| *** rascasoft has joined #openstack-infra | 00:04 | |
| *** jamesmcarthur has joined #openstack-infra | 00:04 | |
| *** lujinluo has joined #openstack-infra | 00:05 | |
| *** _alastor_ has quit IRC | 00:06 | |
| *** jamesmcarthur has quit IRC | 00:08 | |
| *** lujinluo has quit IRC | 00:09 | |
| *** tpsilva has joined #openstack-infra | 00:09 | |
| *** rascasoft has quit IRC | 00:10 | |
| mtreinish | mordred: heh, I don't know. I just spent several hours today trying to get openstack-health deployed via helm charts that I wrote 4 months ago (so I didn't remember all the details) before I just gave up and did it manually in a vm by hand :) | 00:11 |
|---|---|---|
| *** rtjure has quit IRC | 00:14 | |
| *** wolverineav has quit IRC | 00:15 | |
| *** wolverineav has joined #openstack-infra | 00:15 | |
| *** Swami has quit IRC | 00:16 | |
| *** bobh has joined #openstack-infra | 00:17 | |
| *** wolverineav has quit IRC | 00:20 | |
| *** bobh has quit IRC | 00:22 | |
| *** rtjure has joined #openstack-infra | 00:22 | |
| *** bobh has joined #openstack-infra | 00:27 | |
| *** tosky has quit IRC | 00:28 | |
| *** wolverineav has joined #openstack-infra | 00:38 | |
| *** wolverineav has quit IRC | 00:43 | |
| openstackgerrit | James E. Blair proposed openstack-infra/system-config master: Add gitea k8s resource definitions and playbook https://review.openstack.org/626759 | 00:45 |
| corvus | fungi, mordred: ^ that update creates an initial admin user in gitea -- i think we just might be able to have this thing fully bootstrapped by automation :) | 00:45 |
| *** wolverineav has joined #openstack-infra | 00:46 | |
| fungi | that's one up on gerrit at any rate | 00:46 |
| corvus | well, actually, we've also figured out how to do that for gerrit (via the zuul quickstart) | 00:46 |
| fungi | ahh | 00:46 |
| fungi | tl;dr on how it creates an initial admin account for gerrit? | 00:46 |
| corvus | gimme a minute :) | 00:49 |
| *** hwoarang has quit IRC | 00:49 | |
| fungi | not important, i was curious, just not curious enough to dissect the test setup | 00:50 |
| *** wolverineav has quit IRC | 00:50 | |
| *** jamesmcarthur has joined #openstack-infra | 00:50 | |
| *** hwoarang has joined #openstack-infra | 00:50 | |
| corvus | fungi: oh, hrm. i think it may rely on the 'become' mode. | 00:51 |
| corvus | so maybe not quite what we need yet | 00:51 |
| fungi | ahh, yeah that'd work for a test setup | 00:51 |
| corvus | really good chance that with notedb we could commit the admin user out of band | 00:52 |
| fungi | fwiw, i know a way to do it with gerrit by leveraging the builtin superuser account through the ssh api, but it's obscure and i don't know if it'll continue to work in future versions | 00:52 |
| corvus | yeah -- for some reason that didn't work for me with the zuul quickstart, but i don't recall why (the error was rather obscure, and may have had to do with an interaction with the become plugin) | 00:53 |
| corvus | or, as you say, it may just be broken in the newer releases | 00:54 |
| *** wolverineav has joined #openstack-infra | 00:56 | |
| *** wolverineav has quit IRC | 00:56 | |
| *** wolverineav has joined #openstack-infra | 00:56 | |
| openstackgerrit | James E. Blair proposed openstack-infra/system-config master: Add jinja-init Dockerfile https://review.openstack.org/626626 | 00:58 |
| *** jamesmcarthur has quit IRC | 01:04 | |
| *** sthussey has quit IRC | 01:20 | |
| *** mriedem has quit IRC | 01:20 | |
| *** rh-jelabarre has joined #openstack-infra | 01:24 | |
| *** rascasoft has joined #openstack-infra | 01:31 | |
| *** wolverineav has quit IRC | 01:31 | |
| *** wolverineav has joined #openstack-infra | 01:33 | |
| *** jamesmcarthur has joined #openstack-infra | 01:35 | |
| *** rascasoft has quit IRC | 01:37 | |
| *** wolverineav has quit IRC | 01:37 | |
| *** jamesmcarthur has quit IRC | 01:42 | |
| *** dave-mccowan has quit IRC | 01:50 | |
| *** bhavikdbavishi has joined #openstack-infra | 01:52 | |
| *** armax has quit IRC | 01:59 | |
| *** gyee has quit IRC | 01:59 | |
| *** hongbin has joined #openstack-infra | 02:00 | |
| *** tpsilva has quit IRC | 02:19 | |
| *** rascasoft has joined #openstack-infra | 02:23 | |
| *** rascasoft has quit IRC | 02:31 | |
| *** jamesmcarthur has joined #openstack-infra | 02:32 | |
| *** bhavikdbavishi has quit IRC | 02:35 | |
| *** wuchongyao has joined #openstack-infra | 02:37 | |
| wuchongyao | The external CI systemc reports error "SSHException: not a vaild EC private key file" | 02:39 |
| wuchongyao | I post my public key to openstack-infra email list ,but no response | 02:39 |
| *** wolverineav has joined #openstack-infra | 02:42 | |
| *** wolverineav has quit IRC | 02:47 | |
| *** mrsoul has quit IRC | 02:58 | |
| *** ykarel|away has joined #openstack-infra | 02:59 | |
| *** jamesmcarthur has quit IRC | 03:03 | |
| *** jamesmcarthur has joined #openstack-infra | 03:06 | |
| *** hamzy_ has quit IRC | 03:10 | |
| *** _alastor_ has joined #openstack-infra | 03:10 | |
| *** rascasoft has joined #openstack-infra | 03:12 | |
| *** _alastor_ has quit IRC | 03:15 | |
| *** psachin has joined #openstack-infra | 03:20 | |
| *** rascasoft has quit IRC | 03:22 | |
| *** jamesmcarthur has quit IRC | 03:24 | |
| *** jamesmcarthur has joined #openstack-infra | 03:25 | |
| *** wuchongyao has quit IRC | 03:29 | |
| *** bhavikdbavishi has joined #openstack-infra | 03:31 | |
| *** jamesmcarthur has quit IRC | 03:48 | |
| *** jamesmcarthur has joined #openstack-infra | 03:48 | |
| *** lbragstad has quit IRC | 03:52 | |
| *** bobh has quit IRC | 03:55 | |
| *** bhavikdbavishi has quit IRC | 04:02 | |
| *** udesale has joined #openstack-infra | 04:05 | |
| *** hwoarang has quit IRC | 04:16 | |
| *** hwoarang has joined #openstack-infra | 04:16 | |
| *** eernst has joined #openstack-infra | 04:22 | |
| *** eernst has quit IRC | 04:24 | |
| *** rh-jelabarre has quit IRC | 04:25 | |
| *** jamesmcarthur has quit IRC | 04:37 | |
| *** jamesmcarthur has joined #openstack-infra | 04:39 | |
| *** janki has joined #openstack-infra | 04:44 | |
| *** bhavikdbavishi has joined #openstack-infra | 04:50 | |
| *** jamesmcarthur has quit IRC | 04:51 | |
| *** ykarel|away is now known as ykarel | 04:53 | |
| *** jamesmcarthur has joined #openstack-infra | 05:00 | |
| *** sambetts_ has quit IRC | 05:00 | |
| *** hwoarang has quit IRC | 05:03 | |
| *** hwoarang has joined #openstack-infra | 05:05 | |
| *** jamesmcarthur has quit IRC | 05:05 | |
| *** sean-k-mooney has quit IRC | 05:18 | |
| *** jamesmcarthur has joined #openstack-infra | 05:32 | |
| *** ykarel has quit IRC | 05:35 | |
| *** jamesmcarthur has quit IRC | 05:38 | |
| *** jamesmcarthur has joined #openstack-infra | 05:39 | |
| *** Douhet has left #openstack-infra | 05:40 | |
| *** jamesmcarthur has quit IRC | 05:43 | |
| *** jamesdenton has quit IRC | 05:45 | |
| *** jamesden_ has joined #openstack-infra | 05:46 | |
| *** wolverineav has joined #openstack-infra | 05:47 | |
| *** hwoarang has quit IRC | 05:49 | |
| *** hwoarang has joined #openstack-infra | 05:50 | |
| *** ykarel has joined #openstack-infra | 05:53 | |
| *** hongbin has quit IRC | 05:55 | |
| *** eernst has joined #openstack-infra | 06:02 | |
| *** bhavikdbavishi has quit IRC | 06:07 | |
| openstackgerrit | Clint 'SpamapS' Byrum proposed openstack-infra/zuul master: Interpret ANSI while streaming logs https://review.openstack.org/626817 | 06:09 |
| *** psachin has quit IRC | 06:16 | |
| *** wolverineav has quit IRC | 06:17 | |
| openstackgerrit | Surya Prakash (spsurya) proposed openstack-infra/project-config master: Add 'Review-Priority' for Kolla repos https://review.openstack.org/626824 | 06:26 |
| *** benj_- has quit IRC | 06:30 | |
| *** benj_ has joined #openstack-infra | 06:34 | |
| *** psachin has joined #openstack-infra | 06:37 | |
| *** janki has quit IRC | 06:38 | |
| *** slaweq has joined #openstack-infra | 07:03 | |
| *** eernst has quit IRC | 07:09 | |
| *** aojea has joined #openstack-infra | 07:09 | |
| *** _alastor_ has joined #openstack-infra | 07:11 | |
| *** rcernin has quit IRC | 07:16 | |
| *** _alastor_ has quit IRC | 07:16 | |
| openstackgerrit | Merged openstack-infra/project-config master: Use renamed template 'integrated-gate-py3' https://review.openstack.org/626087 | 07:19 |
| *** quiquell|off is now known as quiquell | 07:24 | |
| *** pgaxatte has joined #openstack-infra | 07:26 | |
| *** gfidente has joined #openstack-infra | 07:28 | |
| *** quiquell is now known as quiquell|brb | 07:32 | |
| *** wolverineav has joined #openstack-infra | 07:38 | |
| *** e0ne has joined #openstack-infra | 07:39 | |
| *** lujinluo has joined #openstack-infra | 07:41 | |
| *** wolverineav has quit IRC | 07:43 | |
| *** irdr has quit IRC | 07:44 | |
| *** dpawlik has joined #openstack-infra | 07:45 | |
| *** dpawlik has quit IRC | 07:48 | |
| *** dpawlik has joined #openstack-infra | 07:48 | |
| *** ykarel is now known as ykarel|lunch | 07:50 | |
| *** irdr has joined #openstack-infra | 07:50 | |
| *** bhavikdbavishi has joined #openstack-infra | 07:52 | |
| *** bhavikdbavishi has quit IRC | 07:58 | |
| *** quiquell|brb is now known as quiquell | 08:01 | |
| *** rpittau has joined #openstack-infra | 08:02 | |
| *** ginopc has joined #openstack-infra | 08:02 | |
| *** rascasoft has joined #openstack-infra | 08:11 | |
| *** pcaruana has joined #openstack-infra | 08:19 | |
| *** tosky has joined #openstack-infra | 08:27 | |
| *** e0ne_ has joined #openstack-infra | 08:29 | |
| *** e0ne_ has quit IRC | 08:33 | |
| *** e0ne__ has joined #openstack-infra | 08:33 | |
| *** e0ne has quit IRC | 08:33 | |
| *** dmellado has quit IRC | 08:44 | |
| *** ykarel|lunch is now known as ykarel | 08:45 | |
| *** dmellado has joined #openstack-infra | 08:46 | |
| *** dpawlik has quit IRC | 08:52 | |
| *** dpawlik has joined #openstack-infra | 08:53 | |
| *** jpena|off is now known as jpena | 08:53 | |
| *** tosky has quit IRC | 08:54 | |
| *** tosky has joined #openstack-infra | 08:55 | |
| dulek | Hey folks! Is there a way in multinode Zuul + DevStack to somehow put subnode IP into .conf files on controller? | 08:55 |
| *** janki has joined #openstack-infra | 08:56 | |
| *** shardy has joined #openstack-infra | 08:59 | |
| *** e0ne__ has quit IRC | 09:03 | |
| *** e0ne has joined #openstack-infra | 09:04 | |
| dmellado | AJaeger: ^^ | 09:05 |
| *** e0ne has quit IRC | 09:05 | |
| *** e0ne_ has joined #openstack-infra | 09:05 | |
| *** jpich has joined #openstack-infra | 09:06 | |
| *** bhavikdbavishi has joined #openstack-infra | 09:07 | |
| *** psachin has quit IRC | 09:11 | |
| *** bhavikdbavishi has quit IRC | 09:11 | |
| dmellado | dulek: most of the infra folks are US based, so it might've to wait ;) | 09:12 |
| *** bhavikdbavishi has joined #openstack-infra | 09:12 | |
| dmellado | yolanda isn't even here, lol | 09:12 |
| dmellado | rcarrillocruz? | 09:12 |
| dmellado | either | 09:12 |
| dmellado | booh | 09:12 |
| dmellado | tobiash: ^^ ? | 09:12 |
| tobiash | dmellado: unfortunately I don't know devstack | 09:14 |
| dmellado | tobiash: np! thanks in any case! | 09:14 |
| tobiash | the ip addresses can be found in the inventory | 09:14 |
| dmellado | dulek: I'll try to fetch these from the inventory and write a foo file for now | 09:15 |
| *** psachin has joined #openstack-infra | 09:29 | |
| *** derekh has joined #openstack-infra | 09:35 | |
| *** gmann is now known as gmann_pto | 09:43 | |
| *** dpawlik has quit IRC | 09:46 | |
| *** dpawlik has joined #openstack-infra | 09:53 | |
| AJaeger | dmellado, dulek #openstack-qa might be better for that one as it's devstack related | 09:53 |
| *** tosky has quit IRC | 09:54 | |
| *** tosky has joined #openstack-infra | 09:55 | |
| dmellado | AJaeger: thanks, we'll ask there as well | 09:56 |
| dmellado | I'm also just trying to first get the subnode IP and write them into the controller | 09:57 |
| *** spa-87 has joined #openstack-infra | 09:59 | |
| *** alexchadin has joined #openstack-infra | 10:00 | |
| *** ssbarnea has joined #openstack-infra | 10:13 | |
| alexchadin | ttx: ping | 10:14 |
| ttx | alexchadin: o/ | 10:18 |
| *** gfidente has quit IRC | 10:18 | |
| *** e0ne_ has quit IRC | 10:18 | |
| alexchadin | ttx: regarding https://review.openstack.org/#/c/619470/ | 10:18 |
| *** e0ne has joined #openstack-infra | 10:18 | |
| alexchadin | ttx: I've done my work with https://review.openstack.org/#/c/622946/ | 10:18 |
| alexchadin | ttx: could you please review it? | 10:19 |
| ttx | Nice! I'm not a grenade specialist, but I'll have a look | 10:20 |
| ttx | alexchadin: maybe you can restore the governance change and add a Depends-On on that change | 10:20 |
| ttx | that will trigger additional review attention | 10:20 |
| alexchadin | ttx: Done | 10:22 |
| *** jtomasek has joined #openstack-infra | 10:31 | |
| openstackgerrit | Thierry Carrez proposed openstack-infra/system-config master: Update doc for ptgbot to match current behavior https://review.openstack.org/626869 | 10:39 |
| *** pbourke has quit IRC | 10:40 | |
| *** pbourke has joined #openstack-infra | 10:42 | |
| *** alexchadin has quit IRC | 10:43 | |
| *** udesale has quit IRC | 10:46 | |
| *** annp has quit IRC | 10:52 | |
| *** agopi has joined #openstack-infra | 10:58 | |
| *** owalsh_ has joined #openstack-infra | 10:59 | |
| *** owalsh has quit IRC | 11:01 | |
| *** _alastor_ has joined #openstack-infra | 11:13 | |
| *** _alastor_ has quit IRC | 11:18 | |
| *** dpawlik has quit IRC | 11:41 | |
| *** bhavikdbavishi has quit IRC | 11:43 | |
| *** jonher has quit IRC | 11:46 | |
| *** dpawlik has joined #openstack-infra | 11:46 | |
| *** fresta_ has quit IRC | 11:47 | |
| *** gfidente has joined #openstack-infra | 11:52 | |
| *** jamesden_ is now known as jamesdenton | 11:53 | |
| tosky | hi all; in order to add new repositories to an existing project (sahara), do we need to follow the relevant part of the Project Creator's Guide? | 12:06 |
| tosky | namely https://docs.openstack.org/infra/manual/creators.html | 12:06 |
| tosky | when the guide talks about "projects", does it really mean "repositories"? | 12:06 |
| *** jonher has joined #openstack-infra | 12:07 | |
| tosky | also, is the pypi registration step still manual as described there? | 12:07 |
| *** fresta has joined #openstack-infra | 12:07 | |
| *** alexchadin has joined #openstack-infra | 12:10 | |
| *** jpich has quit IRC | 12:11 | |
| *** jpich has joined #openstack-infra | 12:13 | |
| openstackgerrit | Aleksey Zvyagintsev proposed openstack-infra/jeepyb master: wip https://review.openstack.org/626640 | 12:17 |
| *** jpena is now known as jpena|lunch | 12:24 | |
| openstackgerrit | Aleksey Zvyagintsev proposed openstack-infra/jeepyb master: manage_projects:Add option to exit-with-error-code https://review.openstack.org/626640 | 12:25 |
| alexz__ | fungi: would you kindly to review https://review.openstack.org/#/c/626640/? | 12:26 |
| *** panda|off is now known as panda | 12:28 | |
| *** kaiokmo has joined #openstack-infra | 12:32 | |
| frickler | tosky: the guide seems to mix the use of "project" and "repository" a bit, yes, if you have an idea how to improve that, patches are welcome ;) | 12:32 |
| frickler | tosky: it does say "If you add a new repository, you can make it part of an existing official OpenStack project, ..." though, so that should cover your situation | 12:33 |
| frickler | tosky: I'm not sure about the pypi status, though | 12:33 |
| *** e0ne has quit IRC | 12:37 | |
| *** agopi has quit IRC | 12:41 | |
| *** dpawlik has quit IRC | 12:47 | |
| *** derekh has quit IRC | 12:47 | |
| *** e0ne has joined #openstack-infra | 12:49 | |
| *** rpittau is now known as rpittau|lunch | 12:49 | |
| *** derekh has joined #openstack-infra | 12:51 | |
| *** bhavikdbavishi has joined #openstack-infra | 12:57 | |
| *** bobh has joined #openstack-infra | 12:57 | |
| *** alexchadin has quit IRC | 12:58 | |
| *** dkehn has quit IRC | 12:59 | |
| *** jpich has quit IRC | 13:01 | |
| *** jpich has joined #openstack-infra | 13:02 | |
| *** e0ne has quit IRC | 13:04 | |
| *** rlandy has joined #openstack-infra | 13:06 | |
| *** rlandy is now known as rlandy|rover | 13:07 | |
| *** dpawlik has joined #openstack-infra | 13:11 | |
| *** e0ne has joined #openstack-infra | 13:11 | |
| *** dave-mccowan has joined #openstack-infra | 13:14 | |
| *** bhavikdbavishi has quit IRC | 13:15 | |
| *** rh-jelabarre has joined #openstack-infra | 13:15 | |
| tosky | frickler: ack, thanks, at least I can plan some of the steps properly | 13:16 |
| *** jpich has quit IRC | 13:22 | |
| *** jpich has joined #openstack-infra | 13:25 | |
| *** jcoufal has joined #openstack-infra | 13:40 | |
| *** mriedem has joined #openstack-infra | 13:44 | |
| *** dkehn has joined #openstack-infra | 13:45 | |
| *** trown|outtypewww is now known as trown | 13:48 | |
| *** rpittau|lunch is now known as rpittau | 13:50 | |
| *** psachin has quit IRC | 13:53 | |
| *** jtomasek has quit IRC | 13:56 | |
| *** lbragstad has joined #openstack-infra | 13:59 | |
| *** boden has joined #openstack-infra | 14:00 | |
| *** janki has quit IRC | 14:04 | |
| *** shardy has quit IRC | 14:07 | |
| fungi | dmellado: yeah, as others have indicated, ansible inventory contents can be accessed as replacement variables in your configuration templates. naming the nodes or using the names the base devstack playbook gives them should make the template nice and maintainable | 14:07 |
| *** kgiusti has joined #openstack-infra | 14:09 | |
| fungi | tosky: use of the term "project" in the creators guide is mainly because all of gerrit, zuul, storyboard, launchpad and pypi refer to those as "projects" in their documentation, urls and configuration option names. so for example a gerrit "project" is basically a git repository plus some additional configuration and related records in a database | 14:11 |
| alexz__ | fungi: would you kindly to review https://review.openstack.org/#/c/626640/ ? | 14:12 |
| fungi | tosky: frickler: for pypi, if you want to reserve a project entry there you need to register it manually by uploading a package (it can be basically empty except for egg-info/metadata declaring a name and version, the version can even be "0"), but if the project doesn't exist at the time of the first release build our job which pushes the resulting packages to pypi will cause the project to get | 14:12 |
| fungi | created there | 14:12 |
| *** shardy has joined #openstack-infra | 14:13 | |
| fungi | alexz__: working my way through scrollback, i already saw your request | 14:13 |
| alexz__ | ok, thanks! | 14:13 |
| tosky | fungi: about pypi, does it mean that if we rely on the fact that the name won't be stolen, we can just skip the pre-registration and rely on the magic from infra scripts? | 14:14 |
| fungi | tosky: yes, basically the recommended way to "create" a project on pypi these days is to use the twine utility to upload a package for it. that is exactly what our release jobs do already, so it "just works" automagically | 14:15 |
| fungi | tosky: while you're here, are you still working on 548936 to get a grenade native zuul v3 job to replace the legacy one? or do you know if someone else is planning to pick up that effort? | 14:16 |
| tosky | fungi: I was still planning to work on it, I was "just" sidetracked to other stuff | 14:19 |
| tosky | we have few pending changes in sahara that we would like to land before M2 (like the splitting of the plugins from sahara.git, hence the request for new repositories) | 14:19 |
| *** jcoufal has quit IRC | 14:20 | |
| fungi | tosky: no worries, someone was just asking me about 622946 and that reminded me that we likely still have people adding new jobs based on legacy-dsvm-base because grenade isn't v3 native yet | 14:21 |
| tosky | yeah, I know, sorry :/ it's something I'd really like to have as well | 14:22 |
| fungi | so i went looking for the current state of that effort | 14:22 |
| fungi | don't sweat it. we can all only do as much as we can do | 14:22 |
| fungi | and there's definitely no need to be apologetic about tat | 14:23 |
| fungi | er, about that | 14:23 |
| *** _alastor_ has joined #openstack-infra | 14:25 | |
| fungi | alexz__: thanks, that looks really good! i've left a few inline comments where i think it can be further simplified | 14:29 |
| fungi | alexz__: in particular, i don't think it warrants the complexity of a command-line option. i was mostly just concerned that the implementation takes care not to abort in the middle of the loop without at least trying to process all updates | 14:30 |
| *** _alastor_ has quit IRC | 14:30 | |
| fungi | which is plenty safe to cover our use case | 14:30 |
| *** lbragstad has quit IRC | 14:31 | |
| alexz__ | fungi: i posted answer in review. in few w - it will not broke and b-w compatability, if you didn't pass new flag- exit code will be always 0 | 14:32 |
| fungi | alexz__: right, i'm saying i don't think we need to maintain backwards compatibility for the exit code. i feel like always returning nonzero when there was at least one failure is fine | 14:33 |
| *** lbragstad has joined #openstack-infra | 14:33 | |
| fungi | i consider the case of exiting 0 on failure a bug | 14:33 |
| *** jcoufal has joined #openstack-infra | 14:34 | |
| alexz__ | aah, got you. so, should i simple remove those arg at all ? or make it opposite ? | 14:35 |
| openstackgerrit | Thierry Carrez proposed openstack/ptgbot master: Pin irc module to 15.1.1 to avoid import error https://review.openstack.org/626906 | 14:35 |
| openstackgerrit | Thierry Carrez proposed openstack/ptgbot master: Generate PTGbot index page dynamically https://review.openstack.org/626907 | 14:35 |
| *** agopi has joined #openstack-infra | 14:37 | |
| fungi | alexz__: yeah, i'd just remove the command-line option and the conditional which references it | 14:39 |
| alexz__ | ok. will update in a sec | 14:40 |
| fungi | thanks! and sorry if my earlier review comments were confusnig | 14:40 |
| fungi | er, confusing | 14:40 |
| *** verdurin has quit IRC | 14:42 | |
| *** agopi has quit IRC | 14:42 | |
| openstackgerrit | Aleksey Zvyagintsev proposed openstack-infra/jeepyb master: manage_projects: exit with error code https://review.openstack.org/626640 | 14:43 |
| openstackgerrit | Thierry Carrez proposed openstack-infra/puppet-ptgbot master: No longer build index page in puppet-ptgbot https://review.openstack.org/626911 | 14:44 |
| ttx | infra | 14:44 |
| ttx | oops | 14:44 |
| *** verdurin has joined #openstack-infra | 14:44 | |
| alexz__ | fungi: np! thanks for quick response ;) new ps uploaded | 14:44 |
| tosky | hopefully that was not a password | 14:44 |
| ttx | not really :) | 14:45 |
| fungi | tosky: if so, it would have been a pretty terrible password | 14:47 |
| tosky | :) | 14:48 |
| ttx | my password is 1nfra | 14:49 |
| fungi | smart! | 14:53 |
| *** fresta_ has joined #openstack-infra | 14:55 | |
| *** jonher_ has joined #openstack-infra | 14:55 | |
| ttx | fungi: well my provider requires a number, and I use the same password everywhere, so... | 14:56 |
| *** e0ne has quit IRC | 14:57 | |
| fungi | infra-root: heads up, i have approved my stack of changes (626391, 1626392, 626633, 626393) to generalize our global site manifest regexes and hostgroup glob patterns enabling opendev.org replacements for openstack.org servers. i'll keep close tabs on everything today in case this brings unanticipated regressions anywhere | 14:59 |
| *** fresta has quit IRC | 14:59 | |
| *** jonher has quit IRC | 14:59 | |
| *** jonher_ is now known as jonher | 14:59 | |
| fungi | #status log approved changes 626391, 1626392, 626633, 626393 which expand puppet node definitions and ansible hostgroup patterns to also match opendev.org hostnames | 15:00 |
| openstackstatus | fungi: finished logging | 15:01 |
| *** chandankumar is now known as chkumar|out | 15:02 | |
| *** dpawlik has quit IRC | 15:06 | |
| *** jpena|lunch is now known as jpena | 15:06 | |
| *** ykarel is now known as ykarel|afk | 15:08 | |
| openstackgerrit | Tom Barron proposed openstack-infra/devstack-gate master: Capture nfs-ganesha logs and conf https://review.openstack.org/626921 | 15:14 |
| *** ykarel|afk has quit IRC | 15:14 | |
| *** e0ne has joined #openstack-infra | 15:16 | |
| openstackgerrit | Merged openstack-infra/system-config master: Remove deleted review and puppetmaster servers https://review.openstack.org/626391 | 15:18 |
| openstackgerrit | Merged openstack-infra/system-config master: Use regexes for servers in global site manifest https://review.openstack.org/626392 | 15:20 |
| *** shardy has quit IRC | 15:25 | |
| *** shardy has joined #openstack-infra | 15:26 | |
| bnemec | Quick question: Does Gerrit stop sending emails if one bounces? | 15:32 |
| *** hamerins has joined #openstack-infra | 15:32 | |
| bnemec | I migrated my email server last night and it was briefly misconfigured and bounced a gerrit email. | 15:32 |
| bnemec | I haven't gotten any since then. | 15:32 |
| *** hamerins has left #openstack-infra | 15:33 | |
| bnemec | I'm correctly receiving openstack-discuss emails to the same address so it should be working now. | 15:33 |
| *** jcoufal has quit IRC | 15:35 | |
| *** ykarel|afk has joined #openstack-infra | 15:40 | |
| fungi | bnemec: i can check the mta logs. if it defers one message delivery it may defer subsequent messages until the retry time for the first one is reached | 15:41 |
| fungi | but those are usually fairly short retry times | 15:42 |
| fungi | so would expect it to have flushed them out to your server by now | 15:42 |
| *** quiquell is now known as quiquell|off | 15:43 | |
| bnemec | Yeah, it's been since last night. | 15:43 |
| *** ykarel|afk is now known as ykarel | 15:43 | |
| bnemec | 2018-12-20 18:38:24 H=(review01.openstack.org) [#.#.#.#] F=<review@openstack.org> rejected RCPT <openstack@nemebean.com>: relay not permitted | 15:43 |
| * bnemec failed to configure the server as the final destination for his domain, so it was trying to relay everything :-( | 15:43 | |
| bnemec | I suppose it's possible that my reviews have just been that inactive since then. | 15:44 |
| fungi | bnemec: oh, so a permanent rejection not a deferral | 15:44 |
| *** ginux has joined #openstack-infra | 15:44 | |
| bnemec | Yeah | 15:44 |
| fungi | the only deferrals i see in the queue at the moment are for users @sina.com which must be having trouble of some sort | 15:44 |
| AJaeger | bnemec: which review should i comment on so that you get an email? | 15:45 |
| fungi | 2018-12-21 00:38:24 1ga87u-0005Bl-PB ** openstack@nemebean.com R=dnslookup T=remote_smtp H=nemebean.com [192.81.249.155]: SMTP error from remote mail server after RCPT TO:<openstack@nemebean.com>: 550 relay not permitted | 15:45 |
| fungi | that was the last delivery attempt gerrit has made to your address | 15:46 |
| fungi | (time in utc) | 15:46 |
| AJaeger | bnemec: found one - let's try... | 15:46 |
| *** ginopc has quit IRC | 15:46 | |
| *** ginux is now known as ginopc | 15:46 | |
| bnemec | fungi: Yeah, that matches up with the one I pasted. | 15:46 |
| bnemec | Looking at my dashboard, it's possible that's the only email gerrit tried to send since last night. | 15:46 |
| fungi | agreed, so i think there's just been nothing to send since then | 15:46 |
| AJaeger | bnemec: commented on https://review.openstack.org/#/c/597612/ | 15:46 |
| bnemec | So maybe it's just that everyone has already left for the holidays. :-) | 15:46 |
| fungi | 2018-12-21 15:46:44 1gaN0d-000SJS-Gy => openstack@nemebean.com R=dnslookup T=remote_smtp H=nemebean.com [192.81.249.155] C="250 OK id=1gaN0e-00071S-6q" | 15:47 |
| fungi | that looks like it went through? | 15:47 |
| fungi | (just now) | 15:47 |
| bnemec | Yep, got it fine. | 15:47 |
| bnemec | Thanks guys. | 15:47 |
| AJaeger | bnemec, fungi, we should go on vacation as well ;) | 15:47 |
| bnemec | Sorry for the false alarm. | 15:47 |
| openstackgerrit | Thierry Carrez proposed openstack-infra/puppet-ptgbot master: No longer build index page in puppet-ptgbot https://review.openstack.org/626911 | 15:47 |
| fungi | bnemec: no worries! i'll do a better job of breaking stuff, i swear | 15:48 |
| fungi | AJaeger: yeah, a good chunk of today is packing and pre-travel errands for me | 15:48 |
| * bnemec is out til the second after today. | 15:48 | |
| AJaeger | bnemec: enjoy, happy holidays! | 15:48 |
| bnemec | AJaeger: Thanks, you too! | 15:48 |
| fungi | i'll be around, but just to fix what breaks, not really getting anything new done probably | 15:48 |
| AJaeger | fungi: hope nothing breaks ;) - and happy holidays to you as well | 15:49 |
| corvus | what, vacation when we're only at 70% of capacity (rather than our normal 150%)? | 15:49 |
| fungi | thanks, and to you! | 15:49 |
| fungi | corvus: yeah, zuul is "taking it easy" and just a few hundred jobs an hour | 15:50 |
| fungi | er, at just a few hundred | 15:50 |
| AJaeger | corvus: now is is your chance to crab 80 % of capacity ;) | 15:51 |
| fungi | okay, more than a few hundred. approaching 800jph now | 15:52 |
| corvus | shouldn't have said anything | 15:52 |
| fungi | but yeah, there's some headroom | 15:52 |
| fungi | grr, the random run-base failures are still around today | 15:53 |
| fungi | xenial : ok=0 changed=0 unreachable=1 failed=0 | 15:55 |
| fungi | http://logs.openstack.org/33/626633/1/gate/system-config-run-base/6bee14d/job-output.txt.gz#_2018-12-21_15_15_42_416089 | 15:55 |
| *** armax has joined #openstack-infra | 15:56 | |
| corvus | fungi: http://logs.openstack.org/33/626633/1/gate/system-config-run-base/6bee14d/job-output.txt.gz#_2018-12-21_15_11_46_639634 | 15:56 |
| corvus | looks like the controller lost contact with the node (so an intra-cloud problem) | 15:57 |
| fungi | yeah, failed to ssh during fact collection | 15:57 |
| corvus | (since that was from the inner ansible, that was fake-bridge talking to another node) | 15:58 |
| fungi | that was in limestone-regionone, in case we see more like that | 15:58 |
| *** jcoufal has joined #openstack-infra | 15:59 | |
| *** _alastor_ has joined #openstack-infra | 16:00 | |
| fungi | corvus: if you have a very quick moment to check that the test data change in 626633 for our system-config hostgroups validator makes sense to you, then i'll feel less bad about having self-approved it (but if not, no worries) | 16:02 |
| *** adriancz has quit IRC | 16:03 | |
| fungi | we were apparently testing that review.openstack.org was in some specific hostgroups, which needed changing with its removal and i wanted to at least replace it with something else for similar or better coverage | 16:03 |
| corvus | fungi: makes sense to me | 16:04 |
| fungi | thanks! | 16:04 |
| *** rlandy|rover is now known as rlandy|rover|brb | 16:04 | |
| fungi | if these manage to run the gauntlet of random network issues, i ought to be able to try spinning up new storyboard xenial servers with opendev.org instance names serving openstack.org vhosts (or at least find out what breaks next for that) | 16:05 |
| *** bnemec is now known as bnemerryxmas | 16:08 | |
| *** stephenfin is now known as finucannot | 16:14 | |
| finucannot | fungi: Quick one. I note all python-openstackclient bugs have been moved to Storyboard, but there are still some open on Launchpad that I can't find on storyboard https://bugs.launchpad.net/python-openstackclient/+bug/1411190 | 16:14 |
| openstack | Launchpad bug 1411190 in python-openstackclient "live migration on any host" [Medium,In progress] - Assigned to Rui Chen (kiwik-chenrui) | 16:14 |
| finucannot | fungi: Is that an issue or is it intended? | 16:14 |
| fungi | finucannot: https://storyboard.openstack.org/#!/story/1411190 | 16:16 |
| fungi | looks the same to me? | 16:16 |
| *** e0ne has quit IRC | 16:16 | |
| finucannot | That is. I can't figure out search on Storyboard and manually finding through the list of pages didn't bring that up | 16:17 |
| fungi | finucannot: if it's not obvious, we preserve the imported bug number as the story number for convenience | 16:17 |
| finucannot | Ideal | 16:17 |
| finucannot | I wonder if we could post a message to the bugs in Launchpad and auto-close them? | 16:17 |
| finucannot | I assume that's been discussed? | 16:17 |
| *** e0ne has joined #openstack-infra | 16:17 | |
| *** pcaruana has quit IRC | 16:18 | |
| fungi | yeah, simply performing read-only actions to retrieve bug information from lp is time consuming and fragile enough. automated write actions to lp are much more likely to fail | 16:18 |
| fungi | and take far, far longer to complete due to locking on their backend | 16:18 |
| finucannot | fungi: Fair. That they have the same bug number is good info though. Thanks for that | 16:19 |
| fungi | not to mention, requires us to add authentication in the importer, and they very well may throttle or block an account if they see it commenting on thousands of bugs | 16:19 |
| *** smarcet has joined #openstack-infra | 16:19 | |
| *** aojea has quit IRC | 16:20 | |
| *** bobh has quit IRC | 16:24 | |
| *** tpsilva has joined #openstack-infra | 16:27 | |
| *** rlandy|rover|brb is now known as rlandy|rover | 16:30 | |
| *** bobh has joined #openstack-infra | 16:34 | |
| *** wolverineav has joined #openstack-infra | 16:39 | |
| *** gyee has joined #openstack-infra | 16:40 | |
| *** e0ne has quit IRC | 16:40 | |
| *** e0ne has joined #openstack-infra | 16:42 | |
| *** wolverineav has quit IRC | 16:43 | |
| *** e0ne has quit IRC | 16:44 | |
| *** smarcet has quit IRC | 16:45 | |
| openstackgerrit | Merged openstack-infra/system-config master: Rejigger hostgroup tests for server removals https://review.openstack.org/626633 | 16:46 |
| fungi | wham | 16:46 |
| openstackgerrit | Merged openstack-infra/system-config master: Use domain-generic regexes/globs for hostnames https://review.openstack.org/626393 | 16:48 |
| fungi | and bam | 16:48 |
| fungi | now to start tailing logs | 16:48 |
| fungi | looks like our run_all finished moments after 16:45, so the next should begin at the top of the hour, in ~9 minutes | 16:50 |
| fungi | started at 16:15 so looks like we're taking roughly 30 minutes to complete a pulse | 16:51 |
| fungi | yeah, looks like they're consistently running within a few minutes either side of 30, so happening every 30 or 45 (at least looking at the start/end timestamps for the last dozen or so rounds) | 16:53 |
| *** jamesmcarthur has joined #openstack-infra | 16:54 | |
| openstackgerrit | James E. Blair proposed openstack-infra/system-config master: Add gitea k8s resource definitions and playbook https://review.openstack.org/626759 | 16:54 |
| corvus | fungi, mordred: that revision installs gitea, creates a root user, creates a gerrit user, and adds an ssh key | 16:55 |
| *** pgaxatte has quit IRC | 16:56 | |
| fungi | i think i may be starting to grok some of that now | 16:57 |
| corvus | next i'm going to start on a new playbook that lists all the projects in gerrit and creates orgs and repos in gittea | 16:58 |
| corvus | gitea | 16:58 |
| corvus | it's really hard to stop typing that extra t :) | 16:58 |
| *** jpich has quit IRC | 16:59 | |
| fungi | is the reason for the HTTP_PORT=3000 so that it doesn't need to only listen on localhost to avoid conflicting with apache on the egress interface, or to avoid needing to grant the process permission to open a low-numbered port? | 16:59 |
| corvus | the second | 16:59 |
| fungi | k | 17:00 |
| corvus | it runs as git(1000) inside its container | 17:00 |
| fungi | i mean, there are a couple ways to allow non-root to bind a listening socket below 2^10, but this ought to be fine | 17:01 |
| fungi | run_all is going now, i'm watching the log | 17:01 |
| corvus | the openssh daemon (running in another container in the same pod) happily listens on 22 without conflict with its host | 17:01 |
| fungi | got it | 17:01 |
| corvus | fungi: oh, also, apache isn't running here -- there's an octavia load balancer on the public ip which is currently performing the 80->3000 map. | 17:02 |
| fungi | out of curiosity, does it do something special with ssh or is that just in order to avoid running the openssh package from the distro? | 17:02 |
| corvus | fungi: we probably will want to add an apache or nginx into the mix (likely as a k8s ingress controller) to do the ssl termination and any other funny business we want | 17:02 |
| corvus | fungi: nothing special -- just that since this is k8s, it needs to be in a container. we actually are running sshd from the distro | 17:03 |
| corvus | lemme dig up a link | 17:03 |
| fungi | there's no ssh to the server on which kubernetes is running then? i assume it has an init outside the containers | 17:04 |
| corvus | fungi: yes, the k8s server itself is a perfectly normal machine you can ssh to. mordred's k8s deployment work has it running bionic right now. | 17:05 |
| fungi | or it just wants the sshd for the host server to have some additional isolation? | 17:05 |
| fungi | and so wraps it in a container | 17:05 |
| corvus | fungi: but gitea needs a captive ssh to make "git clone ssh://" work, so we run one in a container for it | 17:05 |
| fungi | aha, that's what i was wondering. so it is doing something special to allow git+ssh | 17:06 |
| corvus | fungi: https://review.openstack.org/#/c/626626/3/docker/gitea/Dockerfile has the build steps for the openssh for gitea in lines 100+ | 17:06 |
| fungi | we haven't offered git+ssh on our existing git servers, so didn't know that was coming... or is that in service of allowing gerrit to push over ssh? | 17:06 |
| corvus | the second | 17:07 |
| *** ginopc has quit IRC | 17:07 | |
| corvus | fungi: cloning over ssh won't work unless we allow people to register accounts, which is not in the initial plan | 17:07 |
| fungi | makes sense. need some way to handle ssh keys otherwise | 17:08 |
| fungi | jinja-init is apparently also something i should make time to read up on | 17:09 |
| corvus | fungi: yes -- though it is a very simple script: https://github.com/ObjectifLibre/jinja-init/blob/master/run.py | 17:09 |
| fungi | i guess that's a means of expanding configuration templates at container start time? | 17:09 |
| corvus | fungi: exactly | 17:09 |
| fungi | neat | 17:09 |
| fungi | and it's in python, not go? interesting | 17:10 |
| fungi | i guess because jinja2 is python | 17:10 |
| corvus | ya | 17:10 |
| corvus | due to the sensitivity of the operation, i chose to checkout a specific git sha in the dockerfile | 17:11 |
| corvus | i mean, i trust the authors, but it seemed a sensible precaution. | 17:11 |
| corvus | (basically, the reason you need this is to deal with secret data, so it's an obvious target) | 17:12 |
| fungi | right, sane choice | 17:12 |
| fungi | where does it obtain the jeblair/gitea and jeblair/gitea-openssh images from? | 17:13 |
| fungi | dockerhub? | 17:14 |
| fungi | not seeing any url for those | 17:14 |
| corvus | fungi: yes | 17:14 |
| fungi | thanks | 17:14 |
| *** mriedem is now known as mriedem_afk | 17:14 | |
| corvus | those will be opendev/ as soon as that patch lands and we start building them | 17:14 |
| fungi | i figured | 17:15 |
| fungi | this change is still plenty poc i'm sure judging from the hard-coded ip address url | 17:15 |
| fungi | i was more wondering about the mechanism for obtaining declared images, but sounds like dockerhub is just assumed by kubernetes | 17:16 |
| corvus | fungi: there's an internal registry it tries first, and you can run your own and use it in addition/instead | 17:16 |
| corvus | so it's not a requirement but is a default | 17:17 |
| *** shardy has quit IRC | 17:17 | |
| fungi | run_all is still looking good. it's just now applying iptables rules | 17:18 |
| *** jamesmcarthur has quit IRC | 17:18 | |
| fungi | i'll check back behind those as soon as it completes since that's where we've gone sideways in the past | 17:18 |
| *** gfidente has quit IRC | 17:18 | |
| fungi | it's moved on to snmp now, so checking firewalls | 17:19 |
| fungi | ip{,6}tables -L looking good on the servers i've spot-checked so far | 17:21 |
| corvus | it's annoying that gerrit seems to be the only app that thinks returning invalid json is the right way to return json. | 17:21 |
| corvus | (i'm looking at those 4 characters at the start of every response) | 17:21 |
| fungi | safety! if it returned real json, someone might <vague handwavery> and compromise a browser/webapp through cross-site scripting | 17:21 |
| fungi | yeah, i think it's insane | 17:22 |
| corvus | it reminds me of when ncsu decided to put a disclaimer on every web page it served, and so had apache stick in a bunch of close tags in the middle of the page to make sure it was displayed. made everybody's web pages invalid html. | 17:22 |
| fungi | gah | 17:23 |
| fungi | hey, it worked in internet explorer, what more do you want? | 17:23 |
| mordred | corvus, fungi: correction from above - the k8s cluster is running on xenial, not bionic | 17:24 |
| mordred | corvus, fungi: k8s isn't publishing debs for kube for bionic yet - something about bionic running too new of a version of something | 17:25 |
| corvus | mordred: that made me giggle | 17:25 |
| *** spa-87 has quit IRC | 17:25 | |
| mordred | :) | 17:26 |
| fungi | bleeding edge software needs kubernetes. kubernetes needs ubuntu lts-before-last | 17:26 |
| fungi | it makes perfect sense | 17:26 |
| mordred | fungi: incidentally, I used some already existing ansible to create the k8s - which I have submitted a few PRs upstream to improve which have been merged already | 17:27 |
| fungi | okay, we should be getting very close to the end of the run_all pulse now | 17:28 |
| *** jamesmcarthur has joined #openstack-infra | 17:29 | |
| *** jpena is now known as jpena|off | 17:31 | |
| fungi | 34m15s, no obvious errors | 17:34 |
| *** jamesmcarthur has quit IRC | 17:34 | |
| *** rpittau has quit IRC | 17:35 | |
| *** derekh has quit IRC | 17:36 | |
| fungi | looks like the puppet playbook has been failing on some servers, though i don't think it's new since those changes went in. looking into why | 17:38 |
| fungi | looks like the storyboard ones are likely related to the recent mysql change: | 17:39 |
| fungi | Execution of '/usr/bin/mysql -NBe create database if not exists `storyboard` character set `utf8mb4` collate `utf8_general_ci`' returned 1: ERROR 1253 (42000) at line 1: COLLATION 'utf8_general_ci' is not valid for CHARACTER SET 'utf8mb4' | 17:39 |
| fungi | might be a mysql-version-specific issue | 17:40 |
| *** trown is now known as trown|lunch | 17:44 | |
| fungi | guess i'll hack on that some when i get back from errands | 17:45 |
| fungi | puppet-user[5139]: Could not find dependency File[/etc/bind/keys/opendev.org] for File[/etc/bind/keys/opendev.org/Kopendev.org.+008+03945.key] at /opt/system-config/production/modules/openstack_project/manifests/master_nameserver.pp:40 | 17:46 |
| fungi | why is adns1.OPENSTACK.org complaining about that? is it just transitional until we move domains off it? | 17:46 |
| fungi | it's been going on for a while, so also not related to the globbing/regex changes | 17:47 |
| corvus | hrm, may have messed something up with the grouping there in the original changes | 17:48 |
| fungi | well, nothing seems to have gone down as a result of this most recent set of changes, so i'm going to go run errands now and then look into continuing on storyboard database stuff once i'm back | 17:48 |
| fungi | shouldn't be more than a couple hours | 17:49 |
| *** ykarel is now known as ykarel|away | 17:58 | |
| *** boden has quit IRC | 18:07 | |
| openstackgerrit | wes hayutin proposed openstack-infra/elastic-recheck master: add query for intermittent container failure https://review.openstack.org/626962 | 18:10 |
| *** jamesmcarthur has joined #openstack-infra | 18:19 | |
| *** smarcet has joined #openstack-infra | 18:19 | |
| *** jamesmcarthur has quit IRC | 18:23 | |
| *** Dobroslaw has quit IRC | 18:26 | |
| clarkb | corvus: any reason to not approve the opendev git spec now? I see there are some new comments since our meeting. But no -1s and they seem minor | 18:26 |
| clarkb | we can clea nthings up in a followup? | 18:26 |
| corvus | clarkb: yeah, i don't see showstoppers. zane brought up a good additional work item for us to look into, but if we fail at that, i don't think it's a showstopper | 18:27 |
| corvus | mordred: ^ perhaps you'd like to vote on https://review.openstack.org/623033 | 18:28 |
| clarkb | my only other concern is there are realtively few rollcall votes. | 18:28 |
| clarkb | you read my mind :) | 18:28 |
| clarkb | I'll give mordred a chance to look at it so we can have a bit more consensus on it before approving | 18:29 |
| *** smarcet has quit IRC | 18:30 | |
| corvus | http://38.108.68.66/explore/organizations | 18:34 |
| corvus | orgs are created :) | 18:34 |
| openstackgerrit | Merged openstack-infra/infra-specs master: StoryBoard Story Attachments https://review.openstack.org/607377 | 18:35 |
| *** wolverineav has joined #openstack-infra | 18:38 | |
| *** boden has joined #openstack-infra | 18:38 | |
| AJaeger | clarkb, corvus : the opendev git spec looks fine. | 18:38 |
| AJaeger | clarkb: want to merge https://review.openstack.org/#/c/623211 as well? | 18:39 |
| clarkb | AJaeger: there was frickler's concern and fungi had concede possibly waiting until gerrit was restarted | 18:39 |
| clarkb | I odn't mind approving it, but didn't want to do that if they had mostly agreed on that | 18:39 |
| AJaeger | clarkb: ah... | 18:40 |
| AJaeger | clarkb: then let's restart while everybody is away ;) | 18:41 |
| AJaeger | clarkb: want to merge https://review.openstack.org/#/c/617399/ and https://review.openstack.org/#/c/617400 as well? Two nitpicks removing duplicate words... | 18:42 |
| *** jcoufal has quit IRC | 18:42 | |
| *** trown|lunch is now known as trown | 18:45 | |
| *** kaiokmo has quit IRC | 18:47 | |
| openstackgerrit | Merged openstack-infra/infra-specs master: remove the dumplicate words https://review.openstack.org/617399 | 18:53 |
| openstackgerrit | Merged openstack-infra/infra-specs master: remove the dumplicate word https://review.openstack.org/617400 | 18:53 |
| clarkb | alright I'm off to do touristy things but will check in later to see if mordred has weighed in on the opendev spec | 18:53 |
| mordred | clarkb: voted | 18:57 |
| pabelanger | me too! | 18:58 |
| *** evgenyl has joined #openstack-infra | 18:59 | |
| *** auristor has quit IRC | 19:01 | |
| *** rossella_s has quit IRC | 19:05 | |
| *** auristor has joined #openstack-infra | 19:06 | |
| openstackgerrit | Monty Taylor proposed openstack-infra/system-config master: Add resources for deploying rook and xtradb to kuberenets https://review.openstack.org/626054 | 19:07 |
| openstackgerrit | Monty Taylor proposed openstack-infra/system-config master: Run k8s-on-openstack to manage k8s control plane https://review.openstack.org/626965 | 19:07 |
| mordred | corvus: ^^ reworked that a bit | 19:07 |
| mordred | corvus: it now has the k8s deploy integrated in with run_All and separate from the other things, and has playbooks for rook and pxc | 19:08 |
| evgenyl | Hi everybody, is it a known issue (?) that readthedocs triggers fail http://zuul.openstack.org/builds?job_name=trigger-readthedocs-webhook ? | 19:08 |
| *** ykarel|away has quit IRC | 19:09 | |
| AJaeger | evgenyl: yes ;/ | 19:10 |
| AJaeger | evgenyl: details at http://lists.openstack.org/pipermail/openstack-infra/2018-December/006247.html | 19:10 |
| evgenyl | AJaeger: I thought that there was a problem with configuration on our side. Thank you! | 19:11 |
| *** elod has quit IRC | 19:15 | |
| *** elod has joined #openstack-infra | 19:16 | |
| *** tristanC has quit IRC | 19:17 | |
| *** tristanC has joined #openstack-infra | 19:17 | |
| *** e0ne has joined #openstack-infra | 19:20 | |
| *** spa-87 has joined #openstack-infra | 19:23 | |
| *** wolverineav has quit IRC | 19:28 | |
| *** wolverineav has joined #openstack-infra | 19:29 | |
| *** wolverineav has quit IRC | 19:30 | |
| *** wolverineav has joined #openstack-infra | 19:31 | |
| *** wolverineav has quit IRC | 19:32 | |
| *** wolverineav has joined #openstack-infra | 19:32 | |
| *** wolverineav has quit IRC | 19:37 | |
| *** mriedem has joined #openstack-infra | 19:38 | |
| *** mriedem_afk has quit IRC | 19:41 | |
| openstackgerrit | James E. Blair proposed openstack-infra/system-config master: Add gitea replication https://review.openstack.org/626968 | 19:45 |
| openstackgerrit | James E. Blair proposed openstack-infra/system-config master: Add gitea replication https://review.openstack.org/626968 | 19:46 |
| corvus | fungi, mordred: ^ i think i'll be ready for that in a few hours -- but i don't think it will hurt to go ahead and merge it even before gitea is ready for it -- gerrit handles replication failures gracefully | 19:46 |
| corvus | and i'd like to see how the system handles the full volume of data, and what that replication looks like | 19:47 |
| * fungi is back and catching up | 20:01 | |
| fungi | in a sec i'll check what's going on and maybe just restart gerrit | 20:02 |
| *** bobh has quit IRC | 20:02 | |
| *** bobh has joined #openstack-infra | 20:05 | |
| *** bobh has quit IRC | 20:06 | |
| *** edmondsw has quit IRC | 20:10 | |
| *** spa-87 has quit IRC | 20:21 | |
| *** dave-mccowan has quit IRC | 20:24 | |
| *** dave-mccowan has joined #openstack-infra | 20:32 | |
| *** wolverineav has joined #openstack-infra | 20:34 | |
| fungi | meh, there's still some stuff running, i'll give it time | 20:34 |
| corvus | fungi: https://review.openstack.org/626968 would be nice to get in before restart if you think it's ok | 20:34 |
| fungi | yep, i had just pulled it up to review. good point about needing a gerrit restart for that! | 20:35 |
| fungi | lgtm, if mordred's not really around i'm happy to single-core approve | 20:35 |
| corvus | fungi: let's go ahead and do it; it'll take enough time to get on disk to give him an opportunity to object or +2 before he gets back i guess :) | 20:39 |
| fungi | yeah, coming to the same conclusion. it's quiet in here today | 20:40 |
| fungi | corvus: approved it. as for things pending a gerrit restart, what do you think about approving 471077 and 471078 | 20:42 |
| fungi | though i guess jhesketh had a comment on 471077 i haven't addressed | 20:42 |
| corvus | fungi: both lgtm | 20:44 |
| * mordred waves | 20:44 | |
| mordred | I added a late +2 onto the repl patch | 20:44 |
| fungi | it's not late as it hasn't merged yet ;) | 20:44 |
| corvus | fungi: i +2 but did not +W pending what you decide about that comment, but i see no reason you shouldn't +W when ready | 20:45 |
| fungi | yeah, i think it's safe to let footers override footer there | 20:45 |
| *** bobh has joined #openstack-infra | 20:46 | |
| fungi | we can clean up the old singular pattern in the future if we decide it's safe to do so... we've not been really careful about backwards-compatibility in these | 20:46 |
| mordred | I'm not super concerned about backwards compat in these at this point | 20:46 |
| mordred | fungi: +2 on both from me - corvus comment applies from me too - fire at will | 20:47 |
| fungi | we restart gerrit so infrequently now... it's eerie to be doing it for config changes | 20:48 |
| mordred | I agree | 20:48 |
| fungi | >4.5 months since the last gerrit restart | 20:49 |
| fungi | to think we were restarting weekly there for a while at the height of the jvm gc issues | 20:50 |
| mordred | fungi: I'd rather not think about that :) | 20:50 |
| fungi | good point | 20:50 |
| *** bobh has quit IRC | 20:50 | |
| fungi | hah, those two changes are so old i needed to recheck to get a vote from zuul because their verify +1 votes were still from "jenkins" | 20:52 |
| mordred | hahaha | 20:52 |
| corvus | ghost of jenkins past again | 20:52 |
| fungi | probably just a bit of undigested figgy pudding | 20:53 |
| fungi | i have items which aren't mine but require me to wrap them in festive paper covering, so i'll do that for a bit while i wait for things to merge | 20:54 |
| *** spa-87 has joined #openstack-infra | 20:56 | |
| *** wolverineav has quit IRC | 20:59 | |
| *** wolverineav has joined #openstack-infra | 21:00 | |
| openstackgerrit | Merged openstack-infra/system-config master: Add gitea replication https://review.openstack.org/626968 | 21:04 |
| fungi | there's one of the three | 21:04 |
| *** wolverineav has quit IRC | 21:05 | |
| *** rlandy|rover has quit IRC | 21:05 | |
| *** wolverineav has joined #openstack-infra | 21:06 | |
| *** trown is now known as trown|outtypewww | 21:07 | |
| *** wolverineav has quit IRC | 21:21 | |
| *** spa-87 has quit IRC | 21:25 | |
| *** wolverineav has joined #openstack-infra | 21:26 | |
| *** wolverineav has quit IRC | 21:26 | |
| *** wolverineav has joined #openstack-infra | 21:26 | |
| *** efried has quit IRC | 21:27 | |
| *** kencjohnston has quit IRC | 21:29 | |
| *** wolverineav has quit IRC | 21:29 | |
| fungi | this cthulhus and elder signs holiday wrap i got on a kickstarter some 8 years ago has lasted unfathomably longer than i predicted. i still have two unopened rolls of the stuff | 21:31 |
| mordred | cthulu is unfathomable | 21:33 |
| openstackgerrit | Merged openstack-infra/puppet-gerrit master: Allow lists of trackingid footers https://review.openstack.org/471077 | 21:34 |
| corvus | oh, wow. i see what the 2 of you did there. | 21:34 |
| mordred | corvus: see us next week at the copa and the week after that in boca raton | 21:34 |
| *** wolverineav has joined #openstack-infra | 21:35 | |
| * mordred goes back to learning how to bribe foreign officals | 21:35 | |
| fungi | mordred: small, non-consecutively-numbered notes | 21:35 |
| corvus | and get a receipt! | 21:36 |
| openstackgerrit | Merged openstack-infra/system-config master: Add a trackingid for Launchpad bug headers https://review.openstack.org/471078 | 21:37 |
| fungi | you still have a week to rack up itemized deductions for 2018, after all | 21:37 |
| fungi | and there's the last of the three gerrit config changes we were waiting for | 21:37 |
| mordred | incidentally, apparently the answer to all questions regarding gdpr behavior is "ask legal" | 21:37 |
| fungi | and the answer from "legal" is always "no" regardless of the question | 21:38 |
| fungi | kafka would be proud | 21:38 |
| fungi | looks like there's a config pulse underway since 21:30 so these will get supplied in the next pulse starting probably 22:15 (maybe 22:00 if the clouds are on our side) | 21:39 |
| *** wolverineav has quit IRC | 21:39 | |
| * fungi suffocates more material goods in colorful paper while waiting | 21:40 | |
| mordred | fungi: the best part about suffocating them is that it opens the door to acquiring additional material goods in the future | 21:43 |
| fungi | it's the foundation of our nation | 21:47 |
| fungi | consumerism to all, and to all a good night | 21:47 |
| mordred | fungi: wow, it's like the good night is a gift-with-purchase | 21:48 |
| fungi | but wait, there's more | 21:48 |
| corvus | fungi, mordred: huh. in https://try.gitea.io/api/swagger i can't find either the ability to edit a repository, or the ability to alter settings like whether the issue tracker is enabled | 21:48 |
| *** jamesmcarthur has joined #openstack-infra | 21:48 | |
| fungi | we, now, don't delay, order now | 21:49 |
| corvus | we might have to write some go | 21:49 |
| *** boden has quit IRC | 21:49 | |
| fungi | as in, no api methods in gitea for modifying repository configuration? | 21:50 |
| corvus | right | 21:50 |
| corvus | and unlike gerrit, the webui does not appear to use the api | 21:51 |
| mordred | corvus: um - http://38.108.68.66/api/swagger is weird to me | 21:51 |
| corvus | yeah, the ip addr is wrong | 21:51 |
| mordred | corvus: it's trying to fetch http://38.108.68.66/api/swagger | 21:51 |
| mordred | kk. just as long as you're seeing it too | 21:51 |
| corvus | mordred: if you correct the address in the box at the top it'll work | 21:51 |
| corvus | i looked there, but i also used the try url so that i could see if git master has the api method | 21:51 |
| corvus | (since we're running a released version) | 21:52 |
| fungi | try.g.i is continuously deployed from their master branch? | 21:52 |
| corvus | or at least frequently, so i gather | 21:52 |
| fungi | k | 21:52 |
| *** jamesmcarthur has quit IRC | 21:53 | |
| fungi | hopefully that's not because altering the repository configuration requires a restart of the service | 21:53 |
| mordred | corvus: I agree with you - I don't see such an api call | 21:53 |
| corvus | fungi: try is currently running the commit on master | 21:53 |
| corvus | fungi: the option is in the web ui; i think they just haven't plumbed it through | 21:54 |
| corvus | we could probably do a ridiculous thing like perform a mysql query in the interim. | 21:54 |
| fungi | the built-in release management options also seem weird. i guess "releases" aren't directly tied to git tags? | 21:55 |
| corvus | unsure -- i do know tags create releases (and mordred has suggested we hide the releases tab, or at least hide the ability to grab a tarball from a release). i haven't looked into that yet | 21:56 |
| fungi | presumably without arbitrary authenticated users we just wouldn't create "releases" without an accompanying tag | 21:56 |
| fungi | either that or methods like post /repos/{owner}/{repo}/releases are just a no-op | 21:57 |
| *** efried has joined #openstack-infra | 21:57 | |
| fungi | looks like you feed it a "tag_name" when creating the release | 21:57 |
| corvus | weird. i wonder what actually happens then. | 21:58 |
| fungi | could be for post-hoc crration of releases for existing tags in imported repositories or something | 21:59 |
| mordred | I think you could potentially use that API to attach notes or artifacts to a release perhaps | 21:59 |
| mordred | like, I tried querying the releases endpoint for the zuul repo we'd imported that showed releases for each of its tags and it returned nothing | 21:59 |
| mordred | corvus: the settings page does a POST to /org/repo/settings - and it's a fairly simple set of form data | 22:00 |
| fungi | run_all pulse from 21:30:00 didn't complete until 22:00:25 so we're looking at the 22:15 pulse after all for the gerrit config updates | 22:01 |
| corvus | mordred: yeah, but that almost certainly means implementing cookie auth | 22:01 |
| mordred | corvus: yeah. http://paste.openstack.org/show/737938/ is a capture, fwiw | 22:02 |
| corvus | yeah, i opened it up in my debugger earlier -- that's how i determined it's not using the api | 22:02 |
| mordred | corvus: we'd probably get further by just writing some go to expose it as an api | 22:02 |
| corvus | mordred: yes, that is the right thing to do. | 22:03 |
| mordred | corvus: do you want me to work on an interim mysql query for now? | 22:06 |
| corvus | mordred: i was just starting on that... i just discovered that this is stored in the repo_unit table | 22:06 |
| mordred | oh god | 22:07 |
| corvus | http://paste.openstack.org/show/737939/ | 22:07 |
| corvus | it's not as lovely as one might hope | 22:07 |
| mordred | what has happened to my life | 22:07 |
| mordred | I can't unsee this | 22:07 |
| mordred | corvus: can you configure a repo by hand with the settings we'd want? (or alternately, point me to how to log in to the web ui) | 22:08 |
| fungi | eek | 22:08 |
| corvus | mordred: the 'type' colum corresponds to things like code, releases, wiki, etc | 22:09 |
| corvus | yeah, i'll set up the 'test' repo now | 22:09 |
| fungi | that's a wonderfully opaque schema | 22:09 |
| mordred | yay for schemaless designs on a rdbms! | 22:09 |
| fungi | someone really wanted redis/mongo | 22:09 |
| mordred | (I mean, it'll make the query to set it pretty easy to do from ansible - since we just need to to_json a dict and then shove that into a single upsert | 22:10 |
| fungi | but still, gross | 22:10 |
| corvus | mordred: repo_id=2 should be set now | 22:10 |
| * fungi shakes his head at the "new generation" of developers | 22:11 | |
| corvus | http://paste.openstack.org/show/737940/ | 22:11 |
| mordred | corvus: pull requests are disabled? | 22:11 |
| corvus | this is the first part of the schema i've run into which is weird | 22:11 |
| corvus | mordred: hrm. unclear. | 22:11 |
| corvus | i'm digging into that | 22:12 |
| *** efried has quit IRC | 22:12 | |
| mordred | corvus: I'm poking in the db too - but you disabled them in teh ui? | 22:12 |
| corvus | mordred: i did not see an option for that | 22:13 |
| corvus | i'm working on understanding why | 22:13 |
| fungi | if this is the first strange not-really-rdbms table you've encountered, makes me wonder whether gitea is actually more than one application/service which got mashed together at some point | 22:14 |
| *** slaweq has quit IRC | 22:16 | |
| mordred | corvus: https://github.com/go-gitea/gitea/blob/eabbddcd98717ef20d8475e819f403c50f4a9787/templates/repo/settings/options.tmpl#L200 is interesting ... ah - | 22:16 |
| mordred | corvus: https://github.com/go-gitea/gitea/blob/1b3404eacaa679c2c06d6a3b63c83a4bec5649d5/models/repo.go#L659 | 22:17 |
| corvus | mordred: yeah, i'm working on getting content in there; i have to fix perms | 22:17 |
| mordred | neat. | 22:17 |
| fungi | new run_all pulse is underway as of a couple minutes ago. should be able to restart gerrit around 22:45 if the config looks right | 22:17 |
| fungi | zuul jph is down around 100 now | 22:20 |
| fungi | or at least was as of 20 minutes ago but may be trending back up again | 22:20 |
| corvus | mordred: okay, pull requests are unit type 3 | 22:23 |
| corvus | mordred: so to disable pull requests, delete from repo_unit where type=3 | 22:23 |
| corvus | mordred: (i just toggled it on/off for openstack/test, repo id 2) | 22:23 |
| mordred | corvus: fascinating | 22:24 |
| mordred | corvus: and I guess wiki is one of the other types? | 22:24 |
| corvus | mordred: looks like external wiki is type 6, internal wiki is type 5 | 22:26 |
| corvus | so i think we also want to delete type=5 | 22:26 |
| corvus | issue tracker is type=2 | 22:28 |
| corvus | rather, internal issue tracker is type=2 | 22:28 |
| *** kgiusti has left #openstack-infra | 22:28 | |
| corvus | appears external issue tracker is type=7 (which we can use to link to storyboard/lp) | 22:28 |
| mordred | cool. so - for a first-stab, we can just sledgehamer - 'delete from repo_unit where type in (2, 3, 5)' | 22:29 |
| mordred | which will disable all the things | 22:29 |
| corvus | yep | 22:29 |
| mordred | as a second step, we can add type 7 entries - although we could just leave that as a "get the API done" step | 22:29 |
| corvus | i think i'd like to add type 7 | 22:29 |
| corvus | so we can see the links | 22:30 |
| mordred | kk | 22:30 |
| corvus | {"ExternalTrackerURL":"http://storyboard.openstack.org/foo","ExternalTrackerFormat":"http://storyboard.openstack.org/foo/{index}","ExternalTrackerStyle":"numeric"} | 22:30 |
| mordred | yeah - it's not a hard insert | 22:30 |
| mordred | and mysql lets you insert with a join | 22:30 |
| corvus | looks like we got the project-by-name thing running in storyboard just in time :) | 22:30 |
| mordred | it's really a shame that they don't have a unique constraint on repo_id, type | 22:30 |
| mordred | corvus: ++ | 22:31 |
| mordred | corvus: for launchpad projects, it's just bugs.launchpad.net/{short_project} right? | 22:31 |
| corvus | mordred: yeah, though since i'm not parsing projects.yaml i don't have a way to switch between them atm, so let's just do storyboard for now | 22:32 |
| mordred | corvus: oh - are you just grabbing a flat list from gerrit? | 22:33 |
| corvus | mordred: yep | 22:33 |
| mordred | nod. | 22:33 |
| * mordred is pondering best way to do this since we don't have a unique key to use with on duplicate key insert | 22:33 | |
| *** e0ne has quit IRC | 22:35 | |
| corvus | mordred: fwiw, i have the repo id available, you don't have to join | 22:35 |
| mordred | well, I was actually thinking we might could do it with a single query | 22:35 |
| mordred | rather than one per project | 22:35 |
| corvus | (i just now realized the significance of your observations about joins earlier) | 22:35 |
| corvus | mordred: i think it needs to be per-project anyway | 22:36 |
| mordred | why's that? | 22:36 |
| corvus | we're going to need to have ansible substitute in the external tracker url for each project | 22:36 |
| corvus | gitea only does templating for the individual issue links | 22:36 |
| corvus | {"ExternalTrackerURL":"https://storyboard.openstack.org/#!/project/USER/REPO","ExternalTrackerFormat":"https://storyboard.openstack.org/#!/story/{index}","ExternalTrackerStyle":""} | 22:36 |
| corvus | mordred: ^ that's the actual value we want | 22:36 |
| corvus | mordred: and USER and REPO need to be hard-coded values | 22:37 |
| mordred | yeah. I'm pretty sure we can do that with a single non-templated query | 22:37 |
| corvus | i'd rather have this be per-project | 22:38 |
| corvus | i don't want to adjust other projects not in the system | 22:38 |
| mordred | ok. | 22:39 |
| *** slaweq has joined #openstack-infra | 22:43 | |
| mordred | corvus: I think this should do the trick then: http://paste.openstack.org/show/737941/ | 22:44 |
| mordred | corvus: since there is no unique key, on duplicate key insert isn't workable, wrapping in an explicit transaction and doing a delete/insert is probably the safest | 22:45 |
| mordred | although if we're going to run this on the whole list every time, that'll be an unfortunate amount of senseless write traffic | 22:46 |
| corvus | mordred: cool, i should be ready to test that in just a few mins. | 22:46 |
| corvus | mordred: for now i'll just have it do it on repo creation | 22:46 |
| mordred | ok cool | 22:47 |
| mordred | in that case it shouldn't be terrible | 22:47 |
| *** slaweq has quit IRC | 22:47 | |
| fungi | last run_all pulse ended at 22:45:55 so checking gerrit config on review.o.o now | 22:49 |
| *** mriedem is now known as mriedem_away | 22:49 | |
| *** smcginnis is now known as smcginnis_away | 22:49 | |
| *** lbragstad has quit IRC | 22:50 | |
| fungi | the its-storyboard commentlink has been updated to key off task headers, the launchpad-bug tracking-id has been added and the gitea replication remote is in place | 22:51 |
| fungi | everything is looking sane gerrit config wise | 22:51 |
| *** e0ne has joined #openstack-infra | 22:51 | |
| fungi | nothing in the gate pipeline looks close to reporting | 22:52 |
| fungi | infra-root: any objections to me performing a gerrit restart in the next few minutes? | 22:53 |
| corvus | fungi: wfm | 22:53 |
| *** bobh has joined #openstack-infra | 22:53 | |
| mordred | fungi: go for it | 22:54 |
| fungi | the gerrit config in the gate just got an updated eta of 8 minutes, but i should be able to beat the clock | 22:54 |
| fungi | er, the nova change in the gate | 22:54 |
| fungi | #status log the gerrit service on review.openstack.org is being restarted to pick up new configuration changes, and will return momentarily | 22:54 |
| openstackstatus | fungi: finished logging | 22:54 |
| fungi | er, that should have been a notice | 22:54 |
| fungi | #status notice the gerrit service on review.openstack.org is being restarted to pick up new configuration changes, and will return momentarily | 22:55 |
| openstackstatus | fungi: sending notice | 22:55 |
| fungi | stopping now | 22:55 |
| fungi | and starting | 22:56 |
| -openstackstatus- NOTICE: the gerrit service on review.openstack.org is being restarted to pick up new configuration changes, and will return momentarily | 22:56 | |
| fungi | Job for gerrit.service failed because the control process exited with error code. See "systemctl status gerrit.service" and "journalctl -xe" for details. | 22:56 |
| mordred | hrm | 22:57 |
| corvus | looking | 22:57 |
| corvus | Dec 21 22:56:11 review01 gerrit[59289]: fatal: bad config file line 112 in /home/gerrit2/review_site/etc/gerrit.config | 22:57 |
| corvus | [trackingid "launchpad-bug"] | 22:57 |
| corvus | match = "\#?(\d+)" | 22:57 |
| corvus | needs more '\' ? | 22:57 |
| mordred | yeah- I think it needs 2 more | 22:57 |
| *** bobh has quit IRC | 22:57 | |
| corvus | fungi, mordred: i'm happy to make that change. ok? | 22:58 |
| *** hwoarang has quit IRC | 22:58 | |
| mordred | corvus: ++ | 22:58 |
| fungi | go for it | 22:58 |
| *** hwoarang has joined #openstack-infra | 22:58 | |
| mordred | the storyboard story line below looks correct | 22:58 |
| fungi | i agree it's missing \\ love | 22:58 |
| openstackstatus | fungi: finished sending notice | 22:58 |
| fungi | corvus: make change to the on-disk config or writing up the patch? i'm happy to do whichever one you aren't | 22:59 |
| corvus | fungi: on disk | 22:59 |
| fungi | cool, you going to try to start it or shall i? | 22:59 |
| corvus | i've started it | 22:59 |
| fungi | thanks! | 23:00 |
| corvus | web ui is back up | 23:00 |
| fungi | working on the patch now | 23:00 |
| corvus | and ssh is answering now too | 23:00 |
| corvus | so i think it's "back up" | 23:00 |
| corvus | 19207 tasks in queue | 23:00 |
| mordred | \o/ | 23:01 |
| corvus | on gitea, i'm just deleting the akanda repo over and over | 23:02 |
| openstackgerrit | Jeremy Stanley proposed openstack-infra/system-config master: Add more \\ to launchpad-bug Gerrit tracking-id https://review.openstack.org/627018 | 23:03 |
| fungi | brown bag ^ | 23:03 |
| mordred | corvus: on purpose, or as a test? | 23:03 |
| corvus | mordred: testing | 23:03 |
| mordred | cool | 23:04 |
| corvus | i'm testing repo creation | 23:04 |
| corvus | mordred: what's the mysql arg to run a query on the cmdline? | 23:08 |
| *** rascasoft has quit IRC | 23:09 | |
| corvus | -e i think | 23:09 |
| mordred | corvus: I usually pipe it in ... | 23:10 |
| corvus | mordred: http://paste.openstack.org/show/737942/ | 23:10 |
| mordred | echo "select * from repo_unit where repo_id=2" | ./kubectl -n gitea-db -c database exec -i gitea-pxc-1 | 23:10 |
| mordred | mysql gitea | 23:10 |
| mordred | corvus: hrm - the -e worked - it didn't like begin transaction ... checking | 23:12 |
| corvus | meanwhile i'll sprinkle more \ around that | 23:13 |
| corvus | (but i don't think the \ were the issue; i think they'd be the next issue) | 23:13 |
| mordred | corvus: DUH. | 23:13 |
| mordred | corvus: start transaction, not begin transaction | 23:13 |
| corvus | k | 23:13 |
| mordred | corvus: begin transaction is, I think, sybase | 23:13 |
| mordred | corvus: there's an issue in yuour insert too | 23:15 |
| mordred | corvus: insert into repo_unit (repo_id, `type`, config) ... the repo_id got replaced by the numeric repo id there - but at that point it wants to be the bare word repo_id (column name) | 23:15 |
| corvus | doh, thx | 23:16 |
| mordred | it's correct in the values section :) | 23:16 |
| mordred | and ExternalTrackerStyle seems cut off | 23:16 |
| mordred | but maybe that's just ansible truncating the command? | 23:16 |
| mordred | oh - nevermind | 23:17 |
| mordred | corvus: I'm guessing that woudl be the more \s you were talking about | 23:17 |
| corvus | mordred: hrm, i still can't get the quotes right apparently: http://paste.openstack.org/show/737943/ | 23:21 |
| corvus | oh, sorry about the continuation lines in the first part -- that's from emacs. the command is really all on one line | 23:21 |
| *** e0ne has quit IRC | 23:21 | |
| corvus | corrected -- http://paste.openstack.org/show/737944/ | 23:22 |
| *** eernst has joined #openstack-infra | 23:22 | |
| mordred | corvus: I see it | 23:24 |
| mordred | corvus: it's the 'id' in create_repo_result.json['id'] | 23:24 |
| mordred | corvus: maybe do a set_fact on that so that you don't have to subscript there? | 23:24 |
| corvus | k | 23:24 |
| mordred | also - it seems like there is a missing semicolon at the end of the insert before the commit | 23:25 |
| corvus | also i think we were missing ); at the end of the insert | 23:25 |
| mordred | corvus: oh - you know what ... | 23:25 |
| mordred | corvus: you could set_fact the actual sqlquery too - so you could do it in a multi-line yaml block | 23:25 |
| corvus | good idea | 23:26 |
| mordred | corvus: then just do -e '{{ sql_statement }}' | 23:26 |
| *** jamesmcarthur has joined #openstack-infra | 23:26 | |
| *** dave-mccowan has quit IRC | 23:26 | |
| corvus | cool, i'm down to trivial mysql errors now | 23:29 |
| corvus | type_id -> `type` | 23:29 |
| corvus | mordred: http://paste.openstack.org/show/737945/ | 23:30 |
| clarkb | fungi good to approve the storyboard infra specs docs update post ferrit restart? | 23:30 |
| mordred | yeah - I wasn't sure if we needed to sql quote type or not - couldn't remember | 23:31 |
| mordred | clarkb: mmm. ferrit | 23:31 |
| corvus | mordred: i'm not sure that made the quoting situation better :/ | 23:31 |
| mordred | corvus: it looks like we lost the ' in the query - can I see the actual ansible? | 23:32 |
| mordred | (I thinkm we're really close here) | 23:32 |
| corvus | mordred: yeah, let me move it into a new file so we can iterate faster | 23:32 |
| mordred | corvus: cool. and I can obvs look at it on bridge | 23:33 |
| corvus | (i just added more escapes and am right back to the no closing quote | 23:33 |
| *** dims has quit IRC | 23:33 | |
| corvus | mordred: ansible-playbook -i ~corvus/opendev/gitea/inventory.yaml ~corvus/opendev/gitea/test-playbook.yaml | 23:34 |
| mordred | oh - yeah - because the string interp is still gonna make it wrapped | 23:34 |
| mordred | corvus: ok - lemme try a different quoting approach | 23:36 |
| *** e0ne has joined #openstack-infra | 23:37 | |
| mordred | corvus: boom - I think that worked? | 23:37 |
| *** bobh has joined #openstack-infra | 23:38 | |
| *** e0ne has quit IRC | 23:38 | |
| mordred | hrm. well, it didn't error - but I don't think it did anything either | 23:38 |
| mordred | oh - it totally did - it just didn't do a repo_id | 23:40 |
| corvus | mordred: no? there's some stuff in the db | 23:40 |
| corvus | yeah | 23:40 |
| mordred | yeah - turns out, *I* was in a started transaction in my mysql window | 23:40 |
| mordred | yay for transaction isolation :) | 23:40 |
| corvus | i must have uninterpolated that repo id in values | 23:41 |
| corvus | easy fix | 23:41 |
| corvus | mordred: i'll fix it | 23:41 |
| corvus | mordred: also, is now() going to work for created_unix ? | 23:42 |
| mordred | maybe? which table? | 23:42 |
| *** bobh has quit IRC | 23:42 | |
| corvus | repo_unit | 23:42 |
| corvus | nope, it put in 20181221234259 not 1545435142 | 23:43 |
| mordred | yeah - I think you want ... | 23:43 |
| *** jamesmcarthur has quit IRC | 23:43 | |
| mordred | unix_timestamp() | 23:44 |
| *** slaweq has joined #openstack-infra | 23:44 | |
| corvus | looks good, thanks! | 23:44 |
| mordred | \o/ | 23:44 |
| corvus | i'll plop that back into the main playbook, and run it again | 23:44 |
| mordred | not QUITE as easy as doing a REST call, but not the WORST thing we've ever written by a long stretch | 23:44 |
| corvus | heh. we set the bar low. | 23:45 |
| corvus | mordred: w00t! looks good! | 23:46 |
| corvus | i'm going to run it on all the repos now | 23:47 |
| *** jamesmcarthur has joined #openstack-infra | 23:47 | |
| corvus | er, actually i'm going to make sure it no-ops on an already created repo first | 23:48 |
| mordred | ++ | 23:48 |
| *** slaweq has quit IRC | 23:48 | |
| mordred | I'm excited | 23:48 |
| corvus | okay, running globally now | 23:50 |
| mordred | corvus: woot. and replication is enabled - so in theory a bunch of repos should get created, and then get populated with a crap-ton of data | 23:50 |
| corvus | mordred: yep | 23:50 |
| mordred | wcpgw? | 23:50 |
| clarkb | hopefully replocating off git far not gerrit | 23:52 |
| fungi | clarkb: yeah, it now reflects present reality rather than future reality | 23:52 |
| clarkb | or gerrit may have a sad | 23:52 |
| *** jamesmcarthur has quit IRC | 23:52 | |
| corvus | clarkb: gerrit is replicating *to* it. push. | 23:52 |
| fungi | when push comes to shove | 23:52 |
| * fungi wants a git shove subcommand now | 23:53 | |
| mordred | fungi: easy to write ... | 23:53 |
| fungi | maybe that's just an alias for push --force | 23:53 |
| mordred | fungi: that would be a funny alias for git push -f --all | 23:53 |
| fungi | heh, jinx! | 23:53 |
| corvus | oh, i'm going to accept the host key for gerrit | 23:53 |
| mordred | ++ | 23:54 |
| clarkb | corvus: oh was that added in the restart too? | 23:54 |
| corvus | clarkb: yes | 23:54 |
| mordred | corvus: yah | 23:54 |
| clarkb | gitea is still poc like though right? this is justbuilding out the bits? | 23:55 |
| corvus | clarkb: yes. i just want to get data into it for testing | 23:55 |
| mordred | yup | 23:56 |
| clarkb | We probably dont want to switch to it until we have the whole deployment story automatedor at least written down (for k8s and k8s workload) | 23:56 |
| clarkb | cool | 23:56 |
| corvus | this is still the throwaway testing cluster | 23:56 |
| mordred | yup. although we've actually got pretty much most everything we've been doing automated and reproducible | 23:56 |
| clarkb | rgr | 23:56 |
| corvus | gerrit will be fine if we shut it down | 23:56 |
| mordred | so - I'm fairly confident in our ability to delete the whole thing and create a whole new one from scratch | 23:57 |
| corvus | yes, we've built the whole system probably more than 50 times already :) | 23:57 |
| mordred | right? | 23:57 |
| clarkb | it will pile up replication tasks. Which iant the end of the world as long as wedont oom on that queue growing | 23:57 |
| corvus | what's the way to tell ssh to use an rsa key? | 23:57 |
| mordred | -i ? | 23:57 |
| corvus | clarkb: it maxes out at the number of repos | 23:57 |
| clarkb | oh right | 23:57 |
| corvus | i mean the host key | 23:57 |
| corvus | i need to convince openssh to accept the rsa host key rather than the ecdsa one i think | 23:58 |
| clarkb | ssh keyscan -t rsa | 23:58 |
| clarkb | corvus: ^ ? | 23:59 |
| fungi | if you don't create an ecdsa key it shouldn't try to query the host's ecdsa key anyway | 23:59 |
| corvus | clarkb: that worked, thanks | 23:59 |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!