| mnaser | corvus, clarkb: any easy way to run this stuff locally? | 00:00 |
|---|---|---|
| clarkb | there are js dev docs in tree | 00:03 |
| clarkb | but this quickly gets past my knowlefge set | 00:03 |
| mnaser | i think mordred did some stuff around this, let me dig | 00:04 |
| mnaser | ok | 00:06 |
| mnaser | `yarn install` then `npm start` | 00:06 |
| corvus | mnaser: yeah https://zuul-ci.org/docs/zuul/developer/javascript.html | 00:11 |
| mnaser | ok i'm doing it in a nice and clean way so that when you enter something and hit 'enter', it'll update the url in an angular-y way | 00:11 |
| mnaser | so we do things The Right Way(tm) | 00:12 |
| corvus | mnaser: also, iirc, you started some work to refactor the status page; i think that got deferred due to the angular structural changes, but i think that runway is clear now if you wanted to resume. you know, in your copius spare time. :) | 00:12 |
| mnaser | corvus: yeah i might hack on that sometime | 00:13 |
| corvus | mnaser: "npm run start:openstack" may be helpful for this. or "npm run start:basic" or the other things mentioned in that doc. | 00:13 |
| SpamapS | thank the flying spaghetti monster that we have somebody who wants to javascript. ;) | 00:13 |
| corvus | i'm looking forward to doing more as things settle down there :) | 00:14 |
| mnaser | ok this is a lot more complicated to do cleanly | 00:27 |
| mnaser | because the dashboard is semi-angular and semi-jquery | 00:27 |
| mnaser | ok i'll get it to work for now but | 00:30 |
| mnaser | yeah some work will be needed to clean it up | 00:30 |
| openstackgerrit | Mohammed Naser proposed openstack-infra/zuul master: Remove unnecessary injections https://review.openstack.org/571341 | 00:34 |
| openstackgerrit | Mohammed Naser proposed openstack-infra/zuul master: Enable HTML5 mode and remove hash prefix for status https://review.openstack.org/571342 | 00:34 |
| mnaser | corvus, SpamapS, clarkb: ^ a clean up and a fix :) | 00:34 |
| mnaser | lengthy commit message because i wanted to explain what was going on | 00:34 |
| clarkb | ++ tha ks | 00:36 |
| ilya_mochalov | hello here, I quite new to zuul and overall ecosystem. I need to integrate gitlab webhook for my zuul server. I am straggling to find a right steps to follow. before I only work with github. Thanks everyone in advance :) | 01:01 |
| *** harlowja has quit IRC | 01:19 | |
| *** xinliang has quit IRC | 01:28 | |
| *** portdirect has joined #zuul | 01:33 | |
| *** D3VIATION has joined #zuul | 01:37 | |
| *** xinliang has joined #zuul | 01:40 | |
| *** D3VIATION has quit IRC | 01:41 | |
| openstackgerrit | Brian Rosmaita proposed openstack-infra/zuul-jobs master: Handle -/_ ambiguity in package names https://review.openstack.org/571005 | 01:43 |
| *** D3VIATION has joined #zuul | 02:36 | |
| *** xinliang has quit IRC | 02:53 | |
| *** xinliang has joined #zuul | 02:54 | |
| *** xinliang has quit IRC | 02:54 | |
| *** xinliang has joined #zuul | 02:54 | |
| ilya_mochalov | ohhh...gitlab is not supported currently. Is there are anyone who works on integration? | 02:57 |
| tristanC | ilya_mochalov: probably not, at least not in review.openstack.org... perhaps mordred knows more what is the gitlab driver status? | 03:23 |
| ilya_mochalov | mordred: do you know anything about gitlab integration status? | 03:26 |
| *** ianychoi has quit IRC | 03:51 | |
| *** D3VIATION has quit IRC | 03:53 | |
| *** rlandy|rover|bbl is now known as rlandy|rover | 04:20 | |
| *** rlandy|rover has quit IRC | 04:20 | |
| *** sshnaidm_pto has quit IRC | 05:28 | |
| *** hashar has joined #zuul | 06:03 | |
| *** pcaruana has joined #zuul | 06:43 | |
| tobiash | ilya_mochalov: afaik no one has been working on gitlab integration yet | 06:50 |
| tobiash | If you want to try it, the github driver is probably a good starting point for creating a gitlab driver | 06:52 |
| *** D3VIATION has joined #zuul | 06:53 | |
| *** D3VIATION has quit IRC | 06:57 | |
| ilya_mochalov | Yeap, I will take a look at github driver | 07:06 |
| *** ssbarnea_ has joined #zuul | 07:15 | |
| *** ianychoi has joined #zuul | 07:32 | |
| johanssone | ilya_mochalov: Ping me if you need any help with a gitlab driver. I would love to have one as well. | 07:44 |
| *** jpena|off is now known as jpena | 07:49 | |
| ilya_mochalov | johanddone: Alright | 08:10 |
| ilya_mochalov | anyone can help me with status_url... in zuul.conf I have `status_url=http://my_local_ip:9000/status`.. when I run zuul-web it's gives me 404 on request for ip:port/status | 08:12 |
| ilya_mochalov | ip:port/api/tenant/quickstart/status works though | 08:13 |
| *** GonZo2000 has joined #zuul | 08:17 | |
| *** sshnaidm_pto has joined #zuul | 08:25 | |
| *** electrofelix has joined #zuul | 08:35 | |
| *** GonZo2000 has quit IRC | 08:36 | |
| tristanC | the mqtt driver (https://review.openstack.org/535543) has four +2, can somebody +W ? | 08:43 |
| openstackgerrit | Tristan Cacqueray proposed openstack-infra/zuul master: docs: add Project Testing Interface guide https://review.openstack.org/571420 | 09:09 |
| *** ttx has quit IRC | 09:52 | |
| *** ttx has joined #zuul | 10:11 | |
| *** ttx has quit IRC | 10:14 | |
| *** ttx has joined #zuul | 10:14 | |
| *** jpena is now known as jpena|lunch | 11:00 | |
| *** ttx has quit IRC | 11:07 | |
| *** ttx has joined #zuul | 11:08 | |
| *** ttx has quit IRC | 11:34 | |
| *** ttx has joined #zuul | 11:35 | |
| *** ttx has quit IRC | 11:47 | |
| *** ttx has joined #zuul | 11:47 | |
| *** ttx has quit IRC | 11:51 | |
| *** ttx has joined #zuul | 11:51 | |
| *** elyezer has quit IRC | 11:52 | |
| *** sshnaidm_pto has quit IRC | 12:02 | |
| *** elyezer has joined #zuul | 12:23 | |
| *** jpena|lunch is now known as jpena | 12:25 | |
| *** rlandy has joined #zuul | 12:28 | |
| *** rlandy is now known as rlandy|rover | 12:32 | |
| *** D3VIATION has joined #zuul | 12:48 | |
| *** Guest46098 is now known as robcresswell | 13:14 | |
| *** hughsaunders has joined #zuul | 13:15 | |
| *** sshnaidm_pto has joined #zuul | 13:31 | |
| *** acozine1 has joined #zuul | 14:23 | |
| *** hughsaunders has quit IRC | 14:28 | |
| *** D3VIATION has quit IRC | 14:38 | |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul master: Move SQL web handler to driver https://review.openstack.org/568028 | 15:01 |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul master: Add never_capture test decorator https://review.openstack.org/569515 | 15:01 |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul master: Replace use of aiohttp with cherrypy https://review.openstack.org/567959 | 15:01 |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul master: Convert streaming unit test to ws4py and remove aiohttp https://review.openstack.org/568335 | 15:01 |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul master: Use iterate timeout in streaming tests https://review.openstack.org/571498 | 15:01 |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul master: Use ZuulWebFixture in tests https://review.openstack.org/571499 | 15:01 |
| corvus | i found the problem which was causing the streaming tests to randomly fail; those patches should be good to go now. | 15:02 |
| corvus | (cherrypy re-uses the httpserver if you stop it and start it again, so if a single test runner process ran two zuulweb tests, and one of them listened on 127.0.0.1 and the other listened on :: (dual stack), the second would actually only listen on 127.0.0.1. i added something to force cherrypy to create a new webserver each time, and i also made all of the zuulweb tests consistent so they all use :: now) | 15:04 |
| openstackgerrit | Merged openstack-infra/zuul master: mqtt: add basic reporter https://review.openstack.org/535543 | 15:13 |
| *** pcaruana has quit IRC | 15:28 | |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul master: Move SQL web handler to driver https://review.openstack.org/568028 | 16:10 |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul master: Add never_capture test decorator https://review.openstack.org/569515 | 16:10 |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul master: Replace use of aiohttp with cherrypy https://review.openstack.org/567959 | 16:10 |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul master: Convert streaming unit test to ws4py and remove aiohttp https://review.openstack.org/568335 | 16:10 |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul master: Use iterate timeout in streaming tests https://review.openstack.org/571498 | 16:10 |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul master: Use ZuulWebFixture in tests https://review.openstack.org/571499 | 16:10 |
| corvus | rebase since the mqtt change merged, and a couple fixes to the last 2 patches | 16:10 |
| *** hughsaunders has joined #zuul | 16:15 | |
| *** jpena is now known as jpena|off | 16:16 | |
| corvus | and the stack is all green now; it's ready to go | 16:44 |
| *** sshnaidm_pto has quit IRC | 17:08 | |
| *** sshnaidm_pto has joined #zuul | 17:21 | |
| openstackgerrit | James E. Blair proposed openstack-infra/nodepool master: Add driver interface documentation https://review.openstack.org/571538 | 17:45 |
| pabelanger | tobiash: are you able to refresh your Dockerfile on nodepool / zuul? I haven't had time to look into pbrx yet and wanted to test docker things locally | 17:50 |
| *** electrofelix has quit IRC | 17:54 | |
| *** myoung is now known as myoung|lunch | 17:57 | |
| *** electrofelix has joined #zuul | 17:57 | |
| *** electrofelix has quit IRC | 17:58 | |
| tobiash | pabelanger: http://paste.openstack.org/show/722456/ | 18:01 |
| tobiash | pabelanger: that's my current zuul-base dockerfile | 18:01 |
| pabelanger | tobiash: great! | 18:01 |
| openstackgerrit | Merged openstack-infra/zuul master: Fix new depends-on format matching for prefixed gerrit ui https://review.openstack.org/570006 | 18:02 |
| tobiash | pabelanger: you need to specify ZUUL_BASE_REVISION and ZUUL_SRC_URL during the build | 18:03 |
| pabelanger | tobiash: just export as bash variable? or is it something docker build specific? | 18:04 |
| tobiash | pabelanger: it's docker build specific | 18:05 |
| pabelanger | ack, I'll go read up on it now | 18:05 |
| tobiash | but you also can just add ENV steps in the dockerfile locally | 18:05 |
| tobiash | pabelanger: you should be able to use --build-arg | 18:06 |
| pabelanger | thanks | 18:06 |
| *** pwhalen has quit IRC | 18:24 | |
| *** pwhalen has joined #zuul | 18:33 | |
| *** ilya_mochalov has quit IRC | 18:36 | |
| *** toabctl has quit IRC | 19:33 | |
| *** myoung|lunch is now known as myoung | 19:34 | |
| *** toabctl has joined #zuul | 19:36 | |
| pabelanger | so, trying to get RDOproject working with zuulv3 (under softwarefactory) but for some reason confguration syntax errors don't seem to be reported back to gerrit | 19:41 |
| pabelanger | eg: 2018-05-31 18:42:06,645 INFO zuul.Pipeline.rdoproject.org.check: Configuration syntax error in dynamic layout | 19:41 |
| pabelanger | but looking at debug log, I cannot see why | 19:42 |
| pabelanger | http://paste.openstack.org/show/722466/ | 19:42 |
| clarkb | pabelanger: do you have the pipeline configured to report failure? | 19:46 |
| pabelanger | clarkb: I do, but I proposed https://review.rdoproject.org/r/#/c/13984/ to copypasta check from openstack-infra | 19:49 |
| pabelanger | failure-message is missing | 19:50 |
| *** harlowja has joined #zuul | 20:01 | |
| corvus | pabelanger: failure-message shouldn't be required. it's just to further customize what's already there. | 20:03 |
| corvus | pabelanger: however, it's entirely possible the case of the "verified" vote might be a problem; does reporting work at all? | 20:04 |
| corvus | pabelanger: oh, the reason it wasn't reporting is this: | 20:04 |
| corvus | 2018-05-31 18:42:06,646 DEBUG zuul.Pipeline.rdoproject.org.check: Project rdo-jobs not in pipeline <Pipeline check> for change <Change 0x7f716815f0f0 13964,4> | 20:04 |
| corvus | pabelanger: in order for a change to report failure, the project has to be attached to the pipeline. so you won't see a syntax error on a change which adds the project to the pipeline. | 20:05 |
| pabelanger | corvus: ah, so I am guessing that is why we require system-required template in openstack-infra? | 20:08 |
| corvus | pabelanger: yep, that's part of it | 20:08 |
| pabelanger | okay, that helps | 20:08 |
| pabelanger | the part that confused me, is the previous patch in the stack has a noop check job, I guess that isn't enough for reporting back an error | 20:09 |
| pabelanger | okay, let me create a merge-check pipeline and systemd-required template and retest | 20:11 |
| pabelanger | thanks | 20:11 |
| corvus | pabelanger: we might be able to change that. currently zuul falls back on the currently running layout when reporting if a change doesn't produce a valid configuration. but we might be able to have it walk up the pipeline to find the most recent layout instead. that's in _reportItem() | 20:36 |
| pabelanger | corvus: Sure, that would be helpful. As only a zuul user on RDO, I couldn't see why my jobs were failing... since I don't have root access to SF. I can try and see if I can patch the function and create a test | 20:39 |
| *** harlowja has quit IRC | 21:00 | |
| *** sshnaidm_pto has quit IRC | 21:01 | |
| *** hashar has quit IRC | 21:30 | |
| *** ssbarnea_ has quit IRC | 21:41 | |
| *** acozine1 has quit IRC | 22:00 | |
| openstackgerrit | James E. Blair proposed openstack-infra/nodepool master: Add driver interface documentation https://review.openstack.org/571538 | 22:11 |
| openstackgerrit | James E. Blair proposed openstack-infra/nodepool master: Directly link providers and request handlers https://review.openstack.org/571584 | 22:11 |
| openstackgerrit | James E. Blair proposed openstack-infra/nodepool master: Create a base Driver class https://review.openstack.org/571585 | 22:11 |
| openstackgerrit | James E. Blair proposed openstack-infra/nodepool master: Have Drivers create Providers https://review.openstack.org/571586 | 22:11 |
| *** harlowja has joined #zuul | 22:21 | |
| *** rlandy|rover is now known as rlandy|rover|bbl | 23:09 | |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul master: Add container spec https://review.openstack.org/560136 | 23:18 |
| corvus | tristanC, clarkb, tobiash, SpamapS, mordred, jhesketh, Shrews: ^ updated | 23:18 |
| SpamapS | mmmmm | 23:19 |
| SpamapS | container | 23:19 |
| SpamapS | spec | 23:19 |
| * SpamapS consumes | 23:19 | |
| corvus | i think that addresses the bulk of the comments and what we discussed in person. i'd say the most notable thing is that it no longer favors openshift first, instead focusing on more generally applicable concepts which should be easier to get an initial implementation out the door, easier to adapt to other container techs, and still doesn't preclude building on them with something like openshift to make it | 23:20 |
| corvus | more streamlined. | 23:20 |
| corvus | it's also more than probable i missed something from all the feedbacks; if i did, i apologize in advance for asking you to repeat yourself :) | 23:21 |
| tristanC | corvus: so is the thingpool going to another spec? | 23:23 |
| tristanC | i really like the proposal of SpamapS, and it worked great and this prototype: https://review.openstack.org/#/q/topic:openshift-zuul-build-resource | 23:23 |
| SpamapS | tristanC: cool! | 23:25 |
| corvus | tristanC: that one still covers "request a k8s namespace", so i guess that's a little bit of thingpool. i didn't take it any further than that. | 23:25 |
| corvus | tristanC: at a really quick glance, that seems pretty compatible with what i wrote and was imagining. :) | 23:25 |
| corvus | that's creating a project. a plain k8s driver would do the same thing but just create a namespace. | 23:26 |
| corvus | (so to try to summarize the spec in <512 chars: part 1 is "give me a container running /bin/sh" and part2 is "give me access to a container orchestration engine") | 23:29 |
| tristanC | hum, i fail to see the "request a k8s namespace" part of the spec | 23:33 |
| corvus | tristanC: line 174 | 23:34 |
| tristanC | got it, thanks | 23:35 |
| tristanC | that workflow is really flexible since the core logic can be implemented in the job | 23:36 |
| tristanC | regarding secret sharing from nodepool to zuul, could we make the scheduler create a global key pair? | 23:37 |
| corvus | yeah, i like how we can probably get moving pretty quick on experimenting in-job, and then move things up the stack as we find them useful (eg, the openshift registry) | 23:37 |
| corvus | tristanC: zk support encryption, so i think we should just rely on that | 23:38 |
| corvus | tristanC: unless you want to make sure it's encrypted at rest? in which i guess that would work. it would be one more bit of state the scheduler (and later, multiple schedulers) need to have. but i guess it's no more overhead than the existing keys for secrets? | 23:39 |
| tristanC | well it's an extra bonus feature, not really important. I think the service account token are revoked when the project is deleted | 23:40 |
| clarkb | corvus: you mean zk supports tls over the wire protocol but not on disk right? | 23:40 |
| tristanC | but i liked the idea, and having a predicable location to fetch the public key would simplify the implementation | 23:41 |
| corvus | clarkb: i know it supports over the wire, i don't know about disk | 23:43 |
| tristanC | btw, i recorded a video of the proposed openshift-base job in action here: http://www.softwarefactory-project.io/draft/zuul-openshift.html | 23:44 |
| corvus | tristanC: in other news, i've been reviewing Shrews changes, and wanted to try to get a big-picture view of the driver api. i wrote 571538 through 571586 to do help with that. i think we know enough about dynamic drivers, static, and now containers to better shape the api. i'm hoping with a few more changes we'll get to something where you can look at the docs and see how you would implement a new driver. | 23:46 |
| tristanC | corvus: at a quick glance, those looks good to me. let me know when I should rebase my drivers | 23:47 |
| corvus | tristanC: i think we're heading toward having the actual driver api be very simple, with most of the complexity of the dynamic drivers handled by a second api that builds on top of it. after looking at your openshift driver, i imagine that will mostly mean we'll just remove some unused methods, and maybe rename a few things. :) | 23:48 |
| corvus | tristanC: will do. i'm expecting we'll make a few more changes, then say it's good enough to start landing containers. | 23:49 |
| clarkb | corvus: the getProviderHandler() and getConfig() et al being the second api? | 23:49 |
| clarkb | or rather the bits they are expected to return I guess | 23:49 |
| corvus | i think the ultimate test is if we all look at the docs and say "yep, that looks obvious" :) | 23:49 |
| corvus | clarkb: i think the second api is a bunch of stuff that the openstacknoderequesthandler does | 23:50 |
| corvus | or rather, the noderequesthandler | 23:50 |
| corvus | i think most of its methods are only needed for dynamic drivers, so i think the next thing to do is separate them out | 23:53 |
| corvus | i'm going to try to send out the status update email on monday, so if folks want to update https://etherpad.openstack.org/p/zuul-update-email that's be great :) | 23:57 |
| corvus | i'm going to head out for the day now | 23:58 |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!