Description of problem: When installing OCP4.1 on Baremetal, console stops logging on "Probing EDD(edd=off to disable)... ok" and we cannot see the rest of the log. If host doesn't start correctly, we need to reboot it and manually add it, losing the information of the first boot. QE debugging is being harder without that parameter as default. Version-Release number of selected component (if applicable): OCP4.1 based on any RHCOS8 version How reproducible: Steps to Reproduce: 1.Start any installation of OCP4.1 2.Look at the virtual console of the host during the booting process 3. Actual results: Console information stops on "Probing EDD(edd=off to disable)... ok", no information about network asignation or ignition process is showed Expected results: All log is showed by default Additional info:
David, Can you provide what version of the bare metal artifact you saw this in? Micah looked at a recent build and found the requested change already present.
This should have already been fixed by https://github.com/coreos/coreos-assembler/pull/485.
Sorry Steve, Subject is wrong, we need console=tty0,115200n8, not just console=ttyS0,115200n8 Baremetal servers uses tty0 on the virtual console from KVM or iDRac. My fault, I have updated the subject
I can confirm that I only see `console=ttys0,115200n8` in 410.8.20190506.1 from ART (the one David was reporting on) Checking why that is the case
Ok it turns out ART is using a very old COSA (From Apr. 4). Should be an easy fix
Since the problem described in this bug report should be resolved in a recent advisory, it has been closed with a resolution of ERRATA. For information on the advisory, and where to find the updated files, follow the link below. If the solution does not work for you, open a new bug report. https://access.redhat.com/errata/RHBA-2019:0758