Red Hat OpenShift autoscaling with Cluster Autoscaler

Summary

Red Hat OpenShift supports compute scaling using Cluster Autoscaler integrated with the OpenShift machine API. The mechanism relies on paired ClusterAutoscaler and MachineAutoscaler custom resources to manage cluster-wide limits and per-MachineSet replica boundaries. Scaling decisions evaluate pending pods against node allocatable capacity and hypothetical node capacity annotations while respecting expander strategies like Random, LeastWaste, or Priority. Cluster Autoscaler checks for unnecessary nodes every 10 seconds, scaling down when utilization falls below configured thresholds and workloads can safely migrate. To mitigate provisioning delays, clusters can adopt overprovisioning via low-priority pause containers or use APIs like ProvisioningRequest and CapacityBuffer for predictive scaling.

Takeaways

1. 1

   Cluster Autoscaler requires both ClusterAutoscaler and MachineAutoscaler custom resources; deploying the former alone will not scale the cluster.

2. 2

   Scale-down checks run every 10 seconds and target nodes with utilization below the threshold, excluding nodes hosting local storage, non-controller pods, or restrictive PodDisruptionBudgets.

3. 3

   Overprovisioning reduces pending pod delays by deploying low-priority pause containers (value -10) that production workloads displace via PriorityClass preemption.