| opendevreview | 임채민 proposed openstack/openstack-zuul-jobs master: [WIP] Add ansible play for weblate client configuration https://review.opendev.org/c/openstack/openstack-zuul-jobs/+/921878 | 00:27 |
|---|---|---|
| tkajinam | hmm I noticed that propose-updates job is broken now https://zuul.opendev.org/t/openstack/builds?job_name=propose-updates | 04:03 |
| tkajinam | error: unknown option `trailer=Generated-By:openstack/project-config:playbooks/proposal/propose_update.sh' in p-o-i jobs | 04:04 |
| tkajinam | ModuleNotFoundError: No module named 'requests' in devstack jobs | 04:04 |
| tkajinam | frickler, do you still have any concern with migrating propose-updates to ubuntu-jammy or even noble ? | 04:07 |
| tkajinam | seeing https://github.com/openstack/project-config/blob/master/zuul.d/jobs.yaml#L761-L765 | 04:07 |
| tkajinam | I have to go out now but will check logs this evening again | 04:07 |
| *** ykarel__ is now known as ykarel | 08:00 | |
| frickler | tkajinam: you can fix it for the specific job for now, we did the same with https://review.opendev.org/c/openstack/project-config/+/959382 | 08:25 |
| frickler | tkajinam: we can look into bumping the nodeset for the parent, too, but I'd wait with that until after the release | 08:26 |
| opendevreview | Michal Arbet proposed openstack/pbr master: Add support for local version as defined in PEP 440 https://review.opendev.org/c/openstack/pbr/+/942594 | 09:17 |
| tkajinam | frickler, ah, ok | 12:27 |
| opendevreview | Takashi Kajinami proposed openstack/project-config master: Migrate propose-updates job for p-o-i in noble https://review.opendev.org/c/openstack/project-config/+/961475 | 12:30 |
| tkajinam | frickler, I suspect the job for all repos is broken without the migration so fixing these even now may not harm but I don't have strong opinion about it and will leave the others now as suggested | 12:31 |
| opendevreview | Takashi Kajinami proposed openstack/project-config master: Migrate propose-updates job for p-o-i to noble https://review.opendev.org/c/openstack/project-config/+/961475 | 12:46 |
| opendevreview | Chaemin Lim proposed openstack/openstack-zuul-jobs master: [WIP] Add ansible play for weblate client configuration https://review.opendev.org/c/openstack/openstack-zuul-jobs/+/921878 | 13:21 |
| fungi | tkajinam: yeah sorry about that, i didn't realize that the trailer option was so recent when i added that call, the jobs i initially checked were working, so when we later spotted failures it made more sense to just move the jobs to newer platforms | 14:02 |
| tkajinam | fungi, np about that | 14:19 |
| tkajinam | and I have no problem with migrating these jobs to newer versions instead of keep using the old one | 14:20 |
| tkajinam | so I prepared https://review.opendev.org/c/openstack/project-config/+/961475 to migrate the job for p-o-i only to noble but can update it if you have a different suggestion https://review.opendev.org/c/openstack/project-config/+/961475 | 14:20 |
| tkajinam | (sorry I included the link twice by mistake | 14:21 |
| rosmaita | i'm assuming someone already reported this, but i'm getting 403 from https://zuul.opendev.org/ | 14:26 |
| rosmaita | (can't look in the channel logs because i'm also getting 403 from https://eavesdrop.openstack.org/irclogs/ ) | 14:26 |
| fungi | rosmaita: did it possibly start around 16:20 utc yesterday? that's when https://review.opendev.org/c/opendev/system-config/+/960399 deployed to update the user agent strings we block on those systems (trying to weed out the llm training crawlers that lie in their ua strings) | 14:30 |
| fungi | what user agent string is your browser supplying? | 14:31 |
| rosmaita | looking ... just discovered that it's not an issue with firefox | 14:31 |
| fungi | we don't expect that list of patterns to match legitimate browsers, but it's likely we made a mistake | 14:32 |
| rosmaita | 'Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/129.0.0.0 Safari/537.36' | 14:34 |
| fungi | and that's from the browser that's broken? | 14:35 |
| rosmaita | yes, that what it says it's using | 14:36 |
| fungi | i think that matches the pattern on line 116 as of the patch above | 14:37 |
| fungi | previously we were only matching up to chrome 99.x and now we match up to 129.x | 14:38 |
| *** dansmith_ is now known as dansmith | 14:39 | |
| fungi | so that seems likely to be the culprit. looking at how we should adjust it | 14:39 |
| fungi | looks like mainline chrome is on 140.x right now | 14:39 |
| fungi | oh i see, we were previously blocking up to 129.x for windows builds but not for linux builds | 14:42 |
| fungi | rosmaita: is that a distro packaged chrome/chromium build, or something else? | 14:43 |
| rosmaita | my string looks exactly like line 116 | 14:43 |
| rosmaita | not sure, i think distro packaged for fedora | 14:44 |
| clarkb | rosmaita: fungi: I decided chrome 12X was safe to drop because ~140 is the current version and the oldest lts (only on chromeos iirc) is ~132 | 14:46 |
| clarkb | rosmaita: fungi: if we have to remove that rule we can but can you also check that you don't needt oupdate your system too? | 14:46 |
| rosmaita | well, i do need to update my system, i have been putting it off | 14:47 |
| fungi | it seems like a legitimate enough agent string, even if the software is in need of an update. i wonder if we should only block up through 128.x in case others are on the same packaged version for similar reasons | 14:48 |
| clarkb | fungi: yes they are all legit user agent strings, thats the issue. The bad crawlers are spoofing legit agents so that we're less liekly to block them | 14:48 |
| clarkb | but yes we could also drop the rule. But I think we should only do that if it is reasonable for someone to be running such an old version. Even debian seems to give you the latest chrome version on stable. rosmaita do you mind tell us which distro rlease you're on so that we can double check? | 14:49 |
| rosmaita | Fedora 36 | 14:50 |
| clarkb | which hit EOL in may 2023 | 14:50 |
| rosmaita | :D | 14:50 |
| clarkb | I won't -2 a change to drop the ruel for 129, but I think we might all be better served if rosmaita takes this as a clear indication that an upgrade is a good idea and does that instead | 14:51 |
| clarkb | (I did try to lookup current versions of browsers before extending those rules fwiw) | 14:52 |
| rosmaita | ok, i will make upgrading a priority (just not today) | 14:53 |
| clarkb | as a side note chrome and chromium both identify as chrome in the user agent. Historically chromium said it was chromium but stopped at some point. So in this context it can be either browser | 14:56 |
| clarkb | google does publish a packages for chrome proper that are kept up to date which is I suppose another option for people not wanting to update the entire operating system. Mozilla publishes packages and a self updating tarball for firefox which can be used in a similar way as well | 14:57 |
| rosmaita | i'm using google-chrome on the theory that since my org uses google tools, google probably tests to make sure they work well with google's browser | 14:58 |
| clarkb | right I'm not suggesting that you change the browser. I was just pointing out there are ways to get updated versions of both chrome and firefox on old system that are no longer patching themselves | 14:59 |
| rosmaita | oh, i understand, i was just clarifying which browser i'm using | 14:59 |
| clarkb | as a side note I use google tools in firefox and things work the vast majority of the time. The last real issue I can recall is that you couldn't add a fido u2f key to a google account within firefox but could auth with them once added via chrome/chromium. And that was years ago and I think this all works now with firefox too | 15:00 |
| rosmaita | ok, upgraded browser to Version 140.0.7339.127 (Official Build) (64-bit) and i get 200s now | 15:03 |
| rosmaita | thanks! | 15:03 |
| fungi | i actually have not had any problems with google tools (mail, docs, drive, meet, et cetera) in qutebrowser for that matter | 15:20 |
| clarkb | it wouldn't surprise me if they work extra hard compared to most to ensure broad browser compatiblity with their sites as it likely helps fight off the "chrome should be split out of google" argument | 15:26 |
| opendevreview | 임채민 proposed openstack/project-config master: Add Weblate support to translation update pipeline https://review.opendev.org/c/openstack/project-config/+/961499 | 15:33 |
| opendevreview | Stephen Finucane proposed openstack/project-config master: Initiate retirement of shade https://review.opendev.org/c/openstack/project-config/+/961522 | 17:05 |
| opendevreview | Stephen Finucane proposed openstack/project-config master: Retire shade https://review.opendev.org/c/openstack/project-config/+/961524 | 17:09 |
| *** clif1 is now known as clif | 17:32 | |
| *** mnaser_ is now known as mnaser | 20:36 | |
Generated by irclog2html.py 4.0.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!