| *** ralonsoh_ is now known as ralonsoh | 07:04 | |
| opendevreview | sean mooney proposed openstack/openstack-zuul-jobs master: Migrate OpenStack tox doc jobs to Ubuntu Noble (24.04) https://review.opendev.org/c/openstack/openstack-zuul-jobs/+/935459 | 20:25 |
|---|---|---|
| sean-k-mooney | gmann: ^ https://review.opendev.org/c/openstack/openstack-zuul-jobs/+/935459/2..3/zuul.d/jobs.yaml that shoudl fix it | 20:26 |
| sean-k-mooney | gmann: the tldr is libprce is listed as a "test" profile dep not as a doc or compile dep | 20:29 |
| sean-k-mooney | as far as i can tell compile is not really used in openstack at all | 20:30 |
| sean-k-mooney | its used in opendev, zuul and vexhost tenants | 20:30 |
| sean-k-mooney | but openstack does not split out deps into a compile target for c python modules | 20:31 |
| fungi | looks like you also rebased when you updated that change | 20:31 |
| sean-k-mooney | my guess is on 22.04 libpcre3-dev was installed as a result of something else | 20:31 |
| fungi | i'm having trouble separating the rebase diff from the revision | 20:31 |
| sean-k-mooney | hum i didnt pass -R to git-review | 20:32 |
| sean-k-mooney | fungi: i just add test to the bindep profile | 20:32 |
| fungi | okay, the diff between patchset 2 and 3 also seems to include other edits like removal of playbooks/openafs-rpm-package-build/pre.yaml and switching dnf tasks to package tasks in roles/openafs-rpm-package-build/tasks/main.yaml, along with removal of some stable/2025.1 branch inclusions for py36 jobs in project-templates | 20:34 |
| sean-k-mooney | fungi: i broght it in line with https://opendev.org/zuul/zuul-jobs/commit/3f27b0cd46cde913713fef03a5257705b82536d6 | 20:34 |
| sean-k-mooney | fungi: for me at least those show up in a diffent colour .i.e. not form the currnt commit | 20:35 |
| fungi | yeah, gertty probably lacks the ability to render that nuance from rebases | 20:35 |
| sean-k-mooney | so normlaly if im being extra carful ill pass -R to prevent git-review from doing an automatic rebase | 20:36 |
| sean-k-mooney | i tought it normally prints when it does rebase? | 20:37 |
| sean-k-mooney | oh i bet it was this | 20:37 |
| sean-k-mooney | Your change was committed before the commit hook was installed. | 20:37 |
| sean-k-mooney | Amending the commit to add a gerrit change id. | 20:37 |
| sean-k-mooney | fungi: it propted me to set up the gerrit remote ectra since this was a clean clone of the repo | 20:38 |
| fungi | it's possible that resulted in a rebase, though i'd consider that a git-review bug if so | 20:38 |
| fungi | anyway, the updated change lgtm, thanks@ | 20:39 |
| fungi | s/@/!/ | 20:39 |
| sean-k-mooney | we need to recheck the nova patch to see if it fixes the probelm | 20:41 |
| fungi | yes, of course, that's why i didn't approve it ;) | 20:41 |
| sean-k-mooney | i just kicked it off again | 20:41 |
| sean-k-mooney | so we can see if https://review.opendev.org/c/openstack/nova/+/935577 is happy now | 20:42 |
| sean-k-mooney | anyway im going to finsih for tonight | 20:43 |
| sean-k-mooney | gmann: hopefully that passes and we can unblock that, ill take a look agian in the morning at the results | 20:43 |
| sean-k-mooney | ok it started runnign already and its not including 2025-03-10 20:46:14.090021 | ubuntu-noble | libpcre3-dev | 20:46 |
| sean-k-mooney | so we should be good | 20:47 |
| sean-k-mooney | i think that means we dont actully need the nova patch at all | 20:47 |
| fungi | great news, thanks! | 20:48 |
| clarkb | is this taht random regex library that openstack wrote depending on libpcre? | 20:49 |
| clarkb | I seem to recall its used for like a single feature and we've had to fix this no wheel for python3.12 problem in a bunch of places | 20:50 |
| clarkb | I can't remember what specifically it was used for but there was discussion of maybe rewriting things to drop the dep which is probably a good idea | 20:50 |
| fungi | i just tried to revisit, but the last failing test result in the nova depends-on change was from november so no logs remaining | 20:59 |
| fungi | no, i take that back, there was a newer run: https://zuul.opendev.org/t/openstack/build/0cd10f89ad134074b438270e770c5bc5 | 21:07 |
| fungi | it's related to https://pypi.org/project/python-pcre/ which was last updated 10 years ago | 21:08 |
| fungi | not anything openstack maintains | 21:08 |
| clarkb | maybe it was a library on top of that that openstack maintains | 21:08 |
| clarkb | fungi: https://opendev.org/openstack/whereto | 21:09 |
| clarkb | that is the only dependency in openstack | 21:09 |
| clarkb | so ya I think we can remove the dep entirely if we just replace the checker tool or use a different regex system etc | 21:10 |
| fungi | oh wow, yes that tool has a very limited use | 21:10 |
| clarkb | fungi: sean-k-mooney https://pypi.org/project/pcre2/ this exists and it sounds like pcre2 regexes are just extensions to pcre. Maybe not perfect for the use case but if its a simple dep replacement could still be worthwile | 21:30 |
| clarkb | https://github.com/PCRE2Project/pcre2/issues/51#issuecomment-1080027348 seems the C level apis are different and there are some additional regex features but using a subset that is pcre should mostly work | 21:37 |
| clarkb | I guess the risk here is that apache2 rules are pcre not pcre2 so we could test with pcre2 and say it looks good then fail when we deploy | 21:37 |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!