Kubernetes Advanced Scheduling
summary
A custom heat-aware scheduler runs as a second scheduler using a kube-scheduler extender (HTTP on localhost, port 8100) and node annotations for boiler temperature; implementation is a small Go extender alongside a vanilla kube-scheduler binary in one pod. A 50-home simulation shows the extender reduces wasted heat and cuts CO2 an extra 10% versus default scheduling; real deployments put ~160 CPUs per home, demonstrated live and scheduled jobs to the coldest boiler until boilers reach capacity.
AI-written summary · about
chapters (12)
- 0:00Intro
- 0:40The old data center
- 3:17Nedda lies
- 4:48Recap
- 5:51Challenges
- 9:14Custom scheduling logic
- 16:42Results
- 18:08Live Demo
- 25:02Baxter Question
- 26:59What can it do for you
- 27:36Wrap up
- 28:32Questions
related talks
01h 29mTutorial: Unleash the Full Potential Of Kubernetes Scheduler: Configuration, Extension And Operation In Production
25mKube-scheduler-evaluator: Evaluating Kubernetes Schedulers at Hyperscaler…
34mWorking your Cluster: Smarter Scheduling Decisions for Your Work...
36mA Toolkit for Simulating Kubernetes Scheduling at Scale
25mScale your Batch / Big Data / AI Workloads Beyond the Kubernetes Scheduler
30mEnhancing the Kubernetes Scheduler for Diverse Workloads in Large Clusters
34mKubernetes Scheduling Features or How Can I Make the System Do What I Want?
27mUnderstanding Kube-Scheduler-Simulator