-
Feature Request
-
Resolution: Unresolved
-
Major
-
None
-
None
-
None
-
Product / Portfolio Work
-
None
-
False
-
-
None
-
None
-
None
-
-
None
-
None
-
None
-
None
-
None
1. Proposed title of this feature request
Add ServiceMonitor / PodMonitors for all HCP workloads
2. What is the nature and description of the request?
Currently, there's only a limited amount of workloads in the control plane for which the control plane operator creates servicemonitors/podmonitors:
oc get servicemonitors.monitoring.rhobs -n ocm-staging-<blurr> NAME AGE audit-webhook 48m catalog-operator 32m cluster-version-operator 32m etcd 32m kube-apiserver 32m kube-controller-manager 32m monitor-multus-admission-controller 44m monitor-ovn-control-plane-metrics 44m node-tuning-operator 32m olm-operator 32m openshift-apiserver 32m openshift-controller-manager 32m openshift-route-controller-manager 32m oc get podmonitors.monitoring.rhobs -n ocm-staging-<blurr> NAME AGE cluster-image-registry-operator 3h17m controlplane-operator 3h35m hosted-cluster-config-operator 3h17m ignition-server 3h17m ingress-operator 3h17m
This should be extended to all control plane workloads to allow SRE to monitor any component via its metrics. As an example, the above has no service/podmonitor for kube-scheduler.
3. Why does the customer need this? (List the business requirements here)
SRE needs access to metrics of each control plane component.
4. List any affected packages or components.
- is related to
-
-
- To Do
-