Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Send system journal logs to console #2399

Closed
wants to merge 1 commit into from

Conversation

stbenjam
Copy link

When debugging early boot issues, especially ones that leave the host without functional networking, it's difficult to debug what broke. This change makes the default to send the systemd journal logs to the console on x86_64. We attempted to do this in openshift with a machineconfig, and this broke arm64 and this config was offered as an alternative.

When debugging early boot issues, especially ones that leave the host
without functional networking, it's difficult to debug what broke. This
change makes the default to send the systemd journal logs to the console
on x86_64.  We attempted to do this in openshift with a machineconfig,
and this broke arm64 and this config was offered as an alternative.
@bgilbert
Copy link
Contributor

Default kernel arguments are a last resort for configuration that really can't be done any other way. If we want to do this, I think we should do it by adding a journald config file to the initrd instead. But I'm not sure it's a good idea to do this everywhere. Would it make sense to add this to debugging docs and leave the defaults as is?

@stbenjam
Copy link
Author

I'm getting pointed many different directions by the coreos folks here. We need the journal logs in CI, and I was asked to back out openshift/release#38015, and this was the recommended approach by @travier

@cgwalters
Copy link
Member

I don't think we should do this by default in FCOS/RHCOS, there'd be a lot of fallout. One in particular is that today printing to the serial console can actually cause large system slowdown, and we explicitly did coreos/fedora-coreos-tracker#1244 to avoid that for basic kernel logs. This would greatly increase the output.

@cgwalters cgwalters closed this Apr 26, 2023
@cgwalters
Copy link
Member

this broke arm64

link/more info?

We need the journal logs in CI

Yes, but generally only on failures. This relates to ye olde coreos/ignition#585

@travier
Copy link
Member

travier commented May 23, 2023

I'm getting pointed many different directions by the coreos folks here. We need the journal logs in CI, and I was asked to back out openshift/release#38015, and this was the recommended approach by @travier

OK, my bad, I hadn't realized that this was just for CI and not something more permanent.

@stbenjam
Copy link
Author

this broke arm64

link/more info?

https://issues.redhat.com/browse/OCPBUGS-11812 is where it broke on ARM

We need the journal logs in CI

Yes, but generally only on failures. This relates to ye olde coreos/ignition#585

I don't think this would help since the failures were mysterious networking failures. If there was something that if network never comes up, the journals got dumped the serial console might help.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

4 participants