Skip to content

Scale to Minimal Replicas

This feature was introduced in Hydrolix version 5.3.

Use scale to minimal to autoscale most components to zero while leaving the cluster available for API calls and the UI. This feature provides extra cost savings by reducing resources used for idle workloads, while still allowing API interactions and scaling up when needed. This setting is off by default.

Enable scale to minimal⚓︎

To enable scale to minimal, add scale_min: true to the spec section of the Hydrolix configuration:

Enable Scale to Minimal
1
2
spec:
  scale_min: true

Active services when scaled to minimal⚓︎

These pods remain active when scale_min: true is set:

  • traefik
  • ui
  • turbine-api
  • turbine-api-worker
  • zookeeper
  • keycloak
  • prometheus
  • validator
  • version

All other pods scale to zero.

Disable scale to minimal⚓︎

To disable scale to minimal, set scale_min: false.

In Hydrolix version 5.11 and later, the scale_min profile allocates adequate memory to init-turbine-api, so disabling scale to minimal doesn't require manual intervention.

In Hydrolix version 5.10 and earlier, setting scale_min: true allocates limited memory to init-turbine-api. This can cause CrashLoopBackOff errors. When reverting to scale_min: false, the operator doesn't trigger a deployment rollout, so turbine-api may remain in CrashLoopBackOff. If this happens, delete the deployment to force recreation:

Force Turbine API Recreation
1
kubectl delete deployment turbine-api -n <namespace>

The operator recreates the deployment with the correct memory settings from the configuration spec. To set the memory explicitly in the spec:

Override Turbine API Memory
1
2
3
4
spec:
  scale:
    turbine-api:
      memory: "6Gi"