Running prioritised workload in Kubernetes
summary
Covers Kubernetes workload placement and resource configuration trade-offs: using requests/limits and zero overcommit for CPU/memory versus allowing CPU overcommit, how preemption works, and how PriorityClass objects define pod importance. Introduces the PriorityClasses used in Netic’s Secure Cloud Stack and discusses issues when controllers or Operators create pods that might not inherit the intended priority.
AI-written summary · about
related talks
45mLimits, Requests, QoS, PriorityClasses: Demystifying Scheduling in K8s!
40mLimits, Requests, QoS... Rethinking What You Thought You Knew About Scheduling in Kubernetes
01mKubernetes: What you need to know ... and why
59mKubernetes - going beyond the basics
29mKubernetes Workload Resiliency in Action: Beyond Basics
21mDemystifying Resource Management in Kubernetes
34mKubernetes Scheduling Features or How Can I Make the System Do What I Want?
32mCoordinate Workloads Colocation: QoS-Oriented Scheduling Enhancement on K8s