k8sdc-cr-k8monitor and ua-mgmt-ua-cloud-monitoring pods are not created after installing Monitoring DataProvider Management

Problem

After installing Monitoring DataProvider Management, the k8sdc-cr-k8monitor and ua-mgmt-ua-cloud-monitoring pods are not created.

Cause

• For remote deployment, the cause is that the cluster namespace is not assigned to a team in Identity and Access.

• For local deployment, it can be caused by the following reasons.

  1. Command in crtSecrets.txt is not executed on the managed cluster.
  2. There is reconcile error in ibm-dc-autoconfig-operator pod.

Resolving the problem

• For remote deployment, assign the cluster namespace to a team for account.

  1. In the IBM Cloud Pak console, click Administer > Identity and Access and then click the Teams tab.
  2. Click a team and then click the Resource tab.
  3. Click Manage resources, select your managed cluster, and then click Save.
  4. In the IBM Cloud Pak console, click Monitoring Health > Infrastructure monitoring to trigger automation, by which pods are automatically created in a short time.

  5. If the previous step does not work, restart the icam-ibm-cem-brokers and icam-ibm-cem-cem-users pods on the IBM Cloud Pak® for Multicloud Management hub cluster.

     oc delete $(oc get pods -n management-monitoring -o name | grep 'cem-brokers\|cem-users') -n management-monitoring
     

• For local deployment, do the following steps.

  1. If you install Monitoring DataProvider Management from OperatorHub, follow steps 11-15 of Local deployment from OperatorHub
  2. If you install Monitoring DataProvider Management by script, follow steps 6-10 of Local deployment by script.
  3. Restart ibm-dc-autoconfig-operator by running the following command:

     kubectl delete $(kubectl get po -l app=ibm-dc-autoconfig-operator -n cp4mcm-cloud-native-monitoring -o name) -n cp4mcm-cloud-native-monitoring