<dustymabe>
jlebon: the day where we push multi-arch cosa images to quay has come!
paragan has joined #fedora-coreos
<jlebon>
dustymabe: woohoo, sweet!
<dustymabe>
jlebon: got to pass code review first.. but as part of that we should discuss a few things (when you're ready)
<jlebon>
dustymabe: sounds good
<jlebon>
here or video?
<dustymabe>
jlebon: here works
<jlebon>
k
<dustymabe>
jlebon: first - syncing from registry.ci - we'll disable that now, correct?
<jlebon>
dustymabe: yeah, we should stage the PR for that so it's ready to merge
<dustymabe>
jlebon: do you know how - could you stage that PR and link me to it?
<jlebon>
dustymabe: yup, can do that
<jlebon>
hmm, we'll want the opposite actually; syncing from quay to registry.ci
<dustymabe>
correct
<dustymabe>
or rather.. should we just use an image stream in the various clusters that consume?
<jlebon>
a few things are pointing there now. but it would be less confusing long-term if we don't have it in both places
<dustymabe>
and the image stream points at quay.io directly?
<jlebon>
we can migrate what we know about, then tweak it to drop at least the :latest tag build and that'll shake out anything else still using it
<jlebon>
you're talking about the pipeline imagestream?
<dustymabe>
yeah, basically any openshift that is consuming the cosa image - stop pointing them at registry.ci, and just point the image streams at quay.io instead
* dustymabe
bbiab
<jlebon>
yup, exactly
rsalveti has joined #fedora-coreos
bgilbert__ has joined #fedora-coreos
jcajka has quit [Quit: Leaving]
paragan has quit [Quit: Leaving]
jbaublitzzz has joined #fedora-coreos
bgilbert__ is now known as bgilbert
Betal has joined #fedora-coreos
jpn has quit [Ping timeout: 240 seconds]
ravanelli has quit [Remote host closed the connection]
ravanelli has joined #fedora-coreos
ravanell_ has joined #fedora-coreos
ravanelli has quit [Ping timeout: 268 seconds]
justJanne has quit [Quit: So, if you care to find me, look to the western sky. As someone told me lately, everyone deserves a chance to fly.]
justJanne has joined #fedora-coreos
justJanne has quit [Client Quit]
justJanne has joined #fedora-coreos
<dustymabe>
saqali_: for https://github.com/coreos/fedora-coreos-tracker/issues/1268 is there anything in the logs that lead us towards the problem? relevant error messages and such? Armed with that information we should be able to open a BZ
saqali_ is now known as saqali
<saqali>
dustymabe: After running a vimdiff between a successful run and todays failed run, I wasn't able to find anything that stood out
<dustymabe>
saqali: i wonder if we're not getting all the logs
ravanell_ has quit [Remote host closed the connection]
<dustymabe>
if that looks good I'll test it and squash it into another commit
<jlebon>
dustymabe: added some comments :)
<dustymabe>
jlebon: i've got some questions :)
<dustymabe>
do I just use shwrapCapture right there in `node{}` to pick up the commit? Will I be in the correct directory at that time? Also what if the job got triggered manually? what does `checkout()` do in that case?
<jlebon>
dustymabe: yes, shwrapCapture should work there. by default the checkout is directly in the workspace and not a subdir IIRC. if triggered manually, it'll still checkout the repo and I *think* just default to `main`, but better to check how we were triggered
<jlebon>
grep for triggered_by_push()
<jlebon>
we can lift that to utils.groovy
<jlebon>
see build-mechanical for how that's used
<dustymabe>
hmm - ok - so if we weren't triggered by push then checkout() will just checkout the main branch?
<dustymabe>
is there any way to tell it what branch to checkout? i.e. user supplied param?
<dustymabe>
or I guess we could fall back to `git ls-remote` in the case that triggered_by_push() is false?
<jlebon>
i added a comment in the PR
<dustymabe>
we're doing a shallow checkout though
<dustymabe>
so that `git checkout` probably won't work
<jlebon>
oh fair. so we can `git fetch --depth=1` before the checkout