Balance the Cluster Autoscaler Across Zones
Configure the autoscaler with balance-similar-node-groups so a zone outage does not leave all pods in one zone.
Task
Configure --balance-similar-node-groups --expander=least-waste so the cluster enforces the policy at admission time instead of relying on manual review.
When to use this
A developer deploys a pod without --balance-similar-node-groups --expander=least-waste and it disrupts a node, causing evictions on a production cluster.
Prerequisites
- A Kubernetes cluster with kubectl access
- Cluster admin or namespace edit RBAC
Steps
1. Configure --balance-similar-node-groups --expander=least-waste in the project configuration file with the recommended defaults.
--balance-similar-node-groups --expander=least-waste
Expected output: The configuration is accepted and --balance-similar-node-groups --expander=least-waste runs without errors on the first invocation.
2. Run the check locally to verify --balance-similar-node-groups --expander=least-waste produces the expected result before pushing.
pnpm test
Expected output: Local run matches CI behavior; no surprises after push.
3. Add the gate to the CI pipeline so --balance-similar-node-groups --expander=least-waste blocks merge on failure.
- name: Run add cluster autoscaler balance
run: pnpm test
Expected output: Pull requests with failing --balance-similar-node-groups --expander=least-waste are blocked from merging.
Troubleshooting
Problem
The check passes locally but fails in CI
Solution
Pin dependency versions and the runtime to match CI exactly; reproduce with the same Docker image or act.
Problem
Configuration is rejected with a schema error
Solution
Check the tool version matches the config schema; older versions may not support newer fields.
Limitations
- The gate only runs on events that trigger the workflow; locally pushed commits bypass it.
- Configuration drift is possible if someone edits settings outside version control.
Enforce at admission
Start by measuring the problem. Configure the autoscaler with balance-similar-node-groups so a zone outage does not leave all pods in one zone. The most common mistake is adding the gate without first confirming it catches the issue you care about. Run –balance-similar-node-groups –expander=least-waste against the current codebase and inspect the output before wiring it into CI.
--balance-similar-node-groups --expander=least-wasteVerify the policy
Once the gate is in place, make failures actionable. A failing –balance-similar-node-groups –expander=least-waste should point to the exact file and line, not a generic error. Pin the tool version in the lockfile so a upstream release does not silently change behavior. Review the gate output in every pull request just like a test result.