| @fungicide:matrix.org | Clark: the zuul community call with the openinfra board of directors is still on for 16:00 utc today, right? | 14:57 |
|---|---|---|
| @clarkb:matrix.org | Last I heard it was still on | 15:02 |
| @clarkb:matrix.org | That is in about an hour | 15:02 |
| @fungicide:matrix.org | yep, just making sure. thanks for confirming! | 15:03 |
| @clarkb:matrix.org | I've started putting potential conversation points on the etherpad. Feel free to update with other ideas. | 15:18 |
| @clarkb:matrix.org | Zuulians https://review.opendev.org/c/zuul/zuul-jobs/+/855402 has the potential for speeding up a number of OpenDev jobs, but also has the potential to have wide ranging impacts if I got something wrong. I'd appreciate careful review as a result, but also feedback on how you think we should appraoch merging that (just send it in and be prepared with a revert?) | 17:40 |
| -@gerrit:opendev.org- MICHAEL KELLY proposed: | 18:43 | |
| - [zuul/zuul-operator] 853585: kopf: Upgrade requirement to 1.35.6 https://review.opendev.org/c/zuul/zuul-operator/+/853585 | ||
| - [zuul/zuul-operator] 853586: Only listen for updates to known secrets https://review.opendev.org/c/zuul/zuul-operator/+/853586 | ||
| - [zuul/zuul-operator] 853592: Allow the specification of storageClassName in PVCs https://review.opendev.org/c/zuul/zuul-operator/+/853592 | ||
| - [zuul/zuul-operator] 853695: Prefix zuul-specific resources with instance name https://review.opendev.org/c/zuul/zuul-operator/+/853695 | ||
| - [zuul/zuul-operator] 853696: Prefix nodepool specific resources with instance name https://review.opendev.org/c/zuul/zuul-operator/+/853696 | ||
| @iwienand:matrix.org | corvus: no particular rush, but https://review.opendev.org/q/topic:zuul-stream-non-string i reworked from your prior comment to put the test in the remote tests. this is catching errors in the zuul_stream.py callback that otherwise go unnoticed. it might be good to do before upgrading ansible, in case of any other issues that causes. | 20:11 |
| @iwienand:matrix.org | zuul-maint: the other one that would be good before ansible ugprades is https://review.opendev.org/c/zuul/zuul/+/853208 which runs zuul_console/stream testing against a 2.7 environment in a container. this will ensure if we do change anything, we don't do something like i did previously and write 3.6+ only code and break older platforms like centos-7 | 20:13 |
| @iwienand:matrix.org | and if anyone is in the mood to review web things; what started out as me noticing a problem with the package: output led to medium-sized stack https://review.opendev.org/q/topic:web-results-strings which converts the console page to PF4, and then makes a few other updates to the modal it shows, etc. thoughts welcome; that stack is in a final state for review atm | 20:16 |
| @iwienand:matrix.org | the only hackish bit there is padding out the data list when mixing expandable sections. i am in disucssion with patternfly about how we can do that better in https://github.com/patternfly/patternfly/issues/5055 | 20:17 |
| @iwienand:matrix.org | that has a few screenshots that show what i mean | 20:18 |
| @clarkb:matrix.org | ianw: on 853208 it looks like the latest patchset brings the debugging _inventory register back. Was that intentional? | 20:36 |
| @iwienand:matrix.org | Clark: ahh, i would say "no" a priori ... let me look, thanks | 20:37 |
| -@gerrit:opendev.org- Ian Wienand proposed: [zuul/zuul] 853208: zuul-stream : Test against a Python 2.7 container https://review.opendev.org/c/zuul/zuul/+/853208 | 20:39 | |
| @iwienand:matrix.org | Clark: ^ thanks ... as I always say, thiry-twoth revision is the charm :) | 20:40 |
| @iwienand:matrix.org | that variable was one of the things that made me have to do so many revisions. turns out it is not safe to use | 20:41 |
| -@gerrit:opendev.org- James E. Blair https://matrix.to/#/@jim:acmegating.com proposed: [zuul/zuul] 856210: Log more info on gerrit 403 errors https://review.opendev.org/c/zuul/zuul/+/856210 | 22:04 | |
| @jim:acmegating.com | ianw: stream topic has +2s and a comment from Clark ; i'll leave it to you to approve | 22:08 |
| @jim:acmegating.com | ianwClark some thoughts on https://review.opendev.org/853208 | 22:19 |
| @jim:acmegating.com | Clark: re testing 855402 -- if that's used in an untrusted playbook, you could triple check it with depends-on. otherwise, i think we can rely on the functional testing and merge it (and fast-revert if it's an oops). | 22:25 |
| @clarkb:matrix.org | that is a good point re depends on. Let me go see how the job that precipitated all this uses it | 22:25 |
| @jim:acmegating.com | (it's the really low level base-job stuff that we can't effectively functional test that i think needs the base-test cycle; i think if we have good func tests we can rely on those) | 22:25 |
| @clarkb:matrix.org | I think this is untrusted in devstack actually. Since we stage it all using stage-output without extra privs then fetch-output grabs it in the base job which is trusted. Let me push a devstack noop change with a depends on and see how it does | 22:27 |
| @clarkb:matrix.org | https://review.opendev.org/c/openstack/devstack/+/856211 has been pushed to test this | 22:31 |
| -@gerrit:opendev.org- Ian Wienand proposed: [zuul/zuul] 856213: zuul_stream: clarify testing environment variables https://review.opendev.org/c/zuul/zuul/+/856213 | 22:46 | |
| @iwienand:matrix.org | corvus: thanks; ^ enhances that comment (or lack thereof). I mentioned in the change I'm a bit hesitant to change it more. Do you think a separate docker network might work the same? I'm not sure if we use that anywhere. it seems we just want an ip that is !127.0.0.1 ... | 22:57 |
| @iwienand:matrix.org | i wonder if adding 127.0.0.3 might just work too...? | 22:58 |
| -@gerrit:opendev.org- Clark Boylan proposed: [zuul/zuul-jobs] 855402: Speed up log file fetching tasks https://review.opendev.org/c/zuul/zuul-jobs/+/855402 | 23:02 | |
| @clarkb:matrix.org | corvus: that was an excellent idea. It already caught one issue :) I've pushed a fix and rechecked the devsatck change | 23:03 |
| -@gerrit:opendev.org- Ian Wienand proposed: [zuul/zuul] 856214: [wip] zuul-stream : try using !127.0.0.1 for loopback https://review.opendev.org/c/zuul/zuul/+/856214 | 23:05 | |
| @jim:acmegating.com | ianw: yeah maybe .3 would work? if so, that'd be great. if not, i'm good with the change as-is. | 23:07 |
| @jim:acmegating.com | ianw: ++ all around -- i didn't +w until we see how 856214 shakes out | 23:10 |
| @iwienand:matrix.org | thanks | 23:10 |
| @jim:acmegating.com | Clark: should we update the func test? | 23:11 |
| @clarkb:matrix.org | corvus: oh yes we probably should. I'll push an update for that once I get some info from the current set of devstack jobs | 23:11 |
| @jim:acmegating.com | (or maybe unit tests, now that i see the error) | 23:11 |
| -@gerrit:opendev.org- Ian Wienand proposed: [zuul/zuul] 856214: [wip] zuul-stream : try using !127.0.0.1 for loopback https://review.opendev.org/c/zuul/zuul/+/856214 | 23:18 | |
| @clarkb:matrix.org | I'm doing both. The functional test is nice because if split into a parent change will ensure the old and new behavior is consistent | 23:18 |
| -@gerrit:opendev.org- Clark Boylan proposed: | 23:36 | |
| - [zuul/zuul-jobs] 855402: Speed up log file fetching tasks https://review.opendev.org/c/zuul/zuul-jobs/+/855402 | ||
| - [zuul/zuul-jobs] 856215: Improve stage-output functional test https://review.opendev.org/c/zuul/zuul-jobs/+/856215 | ||
| @clarkb:matrix.org | corvus: ^ devstack seemed happy with that last update so I've pushed the built in testing update. Split it up so that we can confirm the old behavior is preserved. The old regex stuff isn't super readable but I think I got it right | 23:36 |
| @clarkb:matrix.org | In the base devstack job this ~one minute at https://zuul.opendev.org/t/openstack/build/36cf7d815acf4661be0231e639105287/log/job-output.txt#20968-21032 became 1.2 seconds due to my change | 23:41 |
| @clarkb:matrix.org | and that gap widens as more complicated devsatck based jobs add more files | 23:42 |
| @jim:acmegating.com | Clark: this is one of those "saved millions of hours of cpu time" changes :) | 23:52 |
| @jim:acmegating.com | also, millions of hours of developer time, assuming they're all watching the console log scroll by :) | 23:53 |
| @iwienand:matrix.org | > <@clarkb:matrix.org> In the base devstack job this ~one minute at https://zuul.opendev.org/t/openstack/build/36cf7d815acf4661be0231e639105287/log/job-output.txt#20968-21032 became 1.2 seconds due to my change | 23:53 |
| now you just have to write it in rust to save that last second :) | ||
| @jim:acmegating.com | technically as an ansible module, it could be | 23:54 |
| @jim:acmegating.com | bust most of that second is probably the ansible tqm | 23:54 |
| @jim:acmegating.com | * but most of that second is probably the ansible tqm | 23:54 |
| -@gerrit:opendev.org- Ian Wienand proposed: [zuul/zuul] 856214: [wip] zuul-stream : try using !127.0.0.1 for loopback https://review.opendev.org/c/zuul/zuul/+/856214 | 23:58 | |
| @clarkb:matrix.org | Ya it might be a good idea to skim jobs like that for other unnecessarily slow tasks | 23:59 |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!