Back to all scenarios
Scenario #436
Scaling & Load
Kubernetes v1.25, GCP

Cluster Autoscaler Inconsistencies with Node Pools

Cluster Autoscaler failed to trigger due to node pool constraints.

Find this helpful?
What Happened

Nodes were not scaled when needed because Cluster Autoscaler couldn’t add resources due to predefined node pool limits.

Diagnosis Steps
  • 1Examined autoscaler logs, revealing node pool size limits were blocking node creation.
  • 2Cluster metrics confirmed high CPU usage but no new nodes were provisioned.
Root Cause

Cluster Autoscaler misconfigured node pool limits.

Fix/Workaround
• Increased node pool size limits to allow autoscaling.
• Adjusted autoscaler settings to better handle resource spikes.
Lessons Learned

Autoscaling requires proper configuration of node pools.

How to Avoid
  • 1Ensure that node pool limits are set high enough for scaling.
  • 2Monitor autoscaler logs to catch issues early.
Previous ScenarioNext Scenario