Description of problem: metal jobs failing because "provisioning time limit exceeded; the Packet team will investigate" Example: https://prow.svc.ci.openshift.org/view/gcs/origin-ci-test/logs/release-openshift-ocp-installer-e2e-metal-4.2/183
I met with Zac and Golden @ Packet yesterday to discuss this and they've informed me that this particular message indicates that the device was provisioned but that the host OS never reached running state so this is most likely a failure in PXE / Ignition processes. So it seems like there's definitely something to look into here. In all cases where a running OS became available NetworkManager-wait-online.service is in a failed state because the second interface is not properly configured. This service now blocks other services and this introduces a 300 second delay in the boot process and with hosts rebooting multiple times this has potential to cause job failure. I'm attmepting to only configure the first interface and see if that produces better results.
*** Bug 1772212 has been marked as a duplicate of this bug. ***
We cannot ship without this in 4.3, marking appropriately.
*** This bug has been marked as a duplicate of bug 1775388 ***