Dynamic Pod Resource Boundary Adjustment in Web Scale Clusters
summary
Describes Alibaba's Kubernetes-based approach for co-locating Pods with different QoS classes on the same nodes and dynamically adjusting Pod resource limits (especially under contention) to balance resource efficiency and application SLOs. Presents production results showing 14–30% better cluster utilization, 76–87% reduction in 95th‑percentile tail latency, and 107–163% TPS improvements.
AI-written summary · about
related talks
38mDynamic Pod Resource Boundary Adjustment in Web Scale Clusters
05mResource Roulette: Winning the Kubernetes Allocation Game
23mOptimizing Storage Assignment via Pod Scheduling Under Disturbance Factors
29mRight-Size Your Kubernetes Deployments
23mPractice of Fine-grained Cgroups Resources Scheduling in Kubernetes
15mDemo: Resource(Cost) optimization on planet scale Kubernetes clusters.
36mTaming the Beast: Advanced Resource Management With Kubernetes
28mScaling in Kubernetes Safely on On-Prem KaaS Across 1,300+ Clusters and 40,000+ Nodes