| tristanC | corvus: do we have to revert the react change again? until the publication job is fixed, nothing changed... | 00:18 |
|---|---|---|
| openstackgerrit | Tristan Cacqueray proposed openstack-infra/zuul-jobs master: fetch-javascript-content-tarball: fix tarball relative path https://review.openstack.org/608794 | 00:22 |
| mordred | tristanC: I believe fixing the publicatoin job was also discussed as an alternative | 00:26 |
| mordred | tristanC: I've been travelling so I'm not 100% up to speed | 00:26 |
| mordred | tristanC: but I think the revert patch was put up as an in-parallel effort to fixing the job - although I think corvus brought up a good point which is that we don't have good (any) test coverage of the tarball creation / deployment - nor is it documented - so we should also fix those things | 00:29 |
| tristanC | mordred: sure, but could we do that without reverting? it looks like a messy git history for no good reason | 00:32 |
| corvus | mordred, tristanC: yeah, it's mostly that i can't take the time to babysit untested fixes right now. so if it's up to me, i'm going to revert so that the tree isn't broken, and probably leave it reverted until we have test coverage of what we're actually doing. if someone is around and has time to fix things, that's fine with me. as long as the next thing we do is merge the missing tests. | 00:32 |
| corvus | tristanC: well, it's a messy git history for a very good reason: we don't test what we deploy. | 00:32 |
| mnaser | mordred: side note but to make the dashboard functioning at large load im thinking we should move the filtering to server-side | 00:33 |
| mordred | mnaser: probably not a terrible idea | 00:34 |
| corvus | mnaser: what filtering? | 00:34 |
| mnaser | corvus: dashboard for zuul | 00:35 |
| mordred | corvus, mnaser: like flitering builds by job name I think, right? | 00:35 |
| corvus | mnaser: what filtering in the dashboard for zuul? | 00:35 |
| mnaser | the filters on the status page are client side node | 00:35 |
| mnaser | client side page** | 00:35 |
| mnaser | ok, sorry, i am a bit lacking some sleep | 00:35 |
| mnaser | the textbox next to "filters" on the zuul status page | 00:35 |
| mordred | ah that | 00:36 |
| mnaser | it stills pulls /api/zuul which gets all jobs then filters it locally | 00:36 |
| mnaser | when our queues grow it becomes really hard to use | 00:36 |
| corvus | mordred, tristanC: anyway, i have to run now and won't be back until my morning. mordred, if you want to go with the roll-forward route, feel free to -2 the revert. but i took responsibility for the previous roll-forward and have discharged that responsibility by reverting and rolling back, so... it's your show if you do that :) | 00:36 |
| mnaser | instead maybe we should have something like /api/zuul?filter=foo and get rid of the client side filtering (by client-side as in the browser) | 00:36 |
| mnaser | i'll probably have to do a bit more research on how to make that works | 00:37 |
| mordred | corvus: I would be comfortable with roll forward - but I won't be in ap ositoin to take responsitibility for it for another 2 days or so - so I think today I'm in favor of the revert - but once i'm back online for real, I can heplp drive the roll forward | 00:37 |
| mordred | mnaser: for status page filtering the status server side would be more work I think - but I thinkn tristanC's react rewrite of status helps client-side filtering of status pretty nicely | 00:38 |
| mordred | mnaser: for builds and jobs and stuff like that - I think server side filtering is a great idea, but I think we're already doing a decent amount of it? | 00:38 |
| mnaser | mordred: i've noticed with more terrible internet (hotel wifis, inflight), it really struggles. for example -- curl -vvv zuul.openstack.org/api/status 0.02s user 0.03s system 0% cpu 34.583 total | 00:39 |
| mordred | mnaser: in any case- I agree with and support makingthings betterer as we improve the client/server dance | 00:39 |
| mnaser | thats from my hotel in stockholm right now so it isn't exactly the *worst* | 00:39 |
| clarkb | do peoppe filter a lot? | 00:39 |
| mordred | mnaser: I think we have per-project status api endpoints already - perhaps we just need to use it | 00:40 |
| clarkb | I wonder if that is why no one notices that non og our software works right now | 00:40 |
| mnaser | i have an almost permanent openstack/openstack-ansible filter to pick up all of our roles, etc | 00:40 |
| clarkb | they are tunnel visioned | 00:40 |
| mnaser | i can take blame party for making everything else disappear, heh | 00:40 |
| mordred | clarkb: I filter status almost 100% of the time | 00:43 |
| mnaser | another ui thing i've noticed is it'd be nice if zuul console started from last N lines instead of from the start of a job | 00:46 |
| mordred | mnaser: in terms of loading data? or in scroll position? | 00:46 |
| mnaser | mordred: loading data, for example if i just want to see where the job is so far | 00:47 |
| mnaser | ill click to open the job and then my browser is destroyed loading up everything | 00:47 |
| tristanC | mnaser: i actually started to do that in the rewrite, but to properly implement this we need to update the websocket protocol to support requesting start line | 00:49 |
| mordred | mnaser: that may be harder to implement due to how the stream is implemented on the backend ... but - maybe not, maybe the websocket protocol could be extended to support a "scan til the end" and it could do that seek after opening the file to stream without to much server-side cost | 00:49 |
| mordred | mnaser: that's me just talking out loud - but it sounds like tristanC had a similar thought already :) | 00:49 |
| clarkb | mordred: I never filter and eonder why no one fixes their broken tests :) | 00:49 |
| tristanC | mordred: yeah, shouldn't be that hard, but that would e a follow-up | 00:49 |
| mnaser | yeah i have always thought that it is quite a difficult initial thing to address | 00:49 |
| mordred | clarkb: well - in my case, my tests are rarely roken by other project's tests - and when thye are I do try to go fix the other projects | 00:50 |
| mordred | clarkb: but I hear you | 00:50 |
| tristanC | mordred: once 608794 land, could we zuul enqueue-ref --tenant openstack --trigger gerrit --pipeline post --project openstack-infra/zuul --ref master | 00:52 |
| mordred | tristanC: someone in infra might be able to - my in-flight wifi is too spotty for me to execute tasks like that | 00:55 |
| mordred | BUT - I'm happy to report that klm has in-flight wifi now | 00:55 |
| tristanC | then if the publication worked, we should also land https://review.openstack.org/606601 | 00:55 |
| *** rlandy|bbl is now known as rlandy | 00:59 | |
| *** tristanC has quit IRC | 01:01 | |
| *** tristanC has joined #zuul | 01:01 | |
| openstackgerrit | Merged openstack-infra/zuul-jobs master: fetch-javascript-content-tarball: fix tarball relative path https://review.openstack.org/608794 | 01:41 |
| openstackgerrit | Monty Taylor proposed openstack-infra/zuul master: WIP: docker-compose quickstart example https://review.openstack.org/608344 | 02:12 |
| *** rlandy has quit IRC | 02:22 | |
| openstackgerrit | Merged openstack-infra/zuul master: Doc: executor operations document pause, remove graceful https://review.openstack.org/602455 | 02:36 |
| openstackgerrit | Merged openstack-infra/zuul master: Doc: executor operations - explain jobs will be restarted at restart https://review.openstack.org/603136 | 02:36 |
| dmsimard | Just put up a blog post about AnsibleFest last week, Zuul is a chunk of it: https://dmsimard.com/2018/10/08/ansiblefest-2018-community-project-highlights/ | 03:49 |
| *** nilashishc has joined #zuul | 04:36 | |
| *** pcaruana has joined #zuul | 04:38 | |
| tristanC | mordred: "I just tried overtyping a project filter on status - and because a refresh happened I lost my typing" < i can't reproduce that, the filter input box doesn't clear on refresh | 05:06 |
| openstackgerrit | Tristan Cacqueray proposed openstack-infra/zuul master: web: add build page https://review.openstack.org/597024 | 05:07 |
| openstackgerrit | Tristan Cacqueray proposed openstack-infra/zuul master: web: fix rewrite rule installation doc https://review.openstack.org/608850 | 05:44 |
| *** nilashishc has quit IRC | 06:02 | |
| *** nilashishc has joined #zuul | 06:04 | |
| *** nilashishc has quit IRC | 06:06 | |
| *** nilashishc has joined #zuul | 06:06 | |
| tobiash | tristanC: maybe I just need to accustom to it but I find the darker blue of the progress bars better | 06:08 |
| tobiash | but yay, the new ui is cool :) | 06:10 |
| tristanC | tobiash: it seems like the default for progress-bar-success, we could override the css if you prefer | 06:10 |
| tristanC | tobiash: glad you like it :) | 06:10 |
| tristanC | tobiash: i agree that shade of blue doesn't contrast well with the green | 06:11 |
| tobiash | and I think slightly rounded progress bars may make it look a little bit more elegant | 06:12 |
| tristanC | also, now that it is full screen, we could probably have 4 columns | 06:12 |
| *** quiquell has joined #zuul | 07:19 | |
| quiquell | Good morning | 07:19 |
| quiquell | Hava a question regarding 'vars:' section at jobs | 07:20 |
| quiquell | It's possible to dump them to a file without selecting them one by one or with a regex ? | 07:20 |
| quiquell | I mean something like 'zuul.job.vars' to get the list of all of them | 07:21 |
| *** electrofelix has joined #zuul | 07:25 | |
| openstackgerrit | Simon Westphahl proposed openstack-infra/zuul master: Use merger to get list of files for pull-request https://review.openstack.org/603287 | 08:10 |
| *** sshnaidm|afk is now known as sshnaidm | 08:29 | |
| tobiash | quiquell: currently this is not possible/implemented | 08:30 |
| quiquell | tobiash: ack | 08:32 |
| *** panda has quit IRC | 09:22 | |
| *** panda has joined #zuul | 09:24 | |
| openstackgerrit | Tobias Henkel proposed openstack-infra/nodepool master: Reduce socket connect timeout in nodescan https://review.openstack.org/608897 | 09:28 |
| *** nilashishc has quit IRC | 09:34 | |
| mnaser | i think there is a bug in nodepool when a really small pool exists | 09:47 |
| tobiash | mnaser: what kind of bug? | 09:49 |
| mnaser | tobiash: so i think infra-root will need to supply logs but you can see https://review.openstack.org/#/c/608315/ there is a NODE_FAILURE a few times | 09:49 |
| mnaser | but it could be cloud related but i'm not sure... i dunno why, a recheck works after | 09:50 |
| tobiash | Most 'small pool' bugs should be shaked out until now (mostly quota related) | 09:53 |
| tobiash | But yes, logs should give us more insight | 09:54 |
| *** quiquell is now known as ssquique | 10:16 | |
| *** ssquique is now known as quiquell | 10:22 | |
| *** threestrands has joined #zuul | 11:01 | |
| *** nilashishc has joined #zuul | 11:27 | |
| *** dcastellani has joined #zuul | 12:00 | |
| dcastellani | Good morning | 12:00 |
| *** jesusaur has quit IRC | 12:01 | |
| pabelanger | mnaser: tobiash: we've been running small pools in ansible-network, so far it is working pretty well | 12:03 |
| *** jesusaur has joined #zuul | 12:07 | |
| *** chandankumar has joined #zuul | 12:19 | |
| pabelanger | tristanC: mordred: I like the new UI, but noticed the text in the 'success' green box isn't quite centered top to bottom | 12:25 |
| pabelanger | everything else looks great | 12:25 |
| *** rlandy has joined #zuul | 12:28 | |
| dmsimard | Everyone knows that centering things is like the hardest thing to do in webdev :p | 12:38 |
| dmsimard | http://howtocenterincss.com | 12:39 |
| *** panda has quit IRC | 13:13 | |
| *** panda has joined #zuul | 13:14 | |
| *** quiquell is now known as quiquell|lunch | 13:16 | |
| tristanC | pabelanger: would you mind opening a story with suggestion/improvement? | 13:17 |
| *** rlandy_ has joined #zuul | 13:26 | |
| *** rlandy has quit IRC | 13:26 | |
| *** quiquell|lunch is now known as quiquell | 13:36 | |
| tristanC | https://review.openstack.org/593605 fixes multi filter builds query | 13:40 |
| *** rlandy_ is now known as rlandy | 13:44 | |
| fungi | so many sharp edges in the new dashboard... i feel like i have to be careful not to cut myself | 14:09 |
| dmsimard | We can maybe add radius to round the corners ? | 14:12 |
| dmsimard | ¯\_(ツ)_/¯ | 14:12 |
| dmsimard | Rounded corners! https://i.imgur.com/s9sCVBp.png (I'm not sure which css element is that dark square line) | 14:15 |
| *** mugsie has joined #zuul | 14:15 | |
| fungi | is broken mobile menu drop-down in the new dashboard a known issue? | 14:28 |
| corvus | indeed, if i shrink my window down to really small, clicking the hamburger doesn't do anything | 14:30 |
| fungi | clarkb, AJaeger and i confirmed it over in #openstack-infra a few minutes ago too | 14:31 |
| *** nilashishc has quit IRC | 14:37 | |
| *** quiquell is now known as quiquell|off | 14:39 | |
| dmsimard | Should we round up the different UI bugs/improvement opportunities in an etherpad or something ? | 14:47 |
| pabelanger | I asked yesterday, but didn't get much feedback. Are zuul people saying at a specific hotel in Berlin? Or randomized? | 14:59 |
| *** j^2 has quit IRC | 15:05 | |
| fungi | dmsimard: adding stories for them and tracking those in a worklist/board would make sense | 15:22 |
| dmsimard | sure, I haven't been using storyboard a whole lot recently -- if it's actually used for zuul then yes it's the right place :p | 15:23 |
| fungi | pabelanger: i'm in the swissôtel berlin, augsburger str. 44 (the osf books all the staff into the same hotel in a batch) | 15:23 |
| fungi | dmsimard: yep, https://storyboard.openstack.org/#!/project_group/zuul | 15:24 |
| fungi | a new worklist for tracking the issues with the react-based dashboard might make sense | 15:27 |
| fungi | or just a tag (or could populate an automatic worklist from that tag) | 15:28 |
| *** ianychoi has quit IRC | 15:39 | |
| dmsimard | There's people doing a webinar about Zuul, how cool is that: http://zuul-webinar.codilime.com/info/ | 15:46 |
| dmsimard | from https://www.crowdcast.io/e/zuul/1 | 15:46 |
| pabelanger | neat, watching now | 15:52 |
| pabelanger | ha, also using latest UI for web | 15:52 |
| Diabelko | yes | 15:53 |
| Diabelko | we added it like 40 mins before webinar | 15:54 |
| Diabelko | but shush | 15:54 |
| dmsimard | haha | 15:54 |
| pabelanger | that is pretty awesome | 15:54 |
| dmsimard | Diabelko: great work! | 15:55 |
| Diabelko | thank you | 15:55 |
| corvus | neat! will there be a recording of that we could post to the website? | 16:01 |
| openstackgerrit | Merged openstack-infra/zuul master: Speed up build list query under mysql https://review.openstack.org/605170 | 16:04 |
| dmsimard | I heard something about a recording being eventually available | 16:05 |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul master: WIP: docker-compose quickstart example https://review.openstack.org/608344 | 16:17 |
| corvus | that goes all the way through to successfully running a job and publishing logs! | 16:18 |
| Shrews | corvus: cool. i'll give it a walkthrough now | 16:19 |
| corvus | Shrews: thanks! i need to flesh out the prose in the docs a bit, but all the steps are there in shorthand at the end, so it should be functional | 16:19 |
| *** threestrands has quit IRC | 16:21 | |
| pabelanger | that is neat, great work | 16:33 |
| *** panda has quit IRC | 16:34 | |
| *** panda has joined #zuul | 16:36 | |
| Diabelko | corvus: yes, there'll be a recording of that | 16:38 |
| Diabelko | I'll be happy to share that with you guys later on | 16:38 |
| Diabelko | I just need to get my hands on it first | 16:38 |
| Diabelko | crowdcast is a magical platform ;) | 16:38 |
| corvus | Diabelko: great, good luck and thanks! | 16:45 |
| openstackgerrit | Merged openstack-infra/zuul-jobs master: Retry failed git pushses on workspace setup https://review.openstack.org/608303 | 16:45 |
| Diabelko | speaking of good - is there any zuul-beer planned in Berlin? ;) | 16:47 |
| corvus | i don't think so yet, but that sounds like a great idea | 16:52 |
| Shrews | corvus: hit a gerrit snag. comments in the review | 16:54 |
| *** bhavikdbavishi has joined #zuul | 16:57 | |
| corvus | Shrews: can you run 'git-review -v' and paste the output? | 17:02 |
| *** bhavikdbavishi1 has joined #zuul | 17:02 | |
| Shrews | corvus: oh, doh. had a different host key in known_hosts from when i set it up locally before | 17:03 |
| Shrews | removed that and it works | 17:03 |
| *** bhavikdbavishi has quit IRC | 17:04 | |
| *** bhavikdbavishi1 is now known as bhavikdbavishi | 17:04 | |
| corvus | Shrews: ah, that'll do it! hopefully that won't bite many users... but that does seem like something git-review could do better... | 17:04 |
| corvus | fungi: ^ git-review hides ssh host key errors from users and just says "We don't know where your gerrit is" | 17:05 |
| corvus | i'm going through and taking a bunch of gerrit screenshots to put in there (for the "click become; click your user; etc" instructions) | 17:07 |
| clarkb | corvus: fungi I want to say there is a patch up to fix that from stevenk | 17:07 |
| corvus | Shrews: i don't understand why 'git add zuul.d' wouldn't have worked; were you in the root of the repo? | 17:09 |
| corvus | Shrews: oh, maybe you were in 'zuul.d'? | 17:09 |
| Shrews | corvus: i had the yaml files up one level | 17:15 |
| Shrews | doh | 17:15 |
| corvus | ah good sanity check then :) | 17:16 |
| Shrews | i'm trying to do all of the silly things new users would do!! | 17:16 |
| Shrews | :) | 17:16 |
| Shrews | so now we're building a support-experience base | 17:17 |
| corvus | ++ | 17:17 |
| corvus | this exercise has me wondering if we should make the gerrit event delay configurable. on a sandbox system, there should be no need to wait 10 seconds to process a gerrit event | 17:18 |
| tobiash | ++ | 17:19 |
| tobiash | I thought I had something like that years ago | 17:19 |
| tobiash | https://review.openstack.org/#/c/343562 | 17:26 |
| Shrews | corvus: ok, got through all of it | 17:30 |
| Shrews | and able to see test1 logs | 17:30 |
| Shrews | ooh, i should probably lunch | 17:32 |
| corvus | tobiash: ayep that'd do it :) | 17:34 |
| tobiash | That was for v2 | 17:34 |
| corvus | yes, will need some porting | 17:46 |
| *** electrofelix has quit IRC | 17:50 | |
| fungi | yeah, there were some regressions with error handling a couple git-review releases back. hopefully the fix clarkb mentions actually exists | 18:01 |
| pabelanger | could I get a review on https://review.openstack.org/608472/ we are using it in ansible-network today, but want to upstream | 18:06 |
| pabelanger | do we have a shortlist of things we want to include for 3.3.0? mostly looking to start using the per project SSH keypairs | 18:17 |
| tobiash | is there a release planned this week? | 18:26 |
| tobiash | ah, misunderstood, you want to have a new release? | 18:27 |
| tobiash | I think we should at least wait a few days so we can see that the new web ui behaves correctly (and maybe also round some corners) | 18:29 |
| *** bhavikdbavishi has quit IRC | 18:38 | |
| *** caphrim007 has joined #zuul | 18:51 | |
| *** gouthamr has quit IRC | 19:58 | |
| *** dmellado has quit IRC | 19:58 | |
| *** pcaruana has quit IRC | 20:37 | |
| *** gouthamr has joined #zuul | 20:50 | |
| *** dmellado has joined #zuul | 20:53 | |
| fungi | njohnston has reported in #openstack-infra that the new status dashboard doesn't properly filter when you enter multiple change ids | 21:08 |
| fungi | known issue? | 21:08 |
| fungi | he's going to add a story | 21:10 |
| corvus | did that ever work? | 21:11 |
| corvus | i guess so (that's the report) | 21:11 |
| corvus | there is a possibility none of the zuul authors knew that was a feature :) | 21:12 |
| dmsimard | I didn't know about that either heh | 21:15 |
| *** j^2 has joined #zuul | 21:21 | |
| anticw | fungi: zuul.openstack.org dashboard? i've used multiple ids on that recently (comma sep) | 21:29 |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul master: Add docker-compose quickstart example https://review.openstack.org/608344 | 21:36 |
| fungi | anticw: yep | 21:36 |
| corvus | j^2, tobiash, Shrews, mordred, tristanC: ^ i believe 608344 is complete. please take a look at it, run through it, and leave comments. i'm going to start working on a followup patch which reorganizes the admin manual a bit, now that we have four (4!) fairly redundant installation sections. | 21:38 |
| j^2 | corvus: awesome thanks for this, i can't wait to give it a shot. | 21:50 |
| clarkb | corvus: it is on my todo list to run it, but I've been at a conference with no electrical outlets the last few days and you taught me in austin not to run containers in that situation :P | 21:51 |
| clarkb | I'll see if I can give it a go tomorow | 21:51 |
| corvus | clarkb: haha, yeah i have not been running on battery power :) | 21:52 |
| *** samccann has quit IRC | 22:03 | |
| *** samccann has joined #zuul | 22:03 | |
| fungi | it's funny that conferences don't realize people carry devices which use electricity | 22:16 |
| *** goern has quit IRC | 22:20 | |
| *** panda has quit IRC | 22:37 | |
| *** panda has joined #zuul | 22:39 | |
| *** caphrim007 has quit IRC | 23:15 | |
| *** sshnaidm is now known as sshnaidm|afk | 23:32 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!
