| *** rlandy is now known as rlandy|bbl | 00:11 | |
| *** mrhillsman is now known as mrhillsman_in_mm | 00:11 | |
| *** ChanServ changes topic to "Discussion of the project gating system Zuul | Website: https://zuul-ci.org/ | Docs: https://zuul-ci.org/docs/ | Source: https://git.zuul-ci.org/cgit | Channel logs: http://eavesdrop.openstack.org/irclogs/%23zuul/" | 00:11 | |
| *** ChanServ changes topic to "Discussion of the project gating system Zuul | Website: https://zuul-ci.org/ | Docs: https://zuul-ci.org/docs/ | Source: https://git.zuul-ci.org/ | Channel logs: http://eavesdrop.openstack.org/irclogs/%23zuul/" | 00:12 | |
| *** mrhillsman_in_mm is now known as mrhillsman | 00:13 | |
| *** openstack has joined #zuul | 00:17 | |
| *** ChanServ sets mode: +o openstack | 00:17 | |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul-website master: Update links to git.zuul-ci.org https://review.openstack.org/557099 | 00:18 |
|---|---|---|
| openstackgerrit | James E. Blair proposed openstack-infra/zuul master: Use git.zuul-ci.org https://review.openstack.org/557100 | 00:21 |
| corvus | i'm really happy with the way the channel topic is shaping up :) | 00:33 |
| *** acozine1 has joined #zuul | 00:50 | |
| *** odyssey4me has quit IRC | 00:53 | |
| *** odyssey4me has joined #zuul | 00:53 | |
| *** JasonCL has joined #zuul | 00:55 | |
| *** JasonCL has quit IRC | 00:56 | |
| tristanC | mordred: the web api route change looks good to me, though it seems like we are doing this because we have one html file per interface page and it's boring to rewrite those from tenant scope | 01:08 |
| tristanC | mordred: i think we should look into having a single index.html page served at /, and make it switch controller and crawl the api (e.g. without loading another html file) | 01:09 |
| tristanC | mordred: then that page could load the asset bundle from a /static location, that would also simplify redirection rule a lot | 01:10 |
| tristanC | but given what we have today works, i think it's worthy to just split the html and json route at the top-level | 01:13 |
| openstackgerrit | Tristan Cacqueray proposed openstack-infra/zuul master: autohold: prevent using ',' in change argument https://review.openstack.org/556329 | 01:15 |
| *** JasonCL has joined #zuul | 01:20 | |
| *** JasonCL has quit IRC | 01:25 | |
| *** flepied has quit IRC | 01:49 | |
| *** acozine1 has quit IRC | 01:52 | |
| *** Wei_Liu has quit IRC | 02:17 | |
| *** JasonCL has joined #zuul | 02:17 | |
| *** JasonCL has quit IRC | 02:22 | |
| corvus | tristanC: in fact, i believe that's actually what we have today -- if you load any of the html files, you get the same thing. the angular router determines which page to show based on the url, and when you click the links in the navigation bar, it doesn't reload the app, it just changes the url bar and the page it displays. | 02:48 |
| corvus | tristanC: however, i think we always want to keep the different html urls working -- ie, we always want to be able to deep-link to '/builds', etc, so i think that means this work will continue to be useful. | 02:50 |
| corvus | tristanC: er, there's a possibility that i'm thinking of something further down mordred's stack. i'm a little fuzzy on whether that's true today, or only after the upgrade to angular v5. anyway, it's close enough to being true -- hopefully that will land soon too. :) | 02:52 |
| tristanC | corvus: iiuc, angular routing usually rely on anchor "#" based path for deep-link. though if the angular v5 patch manage to keep the .html path along with angular router, then that sounds good to me too | 03:33 |
| tristanC | glad mordred did the refactor, it's really appreciated | 03:38 |
| tristanC | anyway i left a -1 on 556967 as it seems like it will break the builds page | 03:40 |
| *** sshnaidm_ has joined #zuul | 04:08 | |
| *** sshnaidm has quit IRC | 04:08 | |
| *** sshnaidm_ has quit IRC | 04:32 | |
| *** sshnaidm_ has joined #zuul | 04:48 | |
| *** sshnaidm_ has quit IRC | 05:09 | |
| *** sshnaidm has joined #zuul | 05:24 | |
| *** rlandy|bbl has quit IRC | 05:43 | |
| tristanC | mordred: would you mind rebasing the angular v5 patch on top of the new api route? | 05:45 |
| *** xinliang has quit IRC | 06:52 | |
| *** xinliang has joined #zuul | 07:05 | |
| *** hashar has joined #zuul | 07:14 | |
| *** tosky has joined #zuul | 07:23 | |
| *** electrofelix has joined #zuul | 07:27 | |
| *** flepied has joined #zuul | 07:32 | |
| *** flepied has quit IRC | 07:48 | |
| *** flepied has joined #zuul | 08:36 | |
| *** flepied_ has joined #zuul | 09:05 | |
| *** flepied has quit IRC | 09:07 | |
| *** JasonCL has joined #zuul | 09:13 | |
| *** JasonCL has quit IRC | 09:14 | |
| *** JasonCL has joined #zuul | 09:15 | |
| *** JasonCL has quit IRC | 09:15 | |
| *** JasonCL has joined #zuul | 09:15 | |
| *** JasonCL has quit IRC | 09:38 | |
| *** flepied__ has joined #zuul | 10:05 | |
| *** flepied_ has quit IRC | 10:08 | |
| *** flepied__ is now known as flepied | 10:09 | |
| *** JasonCL has joined #zuul | 10:11 | |
| *** JasonCL has quit IRC | 10:11 | |
| *** sshnaidm has quit IRC | 10:29 | |
| *** sshnaidm has joined #zuul | 10:38 | |
| *** sshnaidm has quit IRC | 10:55 | |
| *** sshnaidm has joined #zuul | 11:07 | |
| *** odyssey4me has quit IRC | 12:22 | |
| *** odyssey4me has joined #zuul | 12:22 | |
| *** JasonCL has joined #zuul | 12:24 | |
| *** JasonCL has quit IRC | 12:30 | |
| *** dkranz has joined #zuul | 12:30 | |
| openstackgerrit | Merged openstack-infra/zuul master: Use requests instead of urllib.request in tests https://review.openstack.org/554057 | 12:36 |
| *** rlandy has joined #zuul | 12:37 | |
| *** sshnaidm is now known as sshnaidm|afk | 12:46 | |
| *** JasonCL has joined #zuul | 12:54 | |
| *** JasonCL has quit IRC | 12:58 | |
| mordred | tristanC: yes- I will rebase the angular v5 patch ... and also - thanks for the review! I'm fixing the stuff you pointed out now - and in one case I was, in fact, thinking of the angular v5 patch (which is set up to handle something for us) and forgot we didn't have that in the angularjs stack yet | 13:03 |
| mordred | tristanC: I have a temporary workaround hacked up - fixing some docs then I'll get both pushed up | 13:03 |
| tristanC | it is hard to juggle with the topic:javascript patches... would be nice if we could merge them soon | 13:11 |
| tristanC | even without the new pipeline, project and job routes, i could rebase them on top of the angularv5 patch if that helps | 13:11 |
| *** JasonCL has joined #zuul | 13:14 | |
| *** JasonCL has quit IRC | 13:24 | |
| *** elyezer has quit IRC | 13:29 | |
| *** electrofelix has quit IRC | 13:29 | |
| *** JasonCL has joined #zuul | 13:34 | |
| *** hashar is now known as hasharAway | 13:39 | |
| *** elyezer has joined #zuul | 13:40 | |
| openstackgerrit | Monty Taylor proposed openstack-infra/zuul master: Reorganize REST API and dashboard urls https://review.openstack.org/556967 | 13:51 |
| openstackgerrit | Monty Taylor proposed openstack-infra/zuul master: Upgrade to webpack 4 https://review.openstack.org/551987 | 13:51 |
| openstackgerrit | Monty Taylor proposed openstack-infra/zuul master: Use typescript for webpack config https://review.openstack.org/555948 | 13:51 |
| openstackgerrit | Monty Taylor proposed openstack-infra/zuul master: Upgrade from angularjs (v1) to angular (v5) https://review.openstack.org/551989 | 13:51 |
| openstackgerrit | Monty Taylor proposed openstack-infra/zuul master: Use glyphicons for status balls https://review.openstack.org/551992 | 13:51 |
| mordred | tristanC: yah - I'm looking forward to getting this stack done :) | 13:52 |
| mordred | tristanC: I'm ont 100% sure https://review.openstack.org/551989 is fully correct with the new routes yet - but there it is rebased and with the first attempt at updating it | 13:53 |
| mordred | tristanC: once we get the api routes landed and updated, I'll make sure 551989 is up to date ... I'll also update the pipeline/project/job route patches - I've mostly just been waiting to finish rebasing those until we're happy with the angular5 patch | 13:54 |
| *** JasonCL has quit IRC | 14:04 | |
| openstackgerrit | David Shrewsbury proposed openstack-infra/nodepool master: Handle ZK session loss during node launch https://review.openstack.org/552538 | 14:16 |
| Shrews | corvus: i know you really hate the use of mock, but I cannot think of a better way to do the test for 552538 ^^ | 14:16 |
| tristanC | mordred: thanks! so i'll wait for you to propose the extra route on top of angular5 patch right? | 14:17 |
| *** JasonCL has joined #zuul | 14:18 | |
| Shrews | tristanC: your fix for session expired reminded me i needed a test for 552538 | 14:18 |
| *** flepied has quit IRC | 14:19 | |
| tristanC | Shrews: how about setting kazoo client internal state like so: https://github.com/python-zk/kazoo/blob/b4967d16b3daa4b1a00c0f408888b4f61e5486cc/kazoo/tests/test_client.py#L487 | 14:20 |
| *** JasonCL has quit IRC | 14:20 | |
| Shrews | tristanC: well, i guess we could, but i dislike poking into external libraries supposedly private areas | 14:21 |
| Shrews | tristanC: and i *think* the tests all share the same zk connection?? if so, that would affect them too | 14:22 |
| Shrews | oh, that might not actually be true | 14:23 |
| tristanC | Shrews: it didn't happen on https://review.openstack.org/#/c/535537/8/nodepool/tests/test_zk.py | 14:23 |
| Shrews | yeah, see above | 14:23 |
| Shrews | i remembered incorrectly | 14:24 |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul master: Remove nullwrap config option https://review.openstack.org/557350 | 14:25 |
| corvus | Shrews: otoh, when you can't think of any other way, that's probably the time to use mock :) -- the main risk here is that we haven't actually reproduced kazoo's behavior correctly (somewhat more likely under adverse circumstances). otoh, the benefit here is it's actually possible to exercise our code, so it's a win. :) | 14:28 |
| mordred | ++ | 14:28 |
| mordred | I find mock, judiciously used, can be helpful in tricky circumstances | 14:29 |
| mordred | tristanC: yah - although if you want to just look at https://review.openstack.org/#/c/551989/14/web/app-routing.module.ts you can see how we're using the t/:tenant/status.html type links with @angular/router (which is prettier than the hack I had to had to the angularjs version) | 14:32 |
| *** JasonCL has joined #zuul | 14:36 | |
| *** JasonCL has quit IRC | 14:38 | |
| *** JasonCL has joined #zuul | 14:39 | |
| fungi | corvus: any chance the source for the lunchtime presentation from the ptg is published anywhere? | 14:49 |
| corvus | fungi: it's https://review.openstack.org/548178 | 14:50 |
| fungi | corvus: aha, perfect, thanks! | 14:50 |
| * fungi reviews | 14:50 | |
| fungi | some of the foundation staff were asking where to find it | 14:50 |
| corvus | we still haven't really tackled 'publishing' gertty stuff to the web head-on | 14:51 |
| corvus | er presentty even | 14:51 |
| fungi | yeah, i warned them it would probably be in rst ;) | 14:51 |
| openstackgerrit | Monty Taylor proposed openstack-infra/zuul master: Reorganize REST API and dashboard urls https://review.openstack.org/556967 | 14:53 |
| mordred | corvus: speaking of presentty - I ran in to more crashing issues in la this weekend - but I was in a hurry so I did not check to see if you'd updated anything since last we poked at things | 14:54 |
| mordred | corvus, tristanC: ^^ I thnk that should take care of the things tristanC pointed out | 14:54 |
| corvus | mordred: i haven't, sorry, my 'free' time has been spoken for. i'm hoping to visit that soon (maybe this weekend) | 14:54 |
| corvus | mordred: cool, i'll review after ingesting food | 14:55 |
| mordred | corvus: also, I pushed up https://review.openstack.org/#/c/557342/ as a followup - it's in the multi-tenant portion which we aren't using in infra anyway, so I figured we could do it as followup and not need to re-review the main puppet patch | 14:56 |
| *** sshnaidm|afk is now known as sshnaidm | 14:57 | |
| LinuxJedi | any Ansible gurus able to help me out with a Zuul job? The environment variables here don't appear to be working: http://paste.openstack.org/show/716312/ | 15:04 |
| LinuxJedi | oh crap! I need more sleep, because I used '=' instead of ':' | 15:05 |
| * LinuxJedi smacks forehead | 15:05 | |
| mordred | LinuxJedi: \o/ | 15:06 |
| LinuxJedi | you'd think I'd have written enough YAML by now to spot that :) | 15:07 |
| mordred | heh | 15:07 |
| *** AJaeger has quit IRC | 15:08 | |
| LinuxJedi | a thanks to you all for your help this week: https://twitter.com/LinuxJedi/status/979013976938942464 | 15:15 |
| *** flepied has joined #zuul | 15:19 | |
| corvus | LinuxJedi: \o/ hey it passed clang *and* noop :) | 15:28 |
| LinuxJedi | everything passes except for scanbuild as scanbuild is newer than when I last ran it and it found a bug ;) | 15:29 |
| corvus | LinuxJedi: maybe a gating ci system will help ;) | 15:29 |
| LinuxJedi | although tests are currently being skipped as there is no mysql server on the builder so I'm fixing that now | 15:29 |
| mordred | LinuxJedi: woot! I sent you a little PR, just for fun | 15:30 |
| LinuxJedi | it is a dead project anyway. So I don't mind it failing. I just wanted something to demo that wasn't too complex and had some CI beforehand (it used Travis CI before) | 15:31 |
| LinuxJedi | mordred: nice! Thanks :) | 15:32 |
| pabelanger | \o/ | 15:34 |
| mordred | LinuxJedi: amusingly enough - lifeless and I originally discussed that tool back when we were still doing drizzle - we just never got far enough for it to be useful - so I figured it would be fitting :) | 15:42 |
| LinuxJedi | mordred: I assume it figures out that certain packages have different names in Fedora, etc... ? | 15:42 |
| fungi | woah... LinuxJedi returns to us! | 15:42 |
| mordred | LinuxJedi: it does not at all | 15:42 |
| mordred | LinuxJedi: it allows you to list such things and mark them appropriately | 15:43 |
| LinuxJedi | mordred: ah, ok. Something I need to be wary of, Ubuntu doesn't always have the best naming system for packages :) | 15:43 |
| LinuxJedi | fungi: unfortunately not permanently. Just to demo Zuul to people more important than me ;) | 15:43 |
| mordred | LinuxJedi: yah. that's actually one of the reasons we've found the tool helpful - attempting to standardize on package names from one distro and having something figure out whatyou meant on another distro is awful | 15:44 |
| fungi | LinuxJedi: sure, but _possibly_ longer if you can make that presentation stick ;) | 15:44 |
| LinuxJedi | mordred: completely agree, sounds good on the surface, but hell when you get into the guts of it | 15:45 |
| mordred | LinuxJedi: exactly | 15:45 |
| mordred | LinuxJedi: whereas - a human just making a list for each distro is easy | 15:45 |
| mordred | fungi: I sent him a bindep patch :) https://github.com/libattachsql/libattachsql/pull/203 | 15:45 |
| fungi | i gathered from context that's what you were talking about | 15:46 |
| LinuxJedi | fungi: if the presentation sticks then it won't be me implementing the final thing. But I'll be consulting I guess. Maybe acting as a middleman as not many here use IRC (mostly the older guys) | 15:46 |
| mordred | tristanC, corvus, pabelanger, tobiash: https://review.openstack.org/#/c/556967/ <-- that has been updated since last it had 2x +2s | 15:46 |
| fungi | LinuxJedi: either way, great to see you around for a bit! | 15:48 |
| LinuxJedi | fungi: great to see you too (and all the familiar nicks) :) | 15:48 |
| fungi | the place hasn't been the same without you ;) | 15:48 |
| LinuxJedi | you mean it has been better? ;) | 15:49 |
| fungi | well, yes, but not because you left | 15:49 |
| * fungi is mildly surprised not many devs at mariadb use irc, as an aside | 15:49 | |
| LinuxJedi | the foundation devs do, not sure about corporation devs. No one on my team does apart from me. I can't tempt them. | 15:50 |
| LinuxJedi | internally we have to use Slack *vomits slightly in mouth* | 15:51 |
| fungi | my sympathies | 15:51 |
| LinuxJedi | technically I am corporation side of things, but I do work with foundation on many things | 15:51 |
| fungi | though i've recently discovered the irc client i'm using (weechat) does have a serviceable slack plugin so when i actually need to interact with certain people via slack it at least _looks_ like any other irc channel to me | 15:52 |
| LinuxJedi | I use IRCCloud and I've heard they have a Slack connection thing too. But I've not had a chance to try it yet | 15:54 |
| *** hasharAway is now known as hashar | 16:04 | |
| *** acozine1 has joined #zuul | 16:05 | |
| clarkb | LinuxJedi: I've heard great things about irccloud frkm people that have historically resisted irc for reasons of simplicty in consumption | 16:12 |
| clarkb | pparebtly works really well for multiple devices and notifications | 16:12 |
| LinuxJedi | clarkb: it solves quite a few problems. I get notifications on my phone/watch instantly. No need to run a ZNC or similar, it stores the scrollback | 16:13 |
| LinuxJedi | I was kinda resistant at first, felt a bit like having a cloud based system own my IRC identity. But it has basically made IRC easier than Slack | 16:14 |
| corvus | mordred, tobiash, SpamapS: i'd like to land https://review.openstack.org/557350 pre-3.0 | 16:16 |
| mordred | corvus: didn't someone say at the ptg that it was important to them for some reason? | 16:16 |
| corvus | mordred: got a name? | 16:17 |
| corvus | i mean, i linked to the email thread where no one said they were using it | 16:17 |
| mordred | corvus: no. I'm only half-remembering the conversatoin - but I think it was related to the fact that on k8s bubblewrap requires containers with elevated privs | 16:18 |
| mordred | corvus: so it might just be that without nullwrap it's not possible to run zuul in a standard non-elevated k8s | 16:18 |
| corvus | mordred: tobiash said that in order to use zuul on openshift, it requires elevated privs, so he uses a non-standard openshift config for that. he also said it may be possible to find more specific capabilities to narrow that down, but he has not done the work. | 16:19 |
| mordred | corvus: yah | 16:19 |
| corvus | mordred: perhaps that's what you're remembering? | 16:19 |
| mordred | corvus: probably so, yeah | 16:20 |
| tobiash | mordred, corvus: rcarrillocruz had an experimental deployment on openshift online which would required nullwrap | 16:20 |
| tobiash | but that cannot be considered production | 16:21 |
| corvus | at any rate, for the moment, running without bwrap is untenable from a security pov. if running in an unpriviledged openshift system is important for someone, then someone needs to either address the privilege situation there, or perhaps the more-native-container work we have upcoming may prove a better solution. | 16:21 |
| mordred | corvus, tobiash: yah - I think it was ricky's deployment in openshift online I was thinking about | 16:22 |
| mordred | that said, I just pinged vbatts to see if he happens to know already the answer to the lowered privs question | 16:23 |
| mordred | corvus: I'm fine removing it - just wanted to do the double-check | 16:23 |
| tobiash | corvus: lgtm | 16:23 |
| corvus | checking is good | 16:23 |
| rcarrillocruz | Yeah, that thing was an experiment. I think writing a driver to use Kube secrets to store Zuul secrets is what we talked at Dublin | 16:24 |
| mordred | rcarrillocruz: ++ | 16:25 |
| mordred | tobiash: have a sec to review updates to https://review.openstack.org/#/c/556967 ? tristanC found some bugs in the previous version | 16:26 |
| tobiash | yah | 16:26 |
| mordred | piddle. I just saw one more little bug ... I'm going to push it up as a fllowup since it won't affect the openstack deployment | 16:28 |
| tobiash | corvus: just got a customer question, is it possible to define the gate queue in a project template? | 16:30 |
| tobiash | or does that have to be in the project stanza? | 16:30 |
| openstackgerrit | Monty Taylor proposed openstack-infra/zuul master: Remove double api prefix from tenants call https://review.openstack.org/557429 | 16:30 |
| pabelanger | I finally got access to openshift.io myself, been trying to find time to play with zuul / nodepool there. Never enough time :) | 16:31 |
| openstackgerrit | Monty Taylor proposed openstack-infra/zuul master: Upgrade to webpack 4 https://review.openstack.org/551987 | 16:36 |
| openstackgerrit | Monty Taylor proposed openstack-infra/zuul master: Use typescript for webpack config https://review.openstack.org/555948 | 16:36 |
| openstackgerrit | Monty Taylor proposed openstack-infra/zuul master: Upgrade from angularjs (v1) to angular (v5) https://review.openstack.org/551989 | 16:36 |
| openstackgerrit | Monty Taylor proposed openstack-infra/zuul master: Use glyphicons for status balls https://review.openstack.org/551992 | 16:36 |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul master: Use git.zuul-ci.org https://review.openstack.org/557100 | 16:41 |
| corvus | pabelanger: thanks :) ^ | 16:41 |
| openstackgerrit | Merged openstack-infra/zuul master: Remove nullwrap config option https://review.openstack.org/557350 | 16:42 |
| corvus | mordred: while you're at it, can you look at https://review.openstack.org/557099 ? | 16:55 |
| *** tosky has quit IRC | 16:59 | |
| *** acozine1 has quit IRC | 17:01 | |
| mordred | corvus: done | 17:03 |
| openstackgerrit | Merged openstack-infra/zuul-website master: Update links to git.zuul-ci.org https://review.openstack.org/557099 | 17:07 |
| *** hashar is now known as hasharDinner | 17:07 | |
| openstackgerrit | Merged openstack-infra/zuul master: Reorganize REST API and dashboard urls https://review.openstack.org/556967 | 17:07 |
| openstackgerrit | Monty Taylor proposed openstack-infra/zuul master: Upgrade from angularjs (v1) to angular (v5) https://review.openstack.org/551989 | 17:08 |
| Shrews | fyi, i need to afk for a while. will be available later this evening | 17:09 |
| openstackgerrit | Tobias Henkel proposed openstack-infra/zuul-website master: Fix doc links in menu https://review.openstack.org/557442 | 17:11 |
| tobiash | corvus: this fixes broken links ^ | 17:11 |
| corvus | tobiash: oops, thanks :) | 17:14 |
| openstackgerrit | Merged openstack-infra/zuul master: Use git.zuul-ci.org https://review.openstack.org/557100 | 17:16 |
| pabelanger | merge all the things! | 17:16 |
| openstackgerrit | James E. Blair proposed openstack-infra/nodepool master: Switch docs to alabaster theme https://review.openstack.org/557445 | 17:18 |
| corvus | noticed while clicking around the docs pages ^ :) | 17:18 |
| openstackgerrit | Merged openstack-infra/zuul-website master: Fix doc links in menu https://review.openstack.org/557442 | 17:19 |
| tobiash | corvus: that was the next thing I had in mind ;) | 17:22 |
| tobiash | mordred: is it possible/easy to make the update interval of the status page configurable? | 17:23 |
| tobiash | some of my customers are impatient when looking at the status page | 17:24 |
| mordred | heh | 17:24 |
| tobiash | it's currently hard coded to 5s if I looked right | 17:25 |
| mordred | tobiash: one of the next things on my list is refactoring the code that fetches things into an angular service | 17:25 |
| tobiash | so I guess that makes it easy | 17:25 |
| corvus | i'd really like to have a system that pushes events out over websockets. | 17:26 |
| mordred | tobiash: should be decently easy to include refresh config while doing that - are you thinking changing the global refresh? or a per-user setting? | 17:26 |
| mordred | corvus: ++ | 17:26 |
| tobiash | mordred: global (per deployment?) | 17:26 |
| mordred | tobiash: nod. that is *very* easy to add | 17:27 |
| corvus | right now, if you look at the openstack status page, you're fetching 1MB every 5s. it's often much larger. | 17:27 |
| tobiash | zuul-web itself caches the status for 1s so I probably would set it in my deployment to 1s or 2s | 17:27 |
| mordred | corvus: or maybe even a streaming gRPC API and corresponding web code :) | 17:27 |
| tobiash | corvus: yah, for openstack 5s is probably on the lower end which is effordable | 17:27 |
| corvus | main thing is -- i think it's okay to make it adjustable in the near term, but we really need to redesign it so we're not moving so much data around (and not redrawing the page so much) | 17:28 |
| tobiash | corvus: so I guess it's already now a problem for openstack | 17:29 |
| mordred | corvus: ++ | 17:33 |
| openstackgerrit | Monty Taylor proposed openstack-infra/zuul master: Add build-time configurable status refresh variable https://review.openstack.org/557450 | 17:33 |
| mordred | tobiash: ^^ there you go | 17:33 |
| mordred | tobiash: if you do npm build:dist -- --define "ZUUL_STATUS_REFRESH=1000" with that patch | 17:34 |
| tobiash | mordred: whoa that was fast | 17:34 |
| tobiash | thank you :) | 17:34 |
| mordred | tobiash: it should build you a dashboard that refreshes every second | 17:34 |
| mordred | tobiash: I understand the dashboard codebase now :) | 17:34 |
| mordred | tobiash, corvus: per corvus' questoin on the patch - I think we can improve upon that patch pretty easily to have it pull the info from /api/info after we get the angular5 patch in shape | 17:39 |
| corvus | mordred: can we maybe defer it until then? | 17:40 |
| corvus | mordred: opening up the door to build-time user-customization of the dashboard seems like a non-trivial thing which i'd like to avoid | 17:41 |
| tobiash | yah, that sounds even better | 17:42 |
| tobiash | it's not urgent for me as I can just stick a minimal patch onto my stack for now | 17:43 |
| tobiash | mordred: will the url changes break anything for a multi tenant deployment below /zuul/ ? | 17:46 |
| mordred | tobiash: not the initial changes. once we land angular5 a multi-tenant sub-url deployment will need to build with --define "ZUUL_BASE_HREF='https://zuul.example.com/zuul'" | 17:51 |
| mordred | or, rather, | 17:51 |
| mordred | --define "ZUUL_BASE_HREF='/zuul'" | 17:51 |
| mordred | (full url is not needed - just the suburl path) | 17:52 |
| tobiash | mordred: how will that work together with 'python setup.py install'? | 17:52 |
| tobiash | that's my current installation method within the container | 17:53 |
| mordred | not particularly well - so let's make sure we solve that method before we land it | 17:53 |
| mordred | tobiash: k. I think I see a way to fix that | 17:54 |
| tobiash | yay | 17:54 |
| mordred | tobiash: also - it's possible we'll be able to solve the need for ZUUL_BASE_HREF before landing that patch as well | 18:06 |
| *** flepied_ has joined #zuul | 18:10 | |
| *** flepied has quit IRC | 18:11 | |
| *** flepied__ has joined #zuul | 18:22 | |
| * jlk is currently uploading a new github3.py release to pypi | 18:25 | |
| *** flepied_ has quit IRC | 18:26 | |
| tobiash | mordred: found a bug | 18:26 |
| tobiash | tenant overview requests /zuul/tenants | 18:26 |
| tobiash | which is actually on /zuul/api/tenants | 18:26 |
| jlk | and done! | 18:27 |
| jlk | corvus: github3.py 1.0.2 is released. | 18:27 |
| *** jimi_|ansible has joined #zuul | 18:29 | |
| *** jimi|ansible has quit IRC | 18:29 | |
| tobiash | mordred: also shift-reload doesn't fix it | 18:29 |
| tobiash | I'll revert to before the url change for now | 18:29 |
| mordred | tobiash: hrm. I thought we had that covered. :( will investigate | 18:30 |
| EmilienM | corvus: hey good morning - in zuul v2.5, do you know if we can have a layout with skip-if + all-files-match-any + a regex that is *negative* for example: https://github.com/rdo-infra/review.rdoproject.org-config/blob/master/zuul/upstream.yaml#L165-L171 (which would basically means: never run except when these 2 files are touched) | 18:31 |
| EmilienM | I think the code for that is here: https://git.openstack.org/cgit/openstack-infra/zuul/tree/zuul/cmd/migrate.py#n1450 | 18:34 |
| EmilienM | but I just need confirmation my snippet can work | 18:34 |
| EmilienM | the other option is to use "files:" layout | 18:36 |
| *** corvus is now known as kermitf | 18:42 | |
| kermitf | jlk: yaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaay! | 18:42 |
| *** kermitf is now known as corvus | 18:42 | |
| jlk | can't keep up with the name changes :D | 18:43 |
| openstackgerrit | Monty Taylor proposed openstack-infra/zuul master: Upgrade to webpack 4 https://review.openstack.org/551987 | 18:43 |
| openstackgerrit | Monty Taylor proposed openstack-infra/zuul master: Use typescript for webpack config https://review.openstack.org/555948 | 18:43 |
| openstackgerrit | Monty Taylor proposed openstack-infra/zuul master: Upgrade from angularjs (v1) to angular (v5) https://review.openstack.org/551989 | 18:43 |
| openstackgerrit | Monty Taylor proposed openstack-infra/zuul master: Remove need for ZUUL_BASE_HREF build time option https://review.openstack.org/557472 | 18:43 |
| corvus | jlk: all you need to know is there is a green frog waving his arms widly in celebration :) | 18:43 |
| jlk | hahah | 18:43 |
| jlk | oh let me do the next thing too | 18:43 |
| mordred | corvus, tobiash: ^^ //review.openstack.org/557472 should remove the need for ZUUL_BASE_HREF - and I updated the job definition to stop using it so we should be able to verify it works (the preview daashboard does not work on logs.o.o otherwise) | 18:44 |
| corvus | https://i.imgur.com/IvFsn8W.gif | 18:46 |
| mordred | tobiash: which ref did you deploy? just through https://review.openstack.org/#/c/556967 ? | 18:47 |
| tobiash | current master | 18:47 |
| tobiash | ;) | 18:47 |
| *** jimi_|ansible has quit IRC | 18:47 | |
| mordred | kk | 18:47 |
| pabelanger | lol, corvus seems excited | 18:47 |
| corvus | EmilienM: i think files: is the way to go there -- any reason not to? | 18:47 |
| clarkb | what is the excitement? | 18:47 |
| tobiash | mordred: to be more precise: fddf8f9d + my staging patches | 18:48 |
| mordred | tobiash: sweet. thanks | 18:48 |
| corvus | clarkb: github3 release | 18:48 |
| EmilienM | corvus: yeah I think files is the way to go, please look https://review.rdoproject.org/r/13130 if you have 2 minutes. Thanks a lot | 18:49 |
| tobiash | now while I'm at breaking production, I might take the latest gh3 release in ;) | 18:49 |
| tobiash | but I think it'l be better to test that in the dev instance... | 18:49 |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul master: Use github3.py >= 1.0.2 https://review.openstack.org/557475 | 18:49 |
| corvus | tobiash, jlk: ^ | 18:50 |
| jlk | crap | 18:50 |
| mordred | tobiash: fddf8f9d5a37b8fc8a2fc0752122358df03c17ae is the nullwrap patch - it shouldn't have the api changes | 18:50 |
| jlk | I had just opened one | 18:50 |
| corvus | jlk: i'm very happy to abandon mine in favor of yours :) | 18:50 |
| tobiash | mordred: oops, that's the ref, I'm currently reverting to | 18:50 |
| mordred | tobiash: (asking followup questions so that I can make sure I'm reproducing properly) | 18:50 |
| mordred | tobiash: ok. cool | 18:50 |
| mordred | tobiash: that makes more sense then | 18:50 |
| tobiash | mordred: I meant 1243cf75 + my patches | 18:50 |
| tobiash | ;) | 18:50 |
| tobiash | sorry | 18:51 |
| corvus | jlk: oh that's what you meant by the next thing. sorry, i thought you were off to do something in the github3.py repo | 18:51 |
| openstackgerrit | Jesse Keating proposed openstack-infra/zuul master: Use latest release of github3.py https://review.openstack.org/557478 | 18:51 |
| mordred | tobiash: sweet! that makes sense (sort of, I'd actually expect that patch to break in the opposite direction - for the web to try to request from /zuul/api/api/tenants | 18:51 |
| jlk | hah yours is more correct | 18:51 |
| jlk | I can close mine and use yours | 18:52 |
| mordred | but I'll figure it out | 18:52 |
| corvus | jlk: too late, i abandoned. better fix yours :) | 18:52 |
| jlk | (I had errant spaces) | 18:52 |
| dmsimard | I'm not exactly sure what are the ramifications (positive or negative) but in 2.5, Ansible added a config lookup i.e, https://gist.github.com/bcoca/1800770fe98975dde28d511ae9597337 | 18:52 |
| openstackgerrit | Jesse Keating proposed openstack-infra/zuul master: Use latest release of github3.py https://review.openstack.org/557478 | 18:52 |
| dmsimard | I suppose since we ban all lookups it's probably not an issue | 18:52 |
| jlk | that's the first thing I've thrown at gerrit since getting my github laptop | 18:53 |
| mordred | tobiash: OH | 18:54 |
| mordred | tobiash: actually - I wonder ... in your container build, is it re-using the source dir? | 18:54 |
| mordred | tobiash: what do you want to be the issue is that since there were already js files in place that they didn't get rebuilt? | 18:55 |
| mordred | s/be/bet/ | 18:55 |
| corvus | jlk: seems to still work! :) | 18:55 |
| jlk | heh, I had to install git review and change the git config to use the right email address for this repo, just minor setup things. | 18:56 |
| tobiash | mordred: I'm not sure I groked your question correctly, my container build basically does: git clone zuul; cd srcdir; pip install requirements; python3 setup.py install | 18:56 |
| tobiash | mordred: and the container is built without cache | 18:57 |
| tobiash | mordred: http://paste.openstack.org/show/716364/ | 19:01 |
| tobiash | that's the snippet that installs zuul in the container | 19:01 |
| tobiash | (but I'm always open for better ways ;) ) | 19:01 |
| mordred | tobiash: ok good | 19:04 |
| jlk | is that drastically different from just `pip install .` ? | 19:07 |
| tobiash | jlk: I think the problem with pip install was no npm installation of zuul-web | 19:08 |
| tobiash | but not sure anymore | 19:08 |
| jlk | ooooh yeah, it might not do some of those extra steps | 19:08 |
| tobiash | corvus, jlk: zuul startup was successful with gh3.py 1.0.2 | 19:13 |
| jlk | oh cool.. CI failed | 19:13 |
| jlk | github3.py>=1.0.2 | 19:14 |
| jlk | er | 19:14 |
| jlk | No matching distribution | 19:14 |
| jlk | Is there caching going on inside of Zuul? | 19:14 |
| tobiash | jlk: openstack is using local mirrors | 19:14 |
| tobiash | maybe it wasn't updated yet? | 19:14 |
| jlk | gotcha. Any idea how frequently they sync? | 19:14 |
| tobiash | my installation at least worked | 19:15 |
| tobiash | and zuul started | 19:15 |
| tobiash | now testing a job | 19:15 |
| tobiash | builds also work | 19:18 |
| *** jimi_|ansible has joined #zuul | 19:20 | |
| *** jimi_|ansible has joined #zuul | 19:20 | |
| tobiash | looks like we don't have to vendor github3.py :) | 19:21 |
| pabelanger | jlk: bandersnatch runs every 5mins | 19:22 |
| tobiash | pabelanger: how long does a run take? | 19:24 |
| pabelanger | not sure, depends how busy pypi is | 19:24 |
| pabelanger | I can check logs | 19:24 |
| tobiash | pabelanger: it's not urgent, we also can just recheck later | 19:24 |
| pabelanger | 2018-03-28 18:30:02,913 DEBUG: Getting https://pypi.python.org/packages/13/74/383f0479f3c104edb97a38b1936c62a2b26e7fd807adb4f7e8be8f195aa3/github3.py-1.0.2.tar.gz (serial None) | 19:25 |
| pabelanger | should be in mirrors now | 19:25 |
| pabelanger | give a recheck and lets see | 19:25 |
| tobiash | is the clock correct? | 19:25 |
| pabelanger | UTC | 19:25 |
| pabelanger | but ya, should be | 19:26 |
| tobiash | 18:30 utc is 50mins ago | 19:26 |
| tobiash | I rechecked just 5mins ago | 19:26 |
| pabelanger | hmm | 19:26 |
| pabelanger | lets move to openstack-infra | 19:27 |
| *** acozine1 has joined #zuul | 19:41 | |
| tobiash | corvus: next week I'll be out of office | 19:43 |
| openstackgerrit | Monty Taylor proposed openstack-infra/zuul master: Improve root api endpoint finding https://review.openstack.org/557429 | 19:50 |
| openstackgerrit | Monty Taylor proposed openstack-infra/zuul master: Add build-time configurable status refresh variable https://review.openstack.org/557450 | 19:50 |
| openstackgerrit | Monty Taylor proposed openstack-infra/zuul master: Upgrade to webpack 4 https://review.openstack.org/551987 | 19:50 |
| openstackgerrit | Monty Taylor proposed openstack-infra/zuul master: Use typescript for webpack config https://review.openstack.org/555948 | 19:51 |
| openstackgerrit | Monty Taylor proposed openstack-infra/zuul master: Upgrade from angularjs (v1) to angular (v5) https://review.openstack.org/551989 | 19:51 |
| openstackgerrit | Monty Taylor proposed openstack-infra/zuul master: Improve root api endpoint finding https://review.openstack.org/557429 | 19:54 |
| openstackgerrit | Monty Taylor proposed openstack-infra/zuul master: Add build-time configurable status refresh variable https://review.openstack.org/557450 | 19:54 |
| openstackgerrit | Monty Taylor proposed openstack-infra/zuul master: Upgrade to webpack 4 https://review.openstack.org/551987 | 19:54 |
| openstackgerrit | Monty Taylor proposed openstack-infra/zuul master: Use typescript for webpack config https://review.openstack.org/555948 | 19:54 |
| openstackgerrit | Monty Taylor proposed openstack-infra/zuul master: Upgrade from angularjs (v1) to angular (v5) https://review.openstack.org/551989 | 19:54 |
| corvus | tobiash: hopefully we can release this week then :) | 19:54 |
| tobiash | is ^ the last blocker? | 19:55 |
| tobiash | one monty patch bomb till release? | 19:56 |
| tobiash | ;;) | 19:56 |
| corvus | afaik, yes | 19:56 |
| corvus | mordred: though this one-liner would be nice https://review.openstack.org/557445 | 19:56 |
| corvus | tobiash: we don't need all of that ... i think; just the first part unless mordred has run into problems that need the upgrade | 19:57 |
| corvus | tobiash: (aiui, mordred is continuing to refresh the later patches to make sure we don't back ourselves into a corner, but they shouldn't be necessary for release) | 19:58 |
| mordred | yah | 19:59 |
| tobiash | mordred: https://review.openstack.org/555185 would also be nice | 20:00 |
| mordred | tobiash: done | 20:00 |
| tobiash | already had a few people asking where to put the depends-on | 20:01 |
| tobiash | :) | 20:01 |
| mordred | tobiash, corvus: I think https://review.openstack.org/557429 should solve the issue tobiash hit. I'm working on trying to verify that right now | 20:01 |
| tobiash | mordred: thanks, I'll try that tomorrow (not in production this time ;) ) | 20:02 |
| mordred | tobiash: :) | 20:02 |
| mordred | tobiash: hopefully I'll have verified it before then :) | 20:02 |
| tobiash | :) | 20:03 |
| openstackgerrit | Merged openstack-infra/nodepool master: Switch docs to alabaster theme https://review.openstack.org/557445 | 20:05 |
| openstackgerrit | Monty Taylor proposed openstack-infra/zuul master: Improve root api endpoint finding https://review.openstack.org/557429 | 20:18 |
| mordred | corvus: ^^ ok. I believe I have verified that does the correct thing locally | 20:18 |
| mordred | corvus: I did this by doing a build of the html, telling apache to serve it from my zuul directory (so that I'd have some suburl in there) | 20:19 |
| mordred | then going to http://localhost/zuul/web/static/tenants.html, opening the web debugger and seeing that it tried to fetch from http://localhost/zuul/web/static/api/tenants | 20:19 |
| mordred | corvus: and then checking the non-suburl case - which also worked | 20:22 |
| mordred | adding rewrite rules now so I can test multi-tenant | 20:23 |
| openstackgerrit | Merged openstack-infra/zuul master: Clarify location of depends-on footer https://review.openstack.org/555185 | 20:23 |
| tobiash | mordred: PS2 didn't work ;) | 20:26 |
| tobiash | rebuilding with ps4 | 20:26 |
| corvus | mordred: what does "return path.split('/').slice(0, -1).join('/') + '/'" do? does that just mean "make sure there's a single / on the end of the path"? | 20:26 |
| mordred | corvus: well, the '/' ensures the / - the other bits are basically "pop off the last path segment" | 20:27 |
| corvus | mordred: oh, that's what -1 means in this context | 20:27 |
| corvus | got it -- that jives with what you just said about your test too | 20:28 |
| mordred | ya - the equiv python would be path.split('/')[:-1].join('/') | 20:28 |
| corvus | not all that different | 20:28 |
| mordred | so - I just realized I can rewrite rule apache to point to the live zuul api service - even with it not having the /api running | 20:28 |
| mordred | I have done that, and localhost/status.html properly shows me status output | 20:28 |
| mordred | hopefully tobiash will similarly get functional results from his test | 20:29 |
| tobiash | we'll see in 10-15minutes | 20:29 |
| * mordred crosses fingers | 20:31 | |
| tobiash | mordred: just wondered why it's still not working | 20:44 |
| tobiash | and then I noticed that I cherry picked the new version of the change but forgot to push up the branch :/ | 20:44 |
| tobiash | I guess it's too late... | 20:44 |
| openstackgerrit | Monty Taylor proposed openstack-infra/zuul master: Use reno for release notes https://review.openstack.org/262597 | 20:47 |
| openstackgerrit | Monty Taylor proposed openstack-infra/zuul master: Use reno for release notes https://review.openstack.org/262597 | 20:50 |
| openstackgerrit | Monty Taylor proposed openstack-infra/zuul master: Use reno for release notes https://review.openstack.org/262597 | 20:52 |
| openstackgerrit | Colleen Murphy proposed openstack-infra/zuul-jobs master: Fix ansible-lint job https://review.openstack.org/557507 | 20:53 |
| openstackgerrit | Monty Taylor proposed openstack-infra/nodepool master: Add reno releasenotes support https://review.openstack.org/536861 | 20:53 |
| *** dkranz has quit IRC | 20:57 | |
| *** acozine1 has quit IRC | 21:02 | |
| tobiash | mordred: dashboard works with https://review.openstack.org/557429 | 21:03 |
| tobiash | in my dev env | 21:03 |
| mordred | tobiash: woot! | 21:03 |
| pabelanger | jlk: tobiash: recheck of github3 patch should work now | 21:03 |
| pabelanger | bandersnatch did a release | 21:03 |
| jlk | trying | 21:04 |
| mordred | tobiash: you ok with landing it and moving forward with rolling it out to openstack? | 21:04 |
| tobiash | mordred: currently reviewing it | 21:04 |
| mordred | sweet | 21:04 |
| mordred | tobiash: thanks for looking at it - I know it's late there | 21:05 |
| tobiash | 23:05 ;) | 21:05 |
| tobiash | +3 | 21:05 |
| tobiash | so, now really finished for today | 21:08 |
| tobiash | cya | 21:08 |
| jlk | Wouldn't be the first time one of us were up to 2300~ plugging away at Zuul... | 21:09 |
| mordred | tobiash: have a good night! | 21:09 |
| tobiash | Thanks | 21:10 |
| openstackgerrit | Monty Taylor proposed openstack-infra/zuul master: Upgrade to webpack 4 https://review.openstack.org/551987 | 21:10 |
| openstackgerrit | Monty Taylor proposed openstack-infra/zuul master: Use typescript for webpack config https://review.openstack.org/555948 | 21:10 |
| openstackgerrit | Monty Taylor proposed openstack-infra/zuul master: Upgrade from angularjs (v1) to angular (v5) https://review.openstack.org/551989 | 21:10 |
| openstackgerrit | Monty Taylor proposed openstack-infra/zuul master: Use glyphicons for status balls https://review.openstack.org/551992 | 21:10 |
| openstackgerrit | Colleen Murphy proposed openstack-infra/zuul-jobs master: Fix ansible-lint job https://review.openstack.org/557507 | 21:11 |
| openstackgerrit | Merged openstack-infra/zuul master: Improve root api endpoint finding https://review.openstack.org/557429 | 21:22 |
| *** rlandy is now known as rlandy|brb | 21:22 | |
| openstackgerrit | Merged openstack-infra/zuul master: Use latest release of github3.py https://review.openstack.org/557478 | 21:38 |
| corvus | welp, that's us at zero release blockers | 21:38 |
| pabelanger | woah, that is awesome | 21:38 |
| corvus | i will start work to prepare for the release; if anyone thinks of anything we missed, now would be a good time to mention it. :) | 21:39 |
| clarkb | release tomorrow? or even now I guess? | 21:39 |
| pabelanger | should we restart all openstack services to pick up latest master as smoke test? | 21:40 |
| clarkb | corvus maybe just make sure the docs updates from bernd and shrews and linuxjedi are in? | 21:40 |
| corvus | pabelanger: yes, i think mordred was going to finish that | 21:40 |
| clarkb | pabelanger: probavly not a bad Idea | 21:40 |
| corvus | clarkb: i think those changes are only in an etherpad right now :| | 21:41 |
| clarkb | :( | 21:41 |
| clarkb | I'm at lax gate to head home now so not sure I can convert those currently. Last flight I didnt have wifif either | 21:41 |
| *** hasharDinner has quit IRC | 21:42 | |
| corvus | clarkb: i know we're not fully integrated with the openstack foundation yet, and we don't expect to do a marketing push around the release, but i figure we should at least give a heads up, so i'll send an email. | 21:43 |
| corvus | then probably release asap after that | 21:43 |
| clarkb | corvus: sounds good | 21:45 |
| clarkb | let me know if you want me to ping them too | 21:45 |
| corvus | clarkb: yeah, if you're about to hop on a plane, maybe just give them a heads up to expect an email from me; i'm writing something now. | 21:46 |
| clarkb | ok doing that now | 21:46 |
| *** rlandy|brb is now known as rlandy | 21:48 | |
| *** harlowja has joined #zuul | 21:53 | |
| -openstackstatus- NOTICE: the zuul web dashboard will experience a short downtime as we roll out some changes - no job execution should be affected | 21:53 | |
| EmilienM | corvus, pabelanger: if I could use 2 min of your time, would you please look at the comment on https://review.rdoproject.org/r/#/c/13130/1/zuul/upstream.yaml and tell if that's true for zuul v2.5 ? I'm trying to stop wasting resources used by RDO CI | 22:04 |
| corvus | EmilienM: i wonder if the job below it is undoing that | 22:07 |
| EmilienM | corvus: ah, so I might have to move the job below to be the first | 22:07 |
| EmilienM | and then the other ones more restrictive, below | 22:08 |
| corvus | EmilienM: tbh, i'm not positive how they interact. let me see if i can answer that | 22:08 |
| EmilienM | thanks | 22:08 |
| corvus | EmilienM: i think either way you'll end up with both 'files' and 'skip-if'. but, if you put the more-specific job after the less-specific one, and add "skip-if: []" to the more-specific one, i think that will remove the skip-if from it. | 22:11 |
| EmilienM | corvus: ok and then I would have to keep skip-if and duplicate the list of "all-files-match-any" from the more general rule into the specific rules, if I want to keep the skip-if, right? | 22:13 |
| EmilienM | in fact it's doing merged as you go down | 22:13 |
| EmilienM | iiuc | 22:13 |
| EmilienM | merge* | 22:13 |
| EmilienM | I'll try something | 22:15 |
| corvus | EmilienM: basically, any job regexes which match the job name get applied to the job in order; later ones override earlier values | 22:16 |
| EmilienM | ok | 22:17 |
| EmilienM | corvus: here is my change: https://review.rdoproject.org/r/#/c/13130/1..2/zuul/upstream.yaml | 22:17 |
| corvus | EmilienM: yep -- assuming all those regexes make sense and match the right things (i'm gonna leave that to you :) i think that should work. | 22:18 |
| EmilienM | ok thanks for your help | 22:19 |
| corvus | EmilienM: oh, wait, i think i spotted a problem | 22:19 |
| * EmilienM stops moving | 22:21 | |
| corvus | EmilienM: i think "skip-if: []" isn't enough to undo that matcher. you might have to do something like "skip-if: [{branch: doesnotexist}]" | 22:22 |
| corvus | EmilienM: (there's a simple "if skip_if_matcher" in the code, so the empty list won't create an override; we'd have to add a matcher that will never match) | 22:22 |
| EmilienM | oh ok I see | 22:24 |
| EmilienM | corvus: what if I don't set skip-if? it'll take the one from above I guess, and that's it? | 22:26 |
| EmilienM | if yes I can remove the fake override, since I'm interested by the "files:" for this one, and the skip-if can't hurt anyway | 22:26 |
| EmilienM | corvus: I removed it: https://review.rdoproject.org/r/#/c/13130/2..3/zuul/upstream.yaml | 22:33 |
| openstackgerrit | Monty Taylor proposed openstack-infra/zuul master: Upgrade from angularjs (v1) to angular (v5) https://review.openstack.org/551989 | 22:45 |
| openstackgerrit | Monty Taylor proposed openstack-infra/zuul master: Use glyphicons for status balls https://review.openstack.org/551992 | 22:45 |
| mordred | corvus: ^^ those now work against the new api server on my local machine. yay | 22:46 |
| corvus | mordred: have we landed enough of the api stack for the openstack live job to work? | 22:49 |
| mordred | corvus: yup | 22:49 |
| corvus | cool | 22:49 |
| pabelanger | mordred: mostly curious, but does any of your javascript changes add back the time remaining on the progress bar of jobs? Want to say we lost that function at some point, but don't know why | 22:55 |
| mordred | pabelanger: I don't think so - I know a change in the not-too-distant-future will bring back the sparklines though | 22:56 |
| mordred | pabelanger: however, we're in a position soon to refactor the status page, which I think will be nice - so it should be easier to reason about re-adding things | 22:56 |
| pabelanger | ++ | 22:57 |
| mordred | btw - this works and I think is cool | 22:58 |
| mordred | $ curl http://zuul.enstack.org/api/info | 22:58 |
| mordred | {"info": {"stats": {"url": null, "prefix": null, "type": "graphite"}, "capabilities": {"job_history": true}, "websocket_url": null, "tenant": "openstack", "rest_api_url": null}} | 22:58 |
| mordred | corvus: ooh. I want to remove something from that data-structure that's not actually used | 22:58 |
| mordred | corvus: (before we release - tiny change coming) | 22:59 |
| clarkb | whats an enstack? | 22:59 |
| mordred | clarkb: wow. a really weird copy/pasta | 22:59 |
| clarkb | o and p werent invited to the arty | 23:01 |
| openstackgerrit | Monty Taylor proposed openstack-infra/zuul master: Remove rest_api_url from info payload https://review.openstack.org/557544 | 23:02 |
| mordred | clarkb, pabelanger, corvus: ^^ vestigal attribute in /info payload | 23:02 |
| corvus | mordred: ++ | 23:04 |
| mordred | corvus: are there any "we're working on" or "this is a temporary branch" messages still in the repo? | 23:05 |
| mordred | I think I remember you getting rid of all of them | 23:05 |
| corvus | okay, so my plan is, after 557544 lands, to tag master of nodepool with 3.0.0 and then send this to the zuul-announce list: https://etherpad.openstack.org/p/qrYmtbSUbX | 23:06 |
| corvus | er, tag master of zuul + nodepool | 23:06 |
| corvus | that sound okay? | 23:06 |
| corvus | mordred: i *think* we got those... but it's a big repo, may be worth some greps | 23:06 |
| mordred | corvus: I see a few - but nothing we can't clean up later | 23:06 |
| clarkb | etherpad lgtm | 23:07 |
| mordred | ++ | 23:07 |
| clarkb | the sitting on plane waiting to push back from gate is the worst btw | 23:07 |
| *** rlandy has quit IRC | 23:10 | |
| pabelanger | looking | 23:11 |
| clarkb | I'm gonna have to find a bottle of something on way home from airport I guess :) | 23:11 |
| pabelanger | etherpad is great! | 23:12 |
| pabelanger | did we want to restart nodepool / zuul services now, before we tag? | 23:13 |
| corvus | pabelanger: yeah, let's do that. 557544 should land in about 10m; let's do it then | 23:32 |
| pabelanger | ack | 23:34 |
| pabelanger | AFK for a few moments but can help | 23:34 |
| openstackgerrit | Merged openstack-infra/zuul master: Remove rest_api_url from info payload https://review.openstack.org/557544 | 23:35 |
| corvus | el, less than 10m :) | 23:35 |
| corvus | moving to -infra for restart discussion | 23:35 |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!
