Configuring Prometheus OpenMetrics monitoring in Unified Agent

When you deploy the Unified Agent, you can select to enable Prometheus OpenMetrics plug-in to handle Prometheus sources.

1. Edit the CR by running the following command:

   kubectl edit ua ua-openmetrics -n ${UA_INSTALL_NAMESPACE}
   

2. Locate the section of [[inputs.openmetrics]].

3. Change enabled: false to enabled: true.

4. Specify the name and urls for the Prometheus OpenMetrics resources and service.
   See the following example:

   spec:
     plugins:
     - data:
         openmetrics.conf: |-
           [[inputs.openmetrics]]
           name = "node"
           urls = ["http://9.46.65.155:9100/metrics"]
           direct = true
       enabled: true
       name: openmetrics
   

5. Save the CR.

6. Verify that the pods are restarted after you save the CR.

   kubectl get pod -n ${UA_INSTALL_NAMESPACE}
   

   You can get the following output:

   NAME                                READY   STATUS              RESTARTS   AGE
   ua-mgmt-ua-cloud-monitoring-987nl   0/1     ContainerCreating   0          9s
   ua-mgmt-ua-cloud-monitoring-kc5fq   1/1     Running             0          5d1h
   ua-mgmt-ua-cloud-monitoring-krp5f   1/1     Running             0          5d1h
   ua-mgmt-ua-cloud-monitoring-mdr9c   1/1     Running             0          29s
   ua-mgmt-ua-cloud-monitoring-mkq62   1/1     Running             0          5d1h
   

7. Verify that the enabled plug-in is loaded:

   kubectl -n ${UA_INSTALL_NAMESPACE} logs ua-mgmt-ua-cloud-monitoring-987nl | grep "Loaded inputs"
   

   See the following example output:

   2020-11-24T07:15:17Z I! Loaded inputs: blockchain cpu mem disk diskio net procstat system processes lwdc nginx openmetrics zipkin jaeger redis