Graylog2 / Graylog2/graylog-helm
RKE2 Deployment Fails, Can't connect to Datanode
- Dominant language
- Go Template
- Stars
- 12
- Forks
- 3
- Avg merge
- 2d 4h
- Merged PRs (30d)
- 13
Description
## Summary
org.graylog2.configuration.IndexerDiscoveryProvider - No datanode is available. Retry #1
### Details
All containers are in the same namespace
From the datanode logs, it appears that the opensearch cluster is never created.
`[ClusterNodeStateTracer] Updating cluster node 08a99607-8fc6-4a7a-b5b2-55d6ad6d1fd9 from UNAVAILABLE to UNAVAILABLE (reason: PROCESS_TERMINATED)`
There are also `Interrupted operation as its client disconnected` errors in the mongodb logs but these may be a result of the graylog nodes not completing setup since they cannot reach datanodes.
### Impact
Graylog cluster never completes deployment. Graylog nodes stuck "In Progress" perpetually.
## How to Reproduce?
kubectl delete -f namespace
kubecstl apply -f release
## Environment
- Helm chart version: 1.0.0
- Graylog version: 7.1.3
- Kubernetes version: v1.34.6+rke2r3
- Cloud provider / platform: RKE2 On Prem
- StorageClass: nfs-subdir-external-provisioner
Contributor guide
Assessment
This issue has not been assessed yet.