On Tue, Jun 09, 2020 at 11:53:41AM +0200, Andrea Bolognani wrote:
On Mon, 2020-06-08 at 19:07 +0200, Andrea Bolognani wrote:
On Mon, 2020-06-08 at 17:34 +0100, Daniel P. Berrangé wrote:
On Mon, Jun 08, 2020 at 05:42:31PM +0200, Andrea Bolognani wrote:
To see this in action, check out the
https://gitlab.com/abologna/libvirt/-/pipelines/153963978
pipeline, and in particular the
https://gitlab.com/abologna/libvirt/-/jobs/585678408 https://gitlab.com/abologna/libvirt/-/jobs/585678413
Ideally the actual build output from Cirrus CI would be pulled into the build output of the GitLab CI job. That way email notifications from failed builds will include the actual compile failure details. This also avoids having to cut+paste non-clickable links to the Cirrus CI job.
I agree that this would be desirable.
Earlier today, while reviewing the code for cirrus-run, I found a reference to a CIRRUS_LOG_FILE environment variable that should allow us to do just that, and was intending to try it out. If that doesn't work out, we should still be able to fetch the full job output from Cirrus CI after the fact.
I tried setting CIRRUS_LOG_FILE and it doesn't seem to work the way we want. Getting the taks log directly from Cirrus CI is definitely possible - you can do so from the Web UI already - but obtaining the correct URL requires calling the GraphQL API and I need some more time to figure out how that works. Possibly make some changes to cirrus-run as well.
Since it's already possible, although not very convenient, to get to the full log, I think this should not block the series and it can come as a follow-up instead. I'll post a v2 that addresses the other issue you pointed out.
Yep, that's fine Regards, Daniel -- |: https://berrange.com -o- https://www.flickr.com/photos/dberrange :| |: https://libvirt.org -o- https://fstop138.berrange.com :| |: https://entangle-photo.org -o- https://www.instagram.com/dberrange :|