The build is failing on ubuntu x86 due to timeout....
# daft-dev
k
The build is failing on ubuntu x86 due to timeout. Is there a way to make a change to main and publish to the same version? The binaries won't change, we only need to change the actions timeout duration. @jay @Sammy Sidhu
j
Unfortunately no, we have to yank I think. Surprised that this issue didn't show up as we were building our nightlies
Let me take a look
j
How do we know this is a timeout? I don't see a timeout specification on the workflow file
k
Wait you're right this is not a timeout
j
Looks like there are a ton of test failures
UnboundLocalError: cannot access local variable 'process' where it is not associated with a value
k
why would
moto_server
error out?
j
Try re-running?
Very odd error
k
I reran once already
and both tests in the matrix for ubuntu x86 are failing the same thing
I'll try one more time though
j
Is this failing on main too?
k
nope
at least not with the most recent commit passes all io integration tests, including the one that is failing on release. it was run before the release I believe
j
Yeah not sure what’s going on. We should dig in more and probably will have to yank and re-release
k
could you take this over? i was feeling a little under the weather and went home early today
j
@Desmond Cheong as oncall— could you take the investigations? If not I’ll take a look later tonight
d
Sure thing
j
Seems like the main problem is that start_server function not being able to start the server correctly 🤷
🤔 1
Yeah don't really know what's going on, there's no error logs coming out of it either:
I'm going to skip the tests on our release pipeline and then cut the release on that, after making sure
main
is green with those changes!
🙌 1
Note there's no need to yank anything since the actual push to pypi didn't happen because of the errors I deleted the tag remotely already as well
d
sounds good. I'm checking whether moto_server is properly defined and has the right settings in a different ci run https://github.com/Eventual-Inc/Daft/actions/runs/13065553280/job/36457213602?pr=3743
I can't seem to repro this issue locally so it's probably some remote setup issue
j
Re-running the publishing now