| Shrews | pabelanger: clarkb: i think i have a programmatic solution to the wedged provider issue you found tonight (it was bugging me we don't handle that better) | 00:02 |
|---|---|---|
| Shrews | will throw something up tomorrow | 00:02 |
| clarkb | Shrews: ok | 00:02 |
| pabelanger | great | 00:03 |
| clarkb | I'm assuming something that avoids blocking there and instead declines requests somehow would work | 00:03 |
| Shrews | clarkb: yep. if request would put us at our max-servers quota, and all of our current nodes are ready&unlocked (meaning not used) and none of them have the needed label, decline it | 00:04 |
| Shrews | i don't think we ever considered the case of us being at quote with all unused nodes (caused here by a zuul restart) | 00:05 |
| Shrews | s/quote/quota/ | 00:06 |
| openstackgerrit | liusheng proposed openstack-infra/zuul master: Fix AttributeError when handle periodic job with github driver https://review.openstack.org/536645 | 00:52 |
| *** myoung|bbl is now known as myoung | 01:12 | |
| *** gigaset has joined #zuul | 01:12 | |
| gigaset | 2!list | 01:13 |
| *** gigaset has left #zuul | 01:16 | |
| *** harlowja has quit IRC | 02:10 | |
| *** Wei_Liu has joined #zuul | 02:30 | |
| *** rlandy|bbl is now known as rlandy | 02:31 | |
| *** rlandy has quit IRC | 03:42 | |
| *** elyezer has quit IRC | 03:48 | |
| *** elyezer has joined #zuul | 03:49 | |
| openstackgerrit | Tristan Cacqueray proposed openstack-infra/nodepool master: Implement a static driver for Nodepool https://review.openstack.org/535553 | 04:08 |
| openstackgerrit | Tristan Cacqueray proposed openstack-infra/nodepool master: nodeutils: use socket.getaddrinfo instead of ipaddress https://review.openstack.org/539086 | 04:08 |
| openstackgerrit | Tristan Cacqueray proposed openstack-infra/nodepool master: Implement a static driver for Nodepool https://review.openstack.org/535553 | 04:16 |
| *** harlowja has joined #zuul | 04:21 | |
| *** harlowja has quit IRC | 04:26 | |
| *** harlowja has joined #zuul | 04:26 | |
| *** harlowja has quit IRC | 04:27 | |
| *** sshnaidm|afk has quit IRC | 05:14 | |
| openstackgerrit | Tristan Cacqueray proposed openstack-infra/nodepool master: Implement a static driver for Nodepool https://review.openstack.org/535553 | 05:27 |
| *** sshnaidm has joined #zuul | 05:31 | |
| *** sshnaidm has quit IRC | 05:43 | |
| *** bhavik1 has joined #zuul | 05:45 | |
| *** bhavik1 has quit IRC | 05:53 | |
| *** jappleii__ has quit IRC | 06:01 | |
| *** sshnaidm has joined #zuul | 06:02 | |
| *** bhavik1 has joined #zuul | 06:14 | |
| *** bhavik1 has quit IRC | 06:36 | |
| tobiash | pabelanger: hrm, it's not as easy as I thought, modifying the test is not sufficient | 07:56 |
| * tobiash needs to do some whiteboarding to really understand nodepool request handling | 07:56 | |
| *** sshnaidm is now known as sshnaidm|afk | 08:15 | |
| *** lennyb has quit IRC | 08:50 | |
| *** jpena|off is now known as jpena | 08:50 | |
| *** sshnaidm|afk is now known as sshnaidm | 09:04 | |
| *** lennyb has joined #zuul | 09:34 | |
| openstackgerrit | Matthieu Huin proposed openstack-infra/zuul master: zuul autohold: allow filtering per commit https://review.openstack.org/536993 | 09:54 |
| *** elyezer has quit IRC | 11:09 | |
| openstackgerrit | Matthieu Huin proposed openstack-infra/zuul master: zuul autohold: allow filtering per commit https://review.openstack.org/536993 | 11:11 |
| *** elyezer has joined #zuul | 11:11 | |
| openstackgerrit | Matthieu Huin proposed openstack-infra/zuul master: [WIP] zuul web: add admin endpoint, enqueue commands https://review.openstack.org/539004 | 11:15 |
| *** sshnaidm is now known as sshnaidm|afk | 12:39 | |
| *** jpena is now known as jpena|lunch | 12:45 | |
| openstackgerrit | Tobias Henkel proposed openstack-infra/nodepool master: Fix relaunch attempts when hitting quota errors https://review.openstack.org/536930 | 13:15 |
| tobiash | pabelanger, Shrews: sooo, this took me half a day but I got the test working... ^^ | 13:15 |
| tobiash | still needs some cleanup but have to think about different things now | 13:16 |
| *** sshnaidm|afk is now known as sshnaidm | 13:19 | |
| *** rlandy has joined #zuul | 13:28 | |
| *** weshay|ruck|afk is now known as weshay|ruck | 13:31 | |
| *** jpena|lunch is now known as jpena | 13:46 | |
| Shrews | clarkb: pabelanger: i found a problem with my proposed solution last night. we're going to have to go with max-ready-age for now until we can come up with something better | 14:22 |
| openstackgerrit | Tristan Cacqueray proposed openstack-infra/zuul-jobs master: Add buildset-artifacts-location and fetch roles https://review.openstack.org/530679 | 14:31 |
| tobiash | hrm, locally that test worked :/ | 14:31 |
| *** elyezer has quit IRC | 14:36 | |
| *** elyezer has joined #zuul | 14:37 | |
| openstackgerrit | Fabien Boucher proposed openstack-infra/zuul-jobs master: Propose to move submit-log-processor-jobs and submit-logstash-jobs in zuul-jobs https://review.openstack.org/537847 | 14:40 |
| openstackgerrit | Tobias Henkel proposed openstack-infra/zuul master: Set remote url on every getRepo in merger https://review.openstack.org/535716 | 15:15 |
| *** myoung is now known as myoung|food | 17:44 | |
| *** electrofelix has quit IRC | 17:49 | |
| *** electrofelix has joined #zuul | 17:50 | |
| *** bhavik1 has joined #zuul | 18:02 | |
| *** bhavik1 has quit IRC | 18:10 | |
| corvus | mordred: good news and bad news: the change to use override-branch to solve the ansible/devel problem should be in production. bad news is that merge failure reports are pipeline-global, so any merge hiccup on ansible will report. due to that, i removed the ansible/ansible project definition until we work it out. | 18:11 |
| mordred | corvus: yah - I saw the removal of ansible/ansible ... and the global merge-failure reports is ... fun isn't it? | 18:11 |
| corvus | mordred: the minimal solution is, i think, to make the merging more robust so that github hiccups don't cause errors. we need to do that regardless... | 18:12 |
| mordred | yup. must do that | 18:12 |
| corvus | mordred: even if we do that, we could have systemic errors though -- like mergers run out of space, or whatever -- main point is -- *something* in the future will cause a spurious merge error, and we'll report it. | 18:12 |
| mordred | or, even a real merge error | 18:13 |
| corvus | mordred: heh, yeah, that could happen too :) | 18:13 |
| corvus | mordred: so i think we have to ask -- if we're *generally* good citizens, but occasionally that happens, is that okay? or do we need to find a way to never report merge failures for specific projects? | 18:13 |
| mordred | corvus: I think the failures exposed an edge-case where the current behavior does not match what people might expect ... which is that a zuul, such as ours, that only has jobs configured to report on a subset of a given repo wouldn't be expected to report merge errors to changes that didn't touch those files | 18:13 |
| corvus | mordred: indeed -- though, in the case of an actual merge failure, how do you know? :) | 18:14 |
| mordred | corvus: I think we do - I'm thinking about openstack third-party CI's too | 18:14 |
| mordred | corvus: that's an excellent questoin! :) | 18:14 |
| mordred | corvus: (sort of thinking of 20 third-party CIs all reporting "I couldn't merge this change" to nova patches) | 18:15 |
| corvus | mordred: yeah, though they generally turn of merge reports for their pipelines | 18:16 |
| mordred | corvus: good point | 18:16 |
| corvus | but we're putting ansible into our existing check pipeline | 18:17 |
| corvus | we could create a new pipeline just for third-party check | 18:17 |
| mordred | corvus: good point. we could also add ansible/ansible in a second tenant with its own pipelines ... although that seems a bit more extreme | 18:18 |
| corvus | i think that would let us proceed with no further zuul changes. or we could look into disabling merge failure reports per-project. i think that's something we may want to do, but that also opens the can-o-worms that SpamapS wanted to open with the project-customized slack reporter, and i kinda wanted to keep that closed until after the release | 18:18 |
| mordred | corvus: yah. project-specific reporter config is a thing that could be desirable, but seems like a thing to talk about in dublin | 18:19 |
| corvus | so maybe we should make a third-party-check pipeline in openstack zuul, add ansible/ansible to it. i think we can still do all the cross-testing we need. gating is, obviously, not a problem right now. | 18:19 |
| corvus | then maybe collapse that back into check later | 18:19 |
| mordred | corvus: yah. I thnk that's a good idea | 18:21 |
| mordred | corvus: that would also let lower its priority as well | 18:21 |
| mordred | s/let/let us/ | 18:21 |
| mordred | corvus: I can make that patch real quick | 18:22 |
| corvus | mordred: kk thx | 18:23 |
| corvus | i'm going to take a little break; i have a bit more energy today, but not quite enough. | 18:24 |
| mordred | corvus: kk. take a break - no need to push yourself and get sicker | 18:24 |
| openstackgerrit | Matthieu Huin proposed openstack-infra/zuul master: zuul autohold: allow filtering per commit https://review.openstack.org/536993 | 18:27 |
| *** myoung|food is now known as myoung | 18:34 | |
| openstackgerrit | Matthieu Huin proposed openstack-infra/zuul master: zuul autohold: allow filtering per commit https://review.openstack.org/536993 | 18:59 |
| *** jpena is now known as jpena|off | 19:00 | |
| mordred | corvus: https://review.openstack.org/#/c/539286/ how's that look? | 19:11 |
| *** harlowja has joined #zuul | 19:17 | |
| tobiash | mordred: wow, I didn't know yet that merge failure can be its own reporting category | 19:19 |
| tobiash | one never stops learning... | 19:23 |
| mrhillsman | mordred corvus have either of you confirmed your cicd f2f attendance yet? | 19:37 |
| mrhillsman | asking because there is an overview on day one and need someone to do zuul | 19:37 |
| mrhillsman | or clarkb | 19:37 |
| corvus | mrhillsman: what event is this? | 19:37 |
| mrhillsman | the one fatih reached out about | 19:37 |
| corvus | mrhillsman: i think i put my name in the etherpad, is there something more formal than that? | 19:38 |
| mrhillsman | cross community f2f cicd | 19:38 |
| clarkb | mrhillsman: last I heard it still wans't confirmed as happening so I haven't bought tickets or ldoging yet but do intend on attending | 19:38 |
| mrhillsman | i think fatih sent to confirmation, let me check to make sure | 19:38 |
| mrhillsman | corvus we are doing the agenda and just wanted to be sure a zuul expert would be available for the overview | 19:39 |
| mordred | mrhillsman: I'll be there for sure | 19:39 |
| clarkb | mrhillsman: I intend on being there | 19:39 |
| mrhillsman | ok cool, he sent an email on the 23rd but i see how it is not clear | 19:40 |
| mrhillsman | he basically just said we have got a room | 19:40 |
| mrhillsman | thx gents, will confirm we have someone to do zuul overview :) | 19:41 |
| mordred | mrhillsman: awesome! thanks | 19:43 |
| mrhillsman | welcome | 20:02 |
| *** sshnaidm is now known as sshnaidm|afk | 20:38 | |
| openstackgerrit | David Shrewsbury proposed openstack-infra/nodepool master: WIP: Provider wedge test https://review.openstack.org/539316 | 20:40 |
| Shrews | ^^ tests the wedge scenario, but still no fix so it should rightly fail | 20:40 |
| Shrews | simplest test case I could conjure up | 20:41 |
| tobiash | corvus, mordred: are we ready to land 535518? | 21:08 |
| tobiash | looks like this might have been forgotten some time ago | 21:09 |
| mordred | tobiash: yah. I thnk so | 21:17 |
| tobiash | ok | 21:18 |
| * tobiash ends his review evening | 21:34 | |
| *** elyezer has quit IRC | 21:34 | |
| openstackgerrit | Merged openstack-infra/zuul master: Remove zuul._projects https://review.openstack.org/535518 | 21:37 |
| *** threestrands has joined #zuul | 21:40 | |
| *** elyezer has joined #zuul | 21:44 | |
| openstackgerrit | Merged openstack-infra/nodepool master: zk: check for client in properties https://review.openstack.org/535559 | 21:58 |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul master: Merger: retry network operations https://review.openstack.org/539356 | 23:11 |
| corvus | mordred: ^ that may hopefully smooth over github network hiccups | 23:11 |
| mordred | corvus: cool! | 23:14 |
| SpamapS | corvus: we should look at bringing in tenacity at some point | 23:37 |
| corvus | SpamapS: heh, yes. i almost wrote that myself, but stopped. :) | 23:41 |
| SpamapS | so much hair on that yak | 23:41 |
| *** elyezer has quit IRC | 23:43 | |
| mordred | SpamapS: don't know if you saw - but we're running shade tests on ansible PRs as of now! | 23:44 |
| *** elyezer has joined #zuul | 23:44 | |
| SpamapS | mordred: ssswwwweeeeeetttt | 23:48 |
| SpamapS | I had a nice internal demo of our zuul today btw. Folks interested in the scalability. :) | 23:48 |
| mordred | SpamapS: yay! | 23:49 |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!