| *** dkranz has quit IRC | 01:11 | |
| *** swest has quit IRC | 01:44 | |
| *** swest has joined #zuul | 01:59 | |
| *** swest has quit IRC | 04:18 | |
| *** swest has joined #zuul | 05:09 | |
| *** swest1 has joined #zuul | 05:12 | |
| *** swest1 has quit IRC | 05:14 | |
| *** swest has quit IRC | 05:14 | |
| *** swest has joined #zuul | 05:15 | |
| *** hashar has joined #zuul | 06:02 | |
| *** AJaeger has quit IRC | 06:12 | |
| *** jpena|off is now known as jpena | 07:10 | |
| *** xinliang has joined #zuul | 07:14 | |
| *** xinliang has quit IRC | 07:14 | |
| *** xinliang has joined #zuul | 07:14 | |
| *** nguyenhai has joined #zuul | 07:54 | |
| *** threestrands has quit IRC | 08:17 | |
| *** ssbarnea_ has joined #zuul | 08:27 | |
| *** sshnaidm|off is now known as sshnaidm | 08:41 | |
| *** AJaeger has joined #zuul | 09:38 | |
| *** sshnaidm has quit IRC | 10:10 | |
| *** sshnaidm has joined #zuul | 11:03 | |
| *** trishnag has quit IRC | 11:06 | |
| *** trishnag has joined #zuul | 11:24 | |
| *** jpena is now known as jpena|lunch | 11:33 | |
| *** pwhalen has quit IRC | 11:42 | |
| openstackgerrit | Andrea Frittoli proposed openstack-infra/zuul-jobs master: Deploy ssh key as root for non-root users https://review.openstack.org/563584 | 11:43 |
|---|---|---|
| *** elyezer has joined #zuul | 12:06 | |
| openstackgerrit | Andrea Frittoli proposed openstack-infra/zuul-jobs master: Deploy ssh key as root for non-root users https://review.openstack.org/563584 | 12:07 |
| *** pwhalen has joined #zuul | 12:14 | |
| *** pwhalen has joined #zuul | 12:14 | |
| *** elyezer has quit IRC | 12:16 | |
| *** rlandy has joined #zuul | 12:19 | |
| *** jpena|lunch is now known as jpena | 12:27 | |
| *** elyezer has joined #zuul | 12:28 | |
| *** dkranz has joined #zuul | 12:35 | |
| *** dmsimard|off is now known as dmsimard | 12:51 | |
| *** pwhalen_ has joined #zuul | 14:32 | |
| *** pwhalen has quit IRC | 14:32 | |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul master: Make config objects freezable https://review.openstack.org/562816 | 14:34 |
| corvus | clarkb, tobiash: that ^ should pass tests now and be ready for real review. | 14:35 |
| *** pwhalen_ is now known as pwhalen | 14:35 | |
| *** pwhalen has quit IRC | 14:35 | |
| *** pwhalen has joined #zuul | 14:35 | |
| corvus | there's a chance that will expose a hole in our testing when we roll that out to a production system. well, there's always a chance, but this one seems like a slightly higher than normal chance. :) | 14:36 |
| tobiash | corvus: ok, will look later at home, currently leaving office | 14:38 |
| corvus | tobiash: no rush | 14:38 |
| clarkb | corvus: https://groups.google.com/forum/#!topic/repo-discuss/O2KIkHEoU38 is an interesting gerrit ml thread | 14:59 |
| clarkb | tl'dr they need zuul | 14:59 |
| *** sshnaidm is now known as sshnaidm|afk | 15:41 | |
| corvus | maybe we should reply to that after the zuul simulation video lands on the website... | 15:43 |
| *** jpena is now known as jpena|brb | 15:43 | |
| *** sshnaidm|afk has quit IRC | 15:46 | |
| *** sshnaidm has joined #zuul | 15:48 | |
| *** sshnaidm has quit IRC | 15:49 | |
| corvus | i just read through the configloader, and i think we've severed all the dirrect references between the different object types. so they're all referenced by name now. | 15:56 |
| corvus | i think that means i can start in on caching the objects themselves instead of the yaml. | 15:57 |
| Shrews | corvus: where are we on the stestr switch? wondering if it's worth fixing that up for mordred for nodepool to get it merged | 16:04 |
| Shrews | i think there were issues with the zuul patch, yeah? | 16:04 |
| Shrews | or maybe that merged already | 16:05 |
| Shrews | ah, merged then reverted | 16:06 |
| corvus | Shrews: i think it merged, was reverted, and maybe hasn't been re-proposed? i think all the issues have been worked out and should be straightforward... | 16:06 |
| Shrews | corvus: i can't find what the actual issues were, but i'll start poking at the nodepool change | 16:07 |
| corvus | Shrews: istr at least one of them was operator error :). there were some other minor things which i don't remember except i think are probably easily resolvable enough if we just want to try it again. | 16:10 |
| *** jpena|brb is now known as jpena | 16:22 | |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul master: Remove layout from ParseContext https://review.openstack.org/563695 | 16:31 |
| *** jpena is now known as jpena|off | 16:50 | |
| *** sshnaidm has joined #zuul | 17:00 | |
| *** sshnaidm is now known as sshnaidm|off | 17:02 | |
| openstackgerrit | Clark Boylan proposed openstack-infra/zuul-jobs master: Better zuul debug info debugging info https://review.openstack.org/561758 | 17:17 |
| tobiash | andreaf: commented on https://review.openstack.org/563584 | 17:19 |
| SpamapS | 2018-04-21 00:04:11.568609 | godaddy-c7 | TASK [Gathering Facts] ********************************************************* | 17:24 |
| SpamapS | 2018-04-21 00:04:27.056581 | godaddy-c7 | ok: [zuul0] | 17:24 |
| SpamapS | derp | 17:25 |
| SpamapS | That setup module is just getting too aggressive | 17:25 |
| SpamapS | who has 16 seconds? :-P | 17:25 |
| clarkb | SpamapS: I think we really only need ti for a tiny fraction of the facts too | 17:26 |
| SpamapS | Actually I pasted that just as I was mis-reading the seconds as minutes | 17:28 |
| SpamapS | I had a job timeout which surprised me and thought it was a 16 minute setup | 17:28 |
| SpamapS | In fact though, the actual thing that took forever was more confusing than that | 17:29 |
| SpamapS | http://paste.openstack.org/show/719759/ | 17:29 |
| SpamapS | Guessing whatever VM I was on had a long hypervisor pause or something. | 17:29 |
| SpamapS | or ssh got routed through a 300 baud modem | 17:29 |
| Shrews | as someone who owned a 300 baud modem, i resemble that remark | 17:31 |
| Shrews | oh the bbs's | 17:31 |
| SpamapS | 16800 zyxel ftw | 17:39 |
| SpamapS | "Screw standards, let's bet this whole BBS world on a brief flash in the pan" | 17:39 |
| tobiash | Shrews, SpamapS thanks, now I feel young :) | 17:52 |
| corvus | tobiash: this will help you feel old: https://www.telnetbbsguide.com/ | 17:54 |
| corvus | must....close....tab.... | 17:58 |
| Shrews | heh, i remember the 1st one I joined required a telephone call from the bbs owner to verify my identity. i must have been about 11 or 12 and this was well before "beware Internet strangers" | 17:58 |
| pabelanger | cool, start playing LORD and VGA planets again | 17:58 |
| corvus | Shrews: one of the first ones i called had a irl bbq every month or so | 17:59 |
| Shrews | corvus: it was obviously before computers were used to AVOID other people :) | 18:00 |
| corvus | Shrews: heh. either that, or we were just doing it wrong :) | 18:00 |
| * SpamapS lost a couple days to trade wars a few months ago ;) | 18:01 | |
| tobiash | corvus: I still feel young ;) | 18:01 |
| corvus | SpamapS: i lost years to tw2002, what's a few more days? | 18:02 |
| SpamapS | so true | 18:02 |
| openstackgerrit | David Shrewsbury proposed openstack-infra/nodepool master: Switch to stestr https://review.openstack.org/536862 | 18:08 |
| Shrews | corvus: so, hrm... now that I've rebased that, I can't seem to get ttrun to output like it used to | 18:12 |
| Shrews | but the tests will run | 18:12 |
| Shrews | perhaps this was the operator error you found before | 18:13 |
| Shrews | ? | 18:13 |
| corvus | Shrews: the, erm, operator error i recall was that i was trying to run a test that didn't exist on the branch i was on (because i switched branches to test out stestr). maybe you need to recreate the venv or something? | 18:15 |
| corvus | Shrews: that was certainly the symptom i had though :) | 18:15 |
| Shrews | i did that with 'tox -epy36 -r' | 18:16 |
| corvus | Shrews: do we have a py36? | 18:16 |
| Shrews | i do | 18:16 |
| Shrews | there's no target in tox.ini | 18:17 |
| corvus | i guess it should work | 18:17 |
| clarkb | pyxy targets are implicit in tox | 18:17 |
| Shrews | if you have python3.5 installed, tox will give you a py35 | 18:17 |
| clarkb | and will just be based off the baseenv if not explicitly defined | 18:17 |
| corvus | just scraping the bottom of the barrel on ideas here :) | 18:17 |
| clarkb | https://review.openstack.org/#/c/553618/16 is quite the daunting diff to review | 18:18 |
| corvus | clarkb: yeah, sorry about that. i'm trying to do it in as small steps as possible; but i don't think i can split that up further (at least, not without major temporary scaffolding or something) | 18:19 |
| clarkb | ya I'll just have to dig through it | 18:19 |
| clarkb | I've done the earlier changes in the stack but not approved them as I get the sort of sense this whole thing should go in together so the end result is consistent quickly | 18:20 |
| clarkb | (rather than being in halfway state for a time) | 18:20 |
| Shrews | mordred: when you are around, I left a question for you on 536862 | 18:20 |
| corvus | theoretically they could go in at any point, but there are a couple of places where it would make sense to merge a batch of changes | 18:21 |
| corvus | merging up to 553618 or 562816 would be good i think. | 18:21 |
| clarkb | Shrews: ttrun basically just runs python -m testtools.run $testnames iirc | 18:21 |
| clarkb | Shrews: you probably want to strace it or find where it does that exec to amke sure it is execing the correct things | 18:22 |
| clarkb | (that should be completely independent of stestr and testr too,) | 18:22 |
| *** harlowja has joined #zuul | 18:44 | |
| Shrews | ianw: Did you ever create an equivalent logging change for Zuul to match https://review.openstack.org/547889 as corvus requested? I'm not seeing it. | 18:45 |
| *** harlowja has quit IRC | 19:07 | |
| *** ssbarnea_ has quit IRC | 19:42 | |
| clarkb | corvus: I'm still not sure I've fully grokked https://review.openstack.org/#/c/553618/16 but I've +2'd it. I did leave a few comments, I think the second two are me just thinking out loud to myself to udnerstand that change but the first may be worth a look | 19:48 |
| corvus | clarkb: responded (with footnotes!) | 20:05 |
| clarkb | tobiash: commented on https://review.openstack.org/#/c/535713/8 there is at least one bug that needs to be fixed. I think my other comment is not a bug just harder to undersatnd (but it may be a bug if I've misunderstood it) | 20:09 |
| clarkb | corvus: your comment implies that voluptuous schema entries are not required by default? For some reason I thought they were | 20:11 |
| corvus | clarkb: ProjectTemplateParser has a 'validate' argument which ProjectParser sets to False; so we skip schema validation in ProjectTemplateParser (because ProjectParser validates the schema for it) | 20:12 |
| clarkb | aha | 20:12 |
| corvus | clarkb: (and in the schema for ProjectParser, name is optional) | 20:12 |
| corvus | projecttemplateparser is actually getting really small these days, we could probably refactor it so that projectparser doesn't delegate to it anymore | 20:13 |
| corvus | (i would like that as sometimes this is difficult to think about) | 20:13 |
| corvus | maybe just pull out the shared bit into a helper function | 20:14 |
| clarkb | corvus: line 101 https://review.openstack.org/#/c/562816/9/zuul/model.py the use of super there is so that Freezable can be used as a mixin? | 20:24 |
| corvus | clarkb: yep | 20:26 |
| corvus | tobiash has blown my mind with his comment on that change | 20:27 |
| corvus | as soon as i'm done with this, i want to change that everywhere | 20:27 |
| corvus | i hate having to type the arguments to super | 20:27 |
| clarkb | ya its a neat new thing I learned recently too | 20:28 |
| clarkb | you still have to use the old style in some cases but not these cases | 20:28 |
| clarkb | (the cases where you are overriding some MRO default) | 20:28 |
| clarkb | ok thats is me through the entire related stack that is passing. Including tobiash's regex project templates | 20:30 |
| clarkb | corvus: https://review.openstack.org/#/c/563695/1 has unhappy test | 20:30 |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul master: Remove 'base' from UnparsedAbideConfig https://review.openstack.org/563757 | 20:32 |
| corvus | clarkb: thanks; that shouldn't be too hard to fix | 20:33 |
| corvus | i'll probably leave it for later, after i get through this current attempt to reuse config objects | 20:33 |
| clarkb | I'm now poking at shrews' stestr change and tox is broken for me so first I must unbreak tox | 20:36 |
| *** dkranz has quit IRC | 20:44 | |
| clarkb | Shrews: ttrun worked for me and py36. But I got an interesting error message. | 20:49 |
| clarkb | /usr/lib64/python3.6/runpy.py:125: RuntimeWarning: 'testtools.run' found in sys.modules after import of package 'testtools', but prior to execution of 'testtools.run'; this may result in unpredictable behaviour | 20:49 |
| clarkb | s/error/warning/ | 20:49 |
| Shrews | clarkb: by "worked" you mean you saw the debug logging output? | 20:50 |
| clarkb | Shrews: oh is that specifically what we are looking for? For that I think you need to flip the env var over to emit it. I'll try that | 20:50 |
| Shrews | clarkb: yes. see my latest comment on that review | 20:51 |
| clarkb | Shrews: ya I had to set the three OS_*_CAPTURE env vars to 0 | 20:53 |
| clarkb | Shrews: the change there updates them to default to true (1) rather than false. With testr we had that inverted. They defaulted to false and testr set them to true in its config file | 20:53 |
| corvus | so testr would capture the output, but ttrun would not | 20:54 |
| corvus | now they both do? | 20:54 |
| Shrews | i'm not sure i like that | 20:54 |
| clarkb | corvus: correct | 20:55 |
| clarkb | I'm looking to see if stestr can set env vars in a similar manner | 20:55 |
| Shrews | why don't we just set them in tox.ini? | 20:55 |
| clarkb | ya that would be one way of doing it | 20:56 |
| clarkb | sincei t is just going to run stestr anyways | 20:56 |
| corvus | Shrews, clarkb: yeah, i think that's what the zuul change did. ++ | 20:56 |
| corvus | oh, though i guess that means that "stestr run" will spew output | 20:57 |
| corvus | basically, i think tox and 'testr run' should be quiet, and ttrun should be verbose | 20:57 |
| clarkb | confirmed stestr run will spew output in that case | 20:58 |
| Shrews | is that a problem though? | 20:58 |
| clarkb | it can be quite confusing to people to have interleaved test output emit that way | 20:59 |
| Shrews | i only ever call tox or ttrun | 20:59 |
| corvus | Shrews: i don't do it as much as ttrun, but i do still run the unit tests locally -- i used 'testr run --failing' a lot last week actually | 20:59 |
| Shrews | i see | 20:59 |
| Shrews | that's a valid use case | 21:00 |
| clarkb | ya not finding a way to do this in stestr itself | 21:04 |
| clarkb | feature request to mtreinish maybe | 21:04 |
| clarkb | Shrews: corvus another option could be to inspect argv[0] in the test run to see how it was invoked | 21:09 |
| clarkb | if running under ttrun then flip that value | 21:09 |
| Shrews | might be just as easy to add the env var feature to ttrun via a new config file | 21:17 |
| corvus | i heard that stestr had something like ttrun; i wonder how that fits in | 21:18 |
| *** hashar has quit IRC | 21:54 | |
| clarkb | before I forget, it would be nice to try and do another zuul reelase soon I think | 22:22 |
| clarkb | there are a few user facing improvments that we should get out there in addition to a bunch of small bugfixes | 22:22 |
| *** threestrands has joined #zuul | 22:41 | |
| *** threestrands has quit IRC | 22:41 | |
| *** threestrands has joined #zuul | 22:41 | |
| corvus | clarkb: ++ should we restart openstack-infra to beta test it? | 22:53 |
| openstackgerrit | Clark Boylan proposed openstack-infra/zuul-jobs master: Split logging of inventory to its own role https://review.openstack.org/563787 | 22:55 |
| clarkb | corvus: ya probably | 23:00 |
| openstackgerrit | Clark Boylan proposed openstack-infra/zuul-jobs master: Stop logging inventory in validate-host https://review.openstack.org/563790 | 23:02 |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!