OpenShift worker nodes are not added to internal load balancer on Azure

Solution Unverified - Updated -

Environment

  • OpenShift 4.10.x
  • Private cluster
  • Microsoft Azure

Issue

  • When scaling up a worker machineset, the newly added nodes are not added to the internal load balancer
  • After making configuration changes to worker nodes and scaling down all of the old nodes and scaling up new nodes, the cluster is no longer accessible via routes

Resolution

  • Modify each of the worker machinesets to add the internalLoadBalancer: option and name.
  • The option should be added in spec.template.spec.providerSpec.value
...
spec:
  replicas: 1
  selector:
    ...
  template:
    metadata:
      ...
    spec:
      metadata: {}
      providerSpec:
        value:
          apiVersion: azureproviderconfig.openshift.io/v1beta1
          credentialsSecret:
            ...
          image:
            ...
          kind: AzureMachineProviderSpec
          internalLoadBalancer: lb-internal          <------------------- ADD IT HERE
          location: us-east-2
          managedIdentity: ocp-clustername-identity
          ...

Root Cause

  • There is a bug in the OpenShift 4.10 code where the internalLoadBalancer: option is not added to the machinseset.
  • JIRA CORS-2698 is open for this issue

Diagnostic Steps

  • Scale up a worker machineset and view the internal Load Balancer in the Azure console
  • The newly created node(s) will not be attached to the Load Balancer

This solution is part of Red Hat’s fast-track publication program, providing a huge library of solutions that Red Hat engineers have created while supporting our customers. To give you the knowledge you need the instant it becomes available, these articles may be presented in a raw and unedited form.

Comments