Kubernetes v1.36 introduces major architectural improvements to workload-aware scheduling.
- •Workload and PodGroup APIs are separated: Workload acts as a static template while PodGroup manages runtime state
- •New PodGroup scheduling cycle enables atomic group evaluation, preventing scheduling deadlocks
- •Topology-aware scheduling co-locates Pods within specific physical or logical domains to reduce network latency
- •Gang scheduling ensures strict all-or-nothing placement of interdependent workloads
- •ResourceClaim support enables Dynamic Resource Allocation for PodGroups
This summary was automatically generated by AI based on the original article and may not be fully accurate.
Kubernetes
Grafana
Google Cloud
Vercel