Troubleshooting data lake issues
Use the information in this topic to troubleshoot issues related to the data lake:
The Redis data is not cleaned up completely
Symptom
When you run the Redis cleanup command, it does not clean all the Redis data and PVCs.
Scope
While performing the Redis data clean-up steps, a prompt is displayed to confirm the deletion of PVCs allotted for Redis data. If you enter a value other than y, Y, n, or N, the cleanup fails to clean up the Redis data.
Resolution
Use the following commands to delete the Redis data manually:
- To delete PVCs for Redis Cluster:
kubectl -n ${NAMESPACE} get pvc --no-headers -l app.kubernetes.io/instance=redis-cluster 2>/dev/null | awk '{print $1}' | xargs kubectl -n ${NAMESPACE} delete pvc --ignore-not-found 2>/dev/null To delete PVCs for redis-redis-ha-server:
kubectl -n ${NAMESPACE} get pvc --no-headers -l app=redis-redis-ha-server 2>/dev/null | awk '{print $1}' | xargs ${KUBECTL_BIN} -n ${NAMESPACE} delete pvc --ignore-not-found 2>/dev/null- To delete PVCs for redis-ha:
kubectl -n ${NAMESPACE} get pvc --no-headers -l app=redis-ha 2>/dev/null | awk '{print $1}' | xargs ${KUBECTL_BIN} -n ${NAMESPACE} delete pvc --ignore-not-found 2>/dev/null
The opensearch-events-monitor pod is continuously in the CrashLoopBackOff state
Symptom
The event search is slower than normal.
Scope
The opensearch-events-monitor pod monitors the following OpenSearch event pods:
opensearch-events-data-0
opensearch-events-data-1
opensearch-events-data-2
opensearch-events-master-0
opensearch-events-master-1
opensearch-events-master-2
If the opensearch-events-monitor pod detects issues with any opensearch-events pods, it tries to remediate the issue. If the issue persists, the opensearch-events-monitor pod gets into the CrashLoopBackOff state.
Resolution
Contact the BMC Support team.