K8s Pod Scheduling - Deep Dive
summary
Detailed, example-rich walk through how Kubernetes schedules by label and topology, showing live demos on a three-node us-west cluster where the autoscaler launches a c5.9xlarge and pods move from Pending to Running. Covers nodeSelector/nodeAffinity for instance types, podAffinity/podAntiAffinity with topologyKey for host vs AZ, the concrete difference between required vs preferred rules and deadlock risk, plus taints/tolerations, PVC locality and GPU node labeling. All manifests and demo code are on GitHub and the examples run as shown.
AI-written summary · about
chapters (6)
- 0:00Start
- 11:16Multiple Replicas Instances Zone
- 15:40Demo
- 16:07Node Selector
- 18:19Pod Affinity
- 19:25Pod Anti Affinity
related talks
52mKubernetes - going beyond the basics
01h 01mKubernetes – going beyond the basics
59mKubernetes - going beyond the basics
50mKubernetes - going beyond the basics
34mKubernetes Scheduling Features or How Can I Make the System Do What I Want?
36mDeep Dive: SIG Scheduling
20mHow Scheduling in Kubernetes Works
01h 29mTutorial: Unleash the Full Potential Of Kubernetes Scheduler: Configuration, Extension And Operation In Production