Resource Roulette: Winning the Kubernetes Allocation Game
Lightningat KubeCon + CloudNativeCon Europe 2025 · on Thursday, 2025-04-03· at 17:00· Level 0 | ICC Auditorium·
summary
Static requests and limits cause either wasted CPU reserved but unused or cluster-wide contention when real use exceeds them; adaptive allocation that divides wide usage intervals into smaller intervals and predicts per-interval requests narrows that gap. Examples include the open-source Vertical Pod Autoscaler and commercial products like StormForge and PepperScale that sample past performance to set requests; market consolidation via recent acquisitions is accelerating adoption. Recommendation: always set requests and continuously adjust them with autoscaling tooling.
AI-written summary · about
related talks
29mRight-Size Your Kubernetes Deployments
38mDynamic Pod Resource Boundary Adjustment in Web Scale Clusters
38mDynamic Pod Resource Boundary Adjustment in Web Scale Clusters
31mFailing to Autoscale?
47mOptimizing Kubernetes Resource Requests/Limits for Cost-Efficiency and Latency
26mResource as You Wish: Collaborative Reservation and Allocation b...
36mCapacity Scheduling for Elastic Resource Sharing in Kubernetes
33mThe Big Kubernetes Mistake Paying More Than Necessary