| tonyb | also https://gitea-lb02.opendev.org/opendev/-/code?q=gitea&t= gives me a 500 but works as expected in prod. I *expect* that's an issue with our test environment. | 00:00 |
|---|---|---|
| tonyb | again just wanting to be sure it is ok/expected | 00:00 |
| Clark[m] | Gitea constantly makes visual changes | 00:07 |
| Clark[m] | We have to update our templates to keep up every upgrade | 00:07 |
| Clark[m] | For search I suspect the issue is we don't have a lot of code in the test environment. But I would've expected a page with no results rather than a 500 | 00:08 |
| tonyb | This is what I get https://usercontent.irccloud-cdn.com/file/n3sJqGTg/image.png | 00:09 |
| tonyb | So it's kinda both? | 00:09 |
| clarkb | its weird if you look at the access log it shows 200s for everything and the browser network log seems to show similar | 00:14 |
| clarkb | both code search methods result in that 500 banner | 00:17 |
| clarkb | I wonder if it is due to this failed favicon.svg request | 00:18 |
| tonyb | Perhaps? I'll poke around locally. | 00:20 |
| clarkb | hrm no internet indicates that is normal reporting for "I don't need ot make this request I can use cached data" | 00:20 |
| clarkb | and chrome doesn't have that aborted request and instead seems to fetch the favicons successfully but the 500 still shows up | 00:20 |
| clarkb | tonyb: ok /var/log/containers/docker-gitea.log has info that is useful | 00:22 |
| clarkb | it is a template rendering error | 00:22 |
| clarkb | I think it is failing to render and adding the 500 content then returning 200 ... | 00:23 |
| clarkb | .../context_response.go:86:HTML() [E] Render failed: failed to render template: explore/code, error: template error: builtin(bindata):code/searchresults:34:83 : executing "code/searchresults" at <Safe>: wrong type for value; expected string; got template.HTML | 00:23 |
| tonyb | Okay | 00:24 |
| clarkb | https://github.com/go-gitea/gitea/issues/28607 | 00:24 |
| clarkb | we can wait for 1.21.4 | 00:25 |
| clarkb | or patch that specific fix on top of 1.21.3 or accept it being broken for a bit. No really good options | 00:25 |
| clarkb | tonyb: good catch :) | 00:27 |
| tonyb | It looks like the fix is in searchresults.tmpl, can we add that to our custom templates until 21.4 ? | 00:27 |
| clarkb | ya there are two template files that need updates. We can either vendor the whole file in our custom templates dir or we can apply the patch to the source code before building | 00:28 |
| clarkb | https://github.com/go-gitea/gitea/pull/28576/files these two files | 00:29 |
| clarkb | vendoring the whole file is somethign we can already do using the existing build setup. I guess we should do that for simplicity (eve nif it is a bit verbose) | 00:30 |
| clarkb | I'll update teh change | 00:30 |
| tonyb | Thanks. I agree that' | 00:30 |
| tonyb | s a good approach and is easy to undo when 1.21.4 comes out | 00:31 |
| opendevreview | Clark Boylan proposed opendev/system-config master: Update gitea to 1.21.3 https://review.opendev.org/c/opendev/system-config/+/897679 | 00:36 |
| opendevreview | Clark Boylan proposed opendev/system-config master: DNM intentional gitea failure to hold a node https://review.opendev.org/c/opendev/system-config/+/848181 | 00:36 |
| clarkb | I'll recycle the node hold | 00:36 |
| clarkb | thats done. Now we wait to confirm that fixes things for us | 00:38 |
| clarkb | side note: if you're going to render an error in your template saying there was a 500 maybe you should also return an http 500 error... | 00:38 |
| tonyb | Yes for sure | 00:38 |
| tonyb | I'm happy to to the etherpad upgrade during my evening, say UTC0500 ? | 00:57 |
| tonyb | and I'm also happy to do the backup pruning today | 00:57 |
| Clark[m] | tonyb: works for me. I've switched over to dinner mode now though | 01:24 |
| tonyb | That's fine. I expect it'll be "fine", and the back-put strategy is also reasonably well defined | 01:26 |
| tonyb | #status log etherpad services will be unavailable briefly while upgraded | 06:17 |
| opendevstatus | tonyb: finished logging | 06:17 |
| tonyb | infra-root: I' | 06:25 |
| tonyb | m also pruning the backups on backup02 | 06:25 |
| opendevreview | Merged opendev/system-config master: Upgrade to etherpad 1.9.5 https://review.opendev.org/c/opendev/system-config/+/904231 | 06:55 |
| tonyb | #status log etherpad services upgrade complete. | 07:02 |
| opendevstatus | tonyb: finished logging | 07:02 |
| tonyb | infra-root: backups pruned. 311Gb reclaimed | 07:16 |
| opendevreview | Jan Marchel proposed openstack/project-config master: Add NebulOuS monitoring-data-persistor and exponential-smoothing-predictor repo https://review.opendev.org/c/openstack/project-config/+/904681 | 08:55 |
| opendevreview | Jan Marchel proposed openstack/project-config master: Add NebulOuS monitoring-data-persistor and exponential-smoothing-predictor repo https://review.opendev.org/c/openstack/project-config/+/904681 | 10:40 |
| opendevreview | Jan Marchel proposed openstack/project-config master: Add NebulOuS monitoring-data-persistor and exponential-smoothing-predictor repo https://review.opendev.org/c/openstack/project-config/+/904681 | 11:33 |
| opendevreview | Merged openstack/project-config master: Remove legacy-tempest-dsvm-nova-libvirt-kvm-apr https://review.opendev.org/c/openstack/project-config/+/904570 | 12:52 |
| fungi | thanks tonyb! | 13:35 |
| opendevreview | Merged openstack/project-config master: Add NebulOuS monitoring-data-persistor and exponential-smoothing-predictor repo https://review.opendev.org/c/openstack/project-config/+/904681 | 14:14 |
| opendevreview | Bernhard Berg proposed zuul/zuul-jobs master: prepare-workspace-git: Add ability to define synced pojects https://review.opendev.org/c/zuul/zuul-jobs/+/887917 | 14:14 |
| frickler | interesting take regarding curl and AI, regarding the recent ML topic. in particular I think the point of ppl using LLM as translator tool seems relevant https://daniel.haxx.se/blog/2024/01/02/the-i-in-llm-stands-for-intelligence/ | 15:16 |
| fungi | yes, i'm likely to be drafting some policy for openstack's mailing lists soon, given the plea i sent to openstack-discuss on the matter seems to have gone unheeded. i too am aware that llm has become a popular way for people who lack confidence in their english to refine spelling, grammar and sentence structure for their messages, and i don't want to discourage that. my primary concern is lack | 15:30 |
| fungi | of citation for important details which do not come from the author's own personal experience: that's essentially plagiarism and professional fraud | 15:30 |
| fungi | and giving people advice pasted from an llm when you don't actually have experience to know that it's accurate can be very dangerous, potentially disastrous for users who assume the author knows what they're talking about and then follows bogus advice | 15:32 |
| frickler | yes, full ack to that | 15:42 |
| Clark[m] | The gerrit community meeting is a week later than usual this month. /me goes back to quiet morning start | 16:02 |
| clarkb | https://172.99.75.44:3081/opendev/system-config is the newly held gitea node | 16:46 |
| clarkb | https://172.99.75.44:3081/explore/code?q=gitea&t= and now search works | 16:46 |
| clarkb | I'm going to work on a new testinfra test to check for the 500 status page content in the responses | 16:46 |
| opendevreview | Clark Boylan proposed opendev/system-config master: Check for gitea template rendering errors https://review.opendev.org/c/opendev/system-config/+/904777 | 17:02 |
| clarkb | there is some automated testing for the thing tonyb found last night. Not comprehensive but I figure better than nothing for now | 17:03 |
| clarkb | infra-root: should we approve https://review.opendev.org/c/opendev/system-config/+/897679/ ? I can be around to help keep an eye on it | 17:03 |
| fungi | clarkb: yes, i'll approve it now, just didn't want to until more people were around | 17:12 |
| clarkb | sounds good | 17:12 |
| opendevreview | Julia Kreger proposed openstack/diskimage-builder master: Enable the no_timer_check option to be removable https://review.opendev.org/c/openstack/diskimage-builder/+/904780 | 17:38 |
| opendevreview | Jan Marchel proposed openstack/project-config master: Add new NebulOuS projects: overlay-network-manager, security-manager https://review.opendev.org/c/openstack/project-config/+/904792 | 18:43 |
| opendevreview | Merged opendev/system-config master: Update gitea to 1.21.3 https://review.opendev.org/c/opendev/system-config/+/897679 | 18:43 |
| clarkb | the deploy job is running but hasn't gotten to the point where it has updated anything yet | 18:46 |
| clarkb | https://gitea09.opendev.org:3081/opendev/system-config is updated now | 18:47 |
| clarkb | the update is a little slow due to some db migrations and bleve code index rebuilding, but seems to be working | 18:47 |
| clarkb | git clone works for me against gitea09 as well as the web service | 18:48 |
| fungi | lgtm too | 18:49 |
| clarkb | we have updated through 11 now | 18:51 |
| clarkb | only 14 remains | 18:55 |
| clarkb | all done and the job reports success as well | 18:58 |
| clarkb | I think I'm happy with the early results here | 19:00 |
| fungi | yep, seems to be fine | 19:00 |
| clarkb | https://review.opendev.org/c/opendev/system-config/+/904777 is a good followup to try and catch those "500 errors" automatically in the future | 19:01 |
| clarkb | I guess I could rewrite that to do a for loop over some urls | 19:04 |
| * clarkb makes that update | 19:04 | |
| opendevreview | Clark Boylan proposed opendev/system-config master: Check for gitea template rendering errors https://review.opendev.org/c/opendev/system-config/+/904777 | 19:08 |
| clarkb | that should make it simpler to add paths to check over time and/or update the template content to check for | 19:08 |
| clarkb | unrelated: there is a discussion on the gerrit mailing list about automatically checking for and detecting plugin compatibility with gerrit proper on gerrit startup. Apparently it isn't common for people to build gerrit the way we do and rebuild everything together to ensure compatibility of api versions | 19:09 |
| clarkb | I'll drop the held nodes for etherpad and gitea upgrade checking at the end of my day if all is well | 19:10 |
| opendevreview | Merged openstack/project-config master: Add 2 new projects to the airship namespace (armada-go & operator) https://review.opendev.org/c/openstack/project-config/+/904239 | 20:06 |
| opendevreview | Julia Kreger proposed openstack/diskimage-builder master: Enable the no_timer_check option to be removable https://review.opendev.org/c/openstack/diskimage-builder/+/904780 | 20:44 |
| clarkb | fwiw I've confirmed that replication is working post gitea upgrade now as well | 21:50 |
| fungi | great, thanks! | 21:51 |
| tonyb | awesome. thanks all. | 22:53 |
| tonyb | clarkb: your additional testing looks good. I want to do some verification on the held node before I +2 it | 22:55 |
| clarkb | ack I won't delete the held node just yet then. | 23:08 |
| opendevreview | Merged openstack/diskimage-builder master: Enable the no_timer_check option to be removable https://review.opendev.org/c/openstack/diskimage-builder/+/904780 | 23:12 |
| *** mtreinish_ is now known as mtreinish | 23:43 | |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!
