Monitoring SMB Connections in Kubernetes with CSI Driver

[uk1988]

Hej

First, thank you for the great work maintaining this driver!

I’m curious how others are monitoring SMB connections in a Kubernetes cluster. We’ve been using this driver for over a year, and recently we encountered an issue we weren’t aware of due to unexpected Kubernetes behavior. A critical pod relying on a PVC mounted via the SMB storage class got stuck during a restart. This happened because the SMB server restarted, resulting in stale mounts, and the pod couldn’t recreate itself since the storage was unavailable.

For context, when deploying our solution to clients’ on-prem systems, we include Grafana and Prometheus for monitoring. We’d like to visualize the SMB connection health on a dashboard—e.g., detecting when SMB mounts are stale or the server is down. However, the metrics available in the csi-smb-controller don’t seem sufficient for this purpose.

How are others addressing this? Are there recommended metrics, workarounds, or additional tools to monitor SMB connection states effectively in a Kubernetes environment?

Thanks!