restart-or-prefer-no-restart policy for shrinks, and node swap (beta, limited-swap mode) permits burst memory to go to swap with caution around performance and security. Live demos show CPU resize succeeds without downtime but memory and swap behavior require careful configuration and cluster admin planning.
AI-written summary · about
chapters (16)
- 0:00Welcome
- 1:44Per-container resource limits struggles
- 2:46Pod resource requests and limits
- 4:31Demo: Pod resource management
- 9:07Stateful workloads challenges
- 10:08In-place Pod resizing
- 11:58Demo: In-place Pod resizing
- 13:24Pod resizing limitations
- 18:39Q&A: Pod resizing adoption
- 19:01Java memory spikes and swap limits
- 20:04Node swap support
- 23:55Demo: Node swap config
- 26:01Swap security and eviction
- 27:05Kubernetes future and workloads
- 29:11Pod grouping proposal
- 34:25Closing
related talks
27mWhat’s in the Box? Resource Management in Kubernetes
29mResize Your Pods w/o Disruptions aka How to Have a Cake and Eat a Cake
38mInside Kubernetes Resource Management (QoS) – Mechanics and Lessons from the Field
29mRight-Size Your Kubernetes Deployments
23mThis is The Way: A Crash Course on the Intricacies of Managing CPUs...
39mStaying in Tune: Optimize Kubernetes for Stability and Utilization
47mOptimizing Kubernetes Resource Requests/Limits for Cost-Efficiency and Latency
35mResource Management WG Deep Dive