| *** hamalq_ has quit IRC | 00:05 | |
| *** saneax has quit IRC | 00:08 | |
| *** saneax has joined #zuul | 00:38 | |
| *** rlandy|ruck|afk is now known as rlandy|ruck | 01:13 | |
| *** rlandy|ruck has quit IRC | 01:46 | |
| *** sgw1 has quit IRC | 02:28 | |
| *** sgw1 has joined #zuul | 02:29 | |
| *** sgw1 has quit IRC | 02:33 | |
| *** Goneri has quit IRC | 02:41 | |
| *** rfolco has quit IRC | 02:41 | |
| *** sgw1 has joined #zuul | 02:59 | |
| *** bhavikdbavishi has joined #zuul | 03:08 | |
| *** bhavikdbavishi has quit IRC | 03:11 | |
| *** bhavikdbavishi has joined #zuul | 03:12 | |
| *** bhavikdbavishi has quit IRC | 03:17 | |
| *** saneax has quit IRC | 03:18 | |
| *** bhavikdbavishi has joined #zuul | 03:22 | |
| *** bhavikdbavishi has quit IRC | 03:29 | |
| *** bhavikdbavishi has joined #zuul | 03:30 | |
| *** bhavikdbavishi has quit IRC | 04:01 | |
| *** vishalmanchanda has joined #zuul | 04:01 | |
| *** bhavikdbavishi has joined #zuul | 04:06 | |
| *** harrymichal has joined #zuul | 04:22 | |
| *** sgw1 has quit IRC | 04:34 | |
| *** wuchunyang has joined #zuul | 04:43 | |
| *** wuchunyang has quit IRC | 04:52 | |
| *** wuchunyang has joined #zuul | 04:58 | |
| *** wuchunyang has quit IRC | 05:02 | |
| *** wuchunyang has joined #zuul | 05:02 | |
| *** wuchunyang has quit IRC | 05:30 | |
| *** wuchunyang has joined #zuul | 05:52 | |
| *** bhavikdbavishi1 has joined #zuul | 06:04 | |
| *** bhavikdbavishi has quit IRC | 06:05 | |
| *** bhavikdbavishi1 is now known as bhavikdbavishi | 06:05 | |
| *** bhavikdbavishi has quit IRC | 06:19 | |
| *** bhavikdbavishi has joined #zuul | 06:20 | |
| *** wuchunyang has quit IRC | 06:24 | |
| *** harrymichal has quit IRC | 06:31 | |
| *** marios has joined #zuul | 06:47 | |
| *** bhavikdbavishi has quit IRC | 06:59 | |
| openstackgerrit | Andrii Ostapenko proposed zuul/zuul-jobs master: Fix issues with buildset registry https://review.opendev.org/741584 | 07:00 |
|---|---|---|
| *** jcapitao has joined #zuul | 07:13 | |
| *** bhavikdbavishi has joined #zuul | 07:31 | |
| openstackgerrit | Benoît Bayszczak proposed zuul/nodepool master: [provider][aws] now performs a multi-describe at once https://review.opendev.org/740328 | 07:32 |
| *** tosky has joined #zuul | 07:33 | |
| openstackgerrit | Benoît Bayszczak proposed zuul/nodepool master: [provider][aws] now performs a multi-describe at once https://review.opendev.org/740328 | 07:34 |
| *** bhavikdbavishi has quit IRC | 07:35 | |
| *** bhavikdbavishi has joined #zuul | 07:45 | |
| *** harrymichal has joined #zuul | 07:53 | |
| *** nils has joined #zuul | 07:54 | |
| *** avass has quit IRC | 08:17 | |
| openstackgerrit | Benoît Bayszczak proposed zuul/nodepool master: [provider][aws] now performs a multi-describe at once https://review.opendev.org/740328 | 08:30 |
| *** harrymichal has quit IRC | 08:33 | |
| *** harrymichal has joined #zuul | 08:33 | |
| openstackgerrit | Benoît Bayszczak proposed zuul/nodepool master: [provider][aws] now performs a multi-describe at once https://review.opendev.org/740328 | 08:34 |
| openstackgerrit | Benoît Bayszczak proposed zuul/nodepool master: [WIP][provider][aws] now performs a multi-describe at once https://review.opendev.org/740328 | 08:35 |
| *** reiterative has joined #zuul | 08:47 | |
| openstackgerrit | Benoît Bayszczak proposed zuul/nodepool master: [WIP][provider][aws] now performs a multi-describe at once https://review.opendev.org/740328 | 08:57 |
| *** bhavikdbavishi has quit IRC | 09:05 | |
| openstackgerrit | Benoît Bayszczak proposed zuul/nodepool master: [WIP][provider][aws] now performs a multi-describe at once https://review.opendev.org/740328 | 09:11 |
| *** bhavikdbavishi has joined #zuul | 09:11 | |
| *** bhavikdbavishi1 has joined #zuul | 09:20 | |
| *** harrymichal has quit IRC | 09:20 | |
| *** bhavikdbavishi has quit IRC | 09:21 | |
| *** bhavikdbavishi1 is now known as bhavikdbavishi | 09:21 | |
| *** marios has quit IRC | 09:38 | |
| openstackgerrit | Benoît Bayszczak proposed zuul/nodepool master: [provider][aws] now performs a multi-describe at once https://review.opendev.org/740328 | 09:41 |
| *** sshnaidm|afk is now known as sshnaidm|off | 10:16 | |
| openstackgerrit | Tobias Henkel proposed zuul/zuul master: Add spec for enhanced regional executor distribution https://review.opendev.org/663413 | 10:53 |
| *** jcapitao is now known as jcapitao_lunch | 10:54 | |
| openstackgerrit | Tobias Henkel proposed zuul/zuul master: Route streams to different zones via finger gateway https://review.opendev.org/664965 | 11:25 |
| openstackgerrit | Tobias Henkel proposed zuul/zuul master: Support ssl encrypted fingergw https://review.opendev.org/664950 | 11:26 |
| *** avass has joined #zuul | 11:33 | |
| *** bhavikdbavishi has quit IRC | 11:50 | |
| *** bhavikdbavishi has joined #zuul | 11:51 | |
| *** rlandy has joined #zuul | 11:56 | |
| *** rlandy is now known as rlandy|ruck | 11:56 | |
| tobiash | tristanC: I commented on the runner spec: https://review.opendev.org/681277 | 12:07 |
| *** rfolco has joined #zuul | 12:08 | |
| tobiash | sorry for the long review delay | 12:08 |
| *** bhavikdbavishi has quit IRC | 12:16 | |
| *** jcapitao_lunch is now known as jcapitao | 12:18 | |
| openstackgerrit | Andrii Ostapenko proposed zuul/zuul-jobs master: Fix issues with buildset registry https://review.opendev.org/741584 | 13:02 |
| openstackgerrit | Fabien Boucher proposed zuul/zuul master: gitlab: support the approval requirement https://review.opendev.org/741637 | 13:31 |
| openstackgerrit | Tobias Henkel proposed zuul/zuul master: Stop falling back to job name for missing url https://review.opendev.org/554056 | 13:39 |
| *** sgw1 has joined #zuul | 13:48 | |
| *** Goneri has joined #zuul | 13:57 | |
| openstackgerrit | Tobias Henkel proposed zuul/zuul master: Add testcase to show inconsistency during merging https://review.opendev.org/619245 | 14:53 |
| *** y2kenny has joined #zuul | 14:58 | |
| y2kenny | For nodepool, for a node to be in the ABORTED state and locked, is nodepool responsible to try again? I haven't implemented "hasRemainingQuota" or "hasProviderQuota" for my driver and I wonder if that's the cause of node being stuck (they were aborted because all available nodes were assigned already.) | 15:10 |
| clarkb | y2kenny: yes I think the expectation is that nodepool will detect the "I'm full state" then block there waiting for resources to be released so that they can be used for subsequent jobs | 15:17 |
| y2kenny | does nodepool stop trying after awhile? My understanding is that the default has*Quota always return true. | 15:18 |
| fungi | keep in mind that launchers are meant to scale in a distributed fashion, so you may have launchers servicing different providers and any which lack capacity to satisfy a request should reject it so another launcher can fulfil it instead | 15:18 |
| y2kenny | the way my node got aborted is via QuotaException. I have been reading the code and I noticed the pause_handler, etc. but I am not sure if there are anything the driver needs to do. | 15:19 |
| y2kenny | like... are there anything the nodepool driver needs to do to "unpause"? | 15:19 |
| y2kenny | actually... I wonder if it's because my launchComplete was not returning correct value | 15:21 |
| fungi | y2kenny: it looks like when passing through _waitForNodeSet() it will un-pause a request automatically when self.hasRemainingQuota(ntype) is truthy | 15:23 |
| y2kenny | ok | 15:24 |
| y2kenny | I have been returning launchesComplete=False if node_state is zk.ABORTED because my driver wasn't using QuotaException initially. That may have been the cause of things getting stuck.. | 15:24 |
| y2kenny | I am guessing there's a cap on the number of ongoing request so when all the request were in Aborted but not complete, nothing move forward. | 15:27 |
| avass | tobiash: I started frankensteining together a win_zuul_console written in powershell. I sort of have a tcp server so far so I might push that as a WIP change later | 15:42 |
| corvus | avass: we're looking at moving the zuul_console stuff to use unix domain sockets so that we don't have to run a server on the remote side; is there any potential to do something similar for win? | 15:44 |
| *** hamalq has joined #zuul | 15:45 | |
| avass | corvus: I have no idea :) | 15:45 |
| avass | corvus: but I'll look around | 15:45 |
| *** hamalq_ has joined #zuul | 15:47 | |
| corvus | neat, windows has apparently added support for unix domain sockets | 15:48 |
| *** jcapitao has quit IRC | 15:48 | |
| avass | yeah, but only for windows 10 :) | 15:48 |
| corvus | avass: and you're running 3.1 right? ;) | 15:49 |
| avass | close, win 7 ;) | 15:49 |
| avass | it's a bit mixed though and we're always trying to get rid of those old machines | 15:49 |
| corvus | avass: iirc, the idea was to have ssh forward the domain socket to the remote node, then have the command module do python logging to the domain socket; the executor plugin would be on the other end of that socket and read the log messages. that way it's all pretty invisible to the worker node. | 15:50 |
| *** hamalq has quit IRC | 15:50 | |
| corvus | avass: the second best option would probably be to do the same thing with tcp port forwarding, though that's more intensive since it requires a listening port on the executor for every connection, and a listening port on the worker node too for the forwarded end (though at least the listener on the worker can listen only on loopback, so there aren't any firewall issues) | 15:52 |
| avass | corvus: you mean this right? https://review.opendev.org/#/c/541434/ | 15:55 |
| corvus | avass: yep that's the one | 15:56 |
| corvus | i mean, it doesn't work yet, but i think it's a sound idea and would be an improvement, so it's worth thinking about new designs that are more like that than the current system | 15:56 |
| avass | yeah agreed, it would also solve the problem we have with running containers as 'separate hosts' on a static node | 15:57 |
| openstackgerrit | Fabien Boucher proposed zuul/zuul master: gitlab: support the labeled event https://review.opendev.org/741667 | 16:07 |
| openstackgerrit | Andrii Ostapenko proposed zuul/zuul-jobs master: Fix issues with buildset registry https://review.opendev.org/741584 | 16:11 |
| openstackgerrit | Andrii Ostapenko proposed zuul/zuul-jobs master: Fix issues with buildset registry https://review.opendev.org/741584 | 16:14 |
| corvus | tobiash: replied to your reply on 681277 | 16:14 |
| tobiash | corvus: that sounds great, another improvement would be the possibility to configure default ignores in a file like ~/.zuul | 16:19 |
| tobiash | avass: awesome :) | 16:19 |
| *** kklimonda has quit IRC | 16:20 | |
| *** kklimonda has joined #zuul | 16:20 | |
| tobiash | corvus, avass: maybe it's even possible to tunnel a unix socket on the executor to a tcp socket on windows | 16:21 |
| tobiash | then we could stick with the unix socket idea and only modify the target for windows | 16:22 |
| avass | tobiash, corvus: I've been trying to do that but can't get it to work. Not sure if I'm giving it the correct path for the windows host :/ | 16:25 |
| *** nils has quit IRC | 16:28 | |
| tobiash | According to this it should be possible to have mixed unix/tcp socket forwardings: https://serverfault.com/questions/127794/forward-local-port-or-socket-file-to-remote-socket-file | 16:28 |
| tobiash | However this relies on ssh which I think we cannot rely on windows simce that is using winrm | 16:29 |
| clarkb | socat? | 16:31 |
| tobiash | I doubt that socat can tunnel through winrm ;) | 16:32 |
| avass | tobiash: yeah I couldn't get unix or tcp ports working over ssh | 16:32 |
| *** yolanda has quit IRC | 16:39 | |
| openstackgerrit | Andrii Ostapenko proposed zuul/zuul-jobs master: Fix issues with buildset registry https://review.opendev.org/741584 | 16:41 |
| openstackgerrit | Andrii Ostapenko proposed zuul/zuul-jobs master: Fix certificate issue with use buildset registry https://review.opendev.org/741584 | 16:43 |
| openstackgerrit | Albin Vass proposed zuul/zuul master: WIP: add win_zuul_console.py https://review.opendev.org/741677 | 16:47 |
| avass | tobiash: that writes a file with the server script and runs that. not sure how to 'daemonize' a process in windows otherwise | 16:49 |
| avass | there's probably a much better way of doing that | 16:49 |
| tobiash | why not, I'm thinking about doing that on linux as well because the forking part is what breaks when using mitogen | 16:55 |
| openstackgerrit | Fabien Boucher proposed zuul/zuul master: gitlab: support the approval requirement https://review.opendev.org/741637 | 17:09 |
| openstackgerrit | Fabien Boucher proposed zuul/zuul master: gitlab: support the labeled event https://review.opendev.org/741667 | 17:09 |
| *** vishalmanchanda has quit IRC | 17:20 | |
| corvus | tobiash: no need to fork if we switch to unix domain | 17:53 |
| tobiash | that's true | 17:57 |
| avass | tobiash: yeah, but it's a bit annoying to keep all of the code in a string. :) | 18:16 |
| *** harrymichal has joined #zuul | 18:19 | |
| *** johnsom has quit IRC | 18:24 | |
| *** johnsom has joined #zuul | 18:25 | |
| openstackgerrit | Albin Vass proposed zuul/zuul master: WIP: add win_zuul_console.py https://review.opendev.org/741677 | 18:51 |
| openstackgerrit | Albin Vass proposed zuul/zuul master: WIP: add win_zuul_console.py https://review.opendev.org/741677 | 18:52 |
| *** kmalloc has quit IRC | 18:53 | |
| *** kmalloc has joined #zuul | 18:53 | |
| openstackgerrit | Merged zuul/zuul master: zuul_stream: avoid exception in _log_streamline https://review.opendev.org/741257 | 18:53 |
| *** rlandy|ruck is now known as rlandy|ruck|mtg | 19:36 | |
| *** maxamillion has quit IRC | 19:53 | |
| *** maxamillion has joined #zuul | 19:54 | |
| *** rlandy|ruck|mtg is now known as rlandy|ruck | 20:29 | |
| openstackgerrit | Andrii Ostapenko proposed zuul/zuul-jobs master: Add ability to use (upload|promote)-docker-image roles in periodic jobs https://review.opendev.org/740560 | 20:38 |
| *** ericsysmin has quit IRC | 20:46 | |
| *** ericsysmin has joined #zuul | 20:46 | |
| *** y2kenny has quit IRC | 20:54 | |
| corvus | does anyone know why i can't copy/paste from the streaming console using the x selection buffer? | 21:06 |
| fungi | confirmed i can reproduce | 21:09 |
| fungi | seems i can highlight and tell my browser to copy and then paste from the alternate (non middle-click) buffer | 21:10 |
| fungi | but normal highlight and then middle-click isn't working there | 21:11 |
| corvus | yeah, highlight in browser, copy, paste in emacs, highlight in emacs, middle-click in terminal does work. it's not ideal. | 21:12 |
| *** rpittau has quit IRC | 21:12 | |
| *** rpittau has joined #zuul | 21:13 | |
| corvus | i guess that's what i get for not just ircing in emacs | 21:13 |
| corvus | like a "real" emacs user would | 21:14 |
| fungi | the terminology for these seems to be "primary selection buffer" (highlight to copy, middle-click to paste) and "clipboard buffer" (highlight and explcitly ask to copy, then explicitly ask to paste) | 21:14 |
| corvus | i always knew it as the selection buffer, but now that you say it's called the "primary" selection buffer, i'm wondering if there are more, cause that would be cool. | 21:14 |
| fungi | https://tronche.com/gui/x/icccm/sec-2.html#s-2.6.1 | 21:15 |
| fungi | there are! | 21:15 |
| fungi | an arbitrary number even | 21:15 |
| fungi | anyway, to summarize, it seems that something about the new console stream implementation has broken the ability for highlighting to populate the primary selection buffer in at least some browsers | 21:17 |
| *** harrymichal has quit IRC | 21:26 | |
| *** hamalq has joined #zuul | 21:28 | |
| *** hamalq_ has quit IRC | 21:31 | |
| *** avass has quit IRC | 21:53 | |
| *** rlandy|ruck has quit IRC | 22:45 | |
| *** tosky_ has joined #zuul | 22:52 | |
| *** tosky has quit IRC | 22:53 | |
| *** tosky_ is now known as tosky | 23:01 | |
| *** rfolco has quit IRC | 23:16 | |
| *** armstrongs has joined #zuul | 23:31 | |
| *** tosky has quit IRC | 23:38 | |
| *** hamalq has quit IRC | 23:38 | |
| *** armstrongs has quit IRC | 23:40 | |
| *** rfolco has joined #zuul | 23:41 | |
Generated by irclog2html.py 2.17.2 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!