Known issues for the scheduling service

The following known issues apply to the scheduling service.

The ibm-cpd-scheduler-spoke-resource-collector pod is in the CrashLoopBackOff state

Applies to: 5.4.0

Fixed in: 5.4.1

The ibm-cpd-scheduler-spoke-resource-collector pod cannot start correctly if the ibm-cpd-scheduler-spoke-resource-collector-kube-system role binding does not exist in the kube-system project.

If the role binding does not exist, the ibm-cpd-scheduler-spoke-resource-collector pod is in the CrashLoopBackOff state.

Important: If you installed the scheduling service on a remote physical location, this problem can occur on either the cluster where the primary instance of IBM® Software Hub is installed or on the remote cluster.
Diagnosing the problem
To confirm that this issue is occurring:
Check the pod on the primary cluster
Run the following command to check the status of the ibm-cpd-scheduler-spoke-resource-collector pod:
oc get pods --namespace=${PROJECT_SCHEDULING_SERVICE} | grep ibm-cpd-scheduler-spoke-resource-collector
Check the pod on the remote cluster
Note: This step applies only if you use remote physical locations to expand your IBM Software Hub deployment.
oc get pods --namespace=${REMOTE_PROJECT_SCHEDULING_SERVICE} | grep ibm-cpd-scheduler-spoke-resource-collector
If the pod is in the CrashLoopBackOff state and the logs contain the following error, proceed to Resolving the problem:
Unable to get configmap/extension-apiserver-authentication in kube-system.
Resolving the problem
To resolve the problem, add the ibm-cpd-scheduler-spoke-resource-collector-kube-system role binding to the kube-system project on the cluster where the pod is in the CrashLoopBackOff state:
The problem is on the primary cluster
  1. Run the following command to create the role binding:
    kubectl create rolebinding ibm-cpd-scheduler-spoke-resource-collector-kube-system \
--namespace=kube-system \
--role=extension-apiserver-authentication-reader \
--serviceaccount=$${PROJECT_SCHEDULING_SERVICE}:ibm-cpd-scheduler-spoke-resource-collector-sa
  2. Delete the ibm-cpd-scheduler-spoke-resource-collector pod:
    oc delete pods --namespace=${PROJECT_SCHEDULING_SERVICE} | grep ibm-cpd-scheduler-spoke-resource-collector
The problem is on the remote cluster
  1. Run the following command to create the role binding:
    kubectl create rolebinding ibm-cpd-scheduler-spoke-resource-collector-kube-system \
--namespace=kube-system \
--role=extension-apiserver-authentication-reader \
--serviceaccount=$${REMOTE_PROJECT_SCHEDULING_SERVICE}:ibm-cpd-scheduler-spoke-resource-collector-sa
  2. Delete the ibm-cpd-scheduler-spoke-resource-collector pod:
    oc delete pods --namespace=${REMOTE_PROJECT_SCHEDULING_SERVICE} | grep ibm-cpd-scheduler-spoke-resource-collector