| @blaisep-sureify:matrix.org | ...would it help to build the docker images with fedora? I have fedora on intel and on arm64 just sitting around. | 02:52 |
|---|---|---|
| @blaisep-sureify:matrix.org | I mean, if you're pulling `FROM docker.io/library/python:${PYTHON_VERSION}-slim-${DEBIAN_VERSION}` as the base image, it's no trouble for me to build you the same image from python.org sources. | 02:57 |
| -@gerrit:opendev.org- Kenny Ho proposed wip: [zuul/zuul-jobs] 843527: fetch-output-openshift: add option to specify container to fetch from https://review.opendev.org/c/zuul/zuul-jobs/+/843527 | 03:02 | |
| -@gerrit:opendev.org- Kenny Ho proposed wip: [zuul/zuul-jobs] 843527: fetch-output-openshift: add option to specify container to fetch from https://review.opendev.org/c/zuul/zuul-jobs/+/843527 | 03:04 | |
| -@gerrit:opendev.org- Kenny Ho proposed wip: [zuul/zuul-jobs] 843527: fetch-output-openshift: add option to specify container to fetch from https://review.opendev.org/c/zuul/zuul-jobs/+/843527 | 03:10 | |
| -@gerrit:opendev.org- Kenny Ho marked as active: [zuul/zuul-jobs] 843527: fetch-output-openshift: add option to specify container to fetch from https://review.opendev.org/c/zuul/zuul-jobs/+/843527 | 03:15 | |
| @fungicide:matrix.org | > <@blaisep-sureify:matrix.org> ...would it help to build the docker images with fedora? I have fedora on intel and on arm64 just sitting around. | 13:35 |
| we have fedora nodes available in the opendev deployment already, it's a matter of wheels for python libraries with compiled c extensions needing to be built for the exact distribution used inside the base image. what we're missing is a wheel cache for the distro those official python docker containers are based on | ||
| @jpew:matrix.org | It would be nice if the Zuul website console showed the number of times a given ansible task was attempted (if it's configured to try multiple times); It's not very clear to users when this occurs ATM | 16:07 |
| @jpew:matrix.org | > <@jpew:matrix.org> It would be nice if the Zuul website console showed the number of times a given ansible task was attempted (if it's configured to try multiple times); It's not very clear to users when this occurs ATM | 16:07 |
| The information is in job-output.json, so I assume the website can get it also | ||
| @blaisep-sureify:matrix.org | ok, it was an opportunity for me to give it a go. I had not built from source in several years. On this new server, it took 13 secs, which is a lot faster than I remember. | 16:31 |
| @clarkb:matrix.org | Zuulians that deploy for master beware that I think the change to add deduplication of jobs for cycles is buggy and cannot be upgraded online without some impact. The cause of this is we unconditionally add deduplicate to attributes here: https://opendev.org/zuul/zuul/src/branch/master/zuul/model.py#L2008 and some jobs may not have that attribute if they were started on the old model | 17:17 |
| @clarkb:matrix.org | * Zuulians that deploy from master beware that I think the change to add deduplication of jobs for cycles is buggy and cannot be upgraded online without some impact. The cause of this is we unconditionally add deduplicate to attributes here: https://opendev.org/zuul/zuul/src/branch/master/zuul/model.py#L2008 and some jobs may not have that attribute if they were started on the old model | 17:18 |
| @clarkb:matrix.org | On the opendev side of things I think it is only affecting jobs that were enqueued while we were in the short period between the old and new model api version | 17:18 |
| @clarkb:matrix.org | for this reason I'm going to see if we can roll forward and just reenqueue things to work around this | 17:18 |
| @clarkb:matrix.org | But it may be necessary to stop the cluster and delete zk state and start on master or v6 | 17:18 |
| @clarkb:matrix.org | We have successfully dequeued the two buildests for the two jobs that were affected. We reenqueued one of them (the other was a periodic buildset and we can wait for the next run) and zuul looks happy now | 17:40 |
| @clarkb:matrix.org | https://paste.opendev.org/show/bHNqoDi2M2f9ExF9s1NH/ is the traceback. I'll give a fix a go but no promises :) | 17:41 |
| -@gerrit:opendev.org- Clark Boylan proposed: [zuul/zuul] 843666: Fix the model API 8 transition https://review.opendev.org/c/zuul/zuul/+/843666 | 17:53 | |
| @clarkb:matrix.org | I think the fix may be as simpel as that, but we probably do want to add more robust testing to eb sure and I haven't started looking at that aspect of this yet | 17:53 |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!