SOLO-5072

PodTerminationTimeoutSoloError

Description

Thrown when pods do not terminate within the allotted time; the message names the namespace and the label selector being waited on. solo waits for matching pods to disappear during teardown, so this means they were still present when the deadline passed — for example a pod is stuck terminating or has a finalizer. It is retryable, since termination often completes shortly after.

Troubleshooting Steps

1. List pods still present: kubectl get pods -n -l <labels.join(’,’)>
2. Describe stuck pods for termination events: kubectl describe pod -n -l <labels.join(’,’)>
3. Check for finalizers blocking deletion: kubectl get pod -n -l <labels.join(’,’)> -o jsonpath=’{.items[*].metadata.finalizers}'
4. Force-delete stuck pods if safe: kubectl delete pod -n -l <labels.join(’,’)> –force –grace-period=0
5. Check solo logs for context: tail -n 100 ~/.solo/logs/solo.log