Managing Large-Scale Kubernetes Clusters Effectively and Reliably
summary
Covers Ant Financial’s automation patterns for operating Kubernetes at multi‑datacenter scale (10,000+ nodes per DC). Describes automated node lifecycle and failure handling using NPD, autoscalers and a custom Operator, practices for cluster deployment and upgrades, and a Prometheus+Operator based risk prevention and fault‑detection/isolation system.
AI-written summary · about
chapters (13)
- 0:00Intro
- 0:17About Me
- 0:39Agenda
- 2:21Background: Cluster Scale
- 4:41Motivation
- 6:38Design Concept: Imperative
- 8:26Design Concept: Declarative
- 10:13Design Concept: Self-recovery
- 12:15Cluster-Operator
- 13:43Machine-Operator
- 16:08Node Final-state Management
- 18:56Fault Self-recovery
- 24:51Risk Prevention
related talks
37mManaging Large-Scale Kubernetes Clusters Effectively and Reliably
34mA story of managing kubernetes cluster with 15k nodes & various workloads
42mMaking Resilient, Self-Adaptive, And Autonomous Kubernetes Clusters
24mDynamic Big Data in the Cloud with Kubernetes
41mKubernetes the Very Hard Way
35mOrchestrating Resilient Data: Harnessing the Strength of Kubernetes with Operators
15mDemo: Resource(Cost) optimization on planet scale Kubernetes clusters.
22mAnatomy of a production grade Kubernetes cluster