Best Practice for Interference Detection and Resource Isolation Enhancement on...

on YouTube ↗28m

" summary

Describes a Kubernetes-based solution used at Kuaishou to detect and diagnose resource interference between latency-sensitive services and batch jobs, using an observation and diagnostic system for rapid identification and troubleshooting. Covers implementation of fine-grained per-service CPU and memory controls to isolate resources and reduce batch-job impact so the platform can safely increase pod density and overall resource utilization.

AI-written summary · about

" related talks

• 32mCoordinate Workloads Colocation: QoS-Oriented Scheduling Enhancement on K8sTao Li, Zuowei Zhang · Kubernetes Batch + HPC Day North America 2022
• 37mCo-Location of CPU and GPU Workloads with High Resource EfficiencyPenghao Cen, Jian He · KubeCon + CloudNativeCon China 2019
• 38mDynamic Pod Resource Boundary Adjustment in Web Scale ClustersCheng Wang, Xiaoyu Zhang · KubeCon + CloudNativeCon + Open Source Summit China 2019
• 29mKubernetes Workload Resiliency in Action: Beyond BasicsNabarun Pal, Akhil Mohan · KubeCon + CloudNativeCon India 2026
• 20mA full-scenario colocation of workloads based on KubernetesDongdong Chen, Lingpeng Chen · KubeCon + CloudNativeCon + Open Source Summit China 2021
• 47mOptimizing Kubernetes Resource Requests/Limits for Cost-Efficiency and LatencyHenning Jacobs · HighLoad++ Moscow 2018
• 38mDynamic Pod Resource Boundary Adjustment in Web Scale ClustersCheng Wang, Xiaoyu Zhang · KubeCon + CloudNativeCon China 2019
• 23mPractice of Fine-grained Cgroups Resources Scheduling in KubernetesQingcan Wang, Xianlu Chen · KubeCon + CloudNativeCon North America 2020 Virtual