ClusterAutoscalerNotScalingUp

Troubleshooting hints

This alert pops up when some pods are in pending state, but the ClusterAutoscaler is not able to scale up (either because it wouldn't allow the pod to schedule anyway or because some conditions prevent the cluster-autoscaler from scaling up, e.g. node pool at max capacity, etc...).

There is no silver bullet to resolve this issue, you first need to identify which context you are in: why is the application not schedulable ? (no node matching scheduling conditions ? not enough resources ?)

You can check for the following information to resolve this issue:

• eventrouter logs (LogQL request: {app="eventrouter"} | json | event_source_component="cluster-autoscaler" | event_reason="NotTriggerScaleUp"): displays NotTriggerScaleUp events generated by cluster-autoscaler. Check for involvedObject, reason and message to identify the proper resolution

• cluster-autoscaler status configMap (kubectl get -n kube-system configmap cluster-autoscaler-status -o jsonpath='{.data.status}'): gives information about nodes and node groups

• you can also fetch additional information on the involvedObject using kubernetes CLI (kubectl describe -n <event_involvedObject_namespace> <event_involvedObject_kind> <event_involvedObject_name>)