Created attachment 1667309 [details]
GRUB needs to be selected manually else this task remains in hold state
Description of problem:
Scaling of worker node is not happening automatically it needs a manual intervention for GRUB selection.
Version-Release number of the following components:
Tested version: registry.svc.ci.openshift.org/ipv6/release:4.3.0-0.nightly-2020-02-05-064223-ipv6.3
Deploy an Openshift IPI cluster in latest fork ipv6/release:4.3.0-0.nightly-2020-02-05-064223-ipv6.3 once the cluster is up
oc get nodes
NAME STATUS ROLES AGE VERSION
master-0.kni7.cloud.lab.eng.bos.redhat.com Ready master 22h v1.16.2
master-1.kni7.cloud.lab.eng.bos.redhat.com Ready master 22h v1.16.2
master-2.kni7.cloud.lab.eng.bos.redhat.com Ready master 22h v1.16.2
worker-0.kni7.cloud.lab.eng.bos.redhat.com Ready worker 25m v1.16.2
worker-1.kni7.cloud.lab.eng.bos.redhat.com Ready worker 20h v1.16.2
worker-2.kni7.cloud.lab.eng.bos.redhat.com Ready worker 21h v1.16.2
Steps to Reproduce:
1. Deploy the cluster
2. Use the link -> https://github.com/openshift-kni/baremetal-deploy/blob/master/install-steps.md#deploy-the-provisioner-node-as-a-worker-node-using-metal3 to provision the worker using the bmh
3. The worker will not come up until you go to idrac and select the GRUB menu manually.
Worker-0 should come up without any manual intervention
worker node should come up without any manual intervention.
We haven't been able to reproduce, please re-open if you can provide more information. It may be an issue that needs to be discussed with the CoreOS team. The baremetal IPI installer isn't involved with grub on the installed machine OS.