Describe the feature you'd like to have
Each Ceph-CSI Pod should have a PodMonitor, so that Prometheus knows how to scrape the CSI metrics.
Some considerations:
- Pods with hostnetworking (by default the node-plugins) can easily cause conflicts when listening on a local port
- ...
Who is the end user and what is the use case where this feature will be valuable?
Through the Prometheus Operator it is possible to get statistics and setup alerting. This helps administrators identifying issues.
How will we know we have a good solution? (acceptance criteria)
Ceph-CSI metrics (from the standard Kubernetes CSI sidecars) become available in Prometheus.
Additional context
Suggested in ceph/ceph-csi#6287. Rook supports ServiceMonitor for the same purpose.
More details in https://prometheus-operator.dev/docs/developer/getting-started/ .
Describe the feature you'd like to have
Each Ceph-CSI Pod should have a PodMonitor, so that Prometheus knows how to scrape the CSI metrics.
Some considerations:
Who is the end user and what is the use case where this feature will be valuable?
Through the Prometheus Operator it is possible to get statistics and setup alerting. This helps administrators identifying issues.
How will we know we have a good solution? (acceptance criteria)
Ceph-CSI metrics (from the standard Kubernetes CSI sidecars) become available in Prometheus.
Additional context
Suggested in ceph/ceph-csi#6287. Rook supports ServiceMonitor for the same purpose.
More details in https://prometheus-operator.dev/docs/developer/getting-started/ .