| *** jamesmcarthur has quit IRC | 00:00 | |
| *** jamesmcarthur has joined #zuul | 00:04 | |
| *** threestrands has joined #zuul | 00:09 | |
| *** jamesmcarthur has quit IRC | 00:13 | |
| *** rlandy has quit IRC | 00:14 | |
| *** jamesmcarthur has joined #zuul | 00:14 | |
| *** jamesmcarthur has quit IRC | 00:22 | |
| *** jamesmcarthur has joined #zuul | 00:54 | |
| *** jamesmcarthur has quit IRC | 01:05 | |
| *** ysandeep|away is now known as ysandeep | 01:16 | |
| *** jamesmcarthur has joined #zuul | 01:30 | |
| *** swest has quit IRC | 01:31 | |
| *** sshnaidm|afk is now known as sshnaidm|off | 01:34 | |
| *** jamesmcarthur has quit IRC | 01:36 | |
| *** cloudnull has quit IRC | 01:43 | |
| *** swest has joined #zuul | 01:46 | |
| *** jamesmcarthur has joined #zuul | 01:46 | |
| *** cloudnull has joined #zuul | 01:51 | |
| *** jamesmcarthur has quit IRC | 01:54 | |
| *** jamesmcarthur has joined #zuul | 01:55 | |
| *** jamesmcarthur has quit IRC | 02:02 | |
| *** jamesmcarthur has joined #zuul | 02:29 | |
| *** jamesmcarthur has quit IRC | 02:43 | |
| *** threestrands has quit IRC | 03:49 | |
| openstackgerrit | Andreas Jaeger proposed zuul/zuul-jobs master: WIP: Sync with configure-mirrors https://review.opendev.org/731293 | 04:17 |
|---|---|---|
| *** jamesmcarthur has joined #zuul | 04:25 | |
| *** evrardjp has quit IRC | 04:33 | |
| *** evrardjp has joined #zuul | 04:33 | |
| *** jamesmcarthur has quit IRC | 04:56 | |
| *** jamesmcarthur_ has joined #zuul | 04:56 | |
| *** sgw has quit IRC | 05:06 | |
| openstackgerrit | Andreas Jaeger proposed zuul/zuul-jobs master: Add configure-os-mirrors role https://review.opendev.org/677578 | 05:18 |
| AJaeger | avass: I squashed our changes together ^ | 05:26 |
| AJaeger | zuul-jobs-main, https://review.opendev.org/677578 is the new configure-os-mirrors role and ready for review, please have a look - and let's discuss how to migrate to it. | 05:30 |
| *** reiterative has quit IRC | 05:40 | |
| *** reiterative has joined #zuul | 05:40 | |
| *** dpawlik has joined #zuul | 06:07 | |
| openstackgerrit | Andreas Jaeger proposed zuul/zuul-jobs master: Fix typo: Deprecrated https://review.opendev.org/731439 | 06:10 |
| openstackgerrit | Merged zuul/zuul master: last3x release note: fix a typo https://review.opendev.org/731372 | 06:12 |
| openstackgerrit | Lida Liu proposed zuul/zuul master: WIP: Add commit id to Change for mqtt reporter https://review.opendev.org/722478 | 06:36 |
| *** yolanda has joined #zuul | 06:43 | |
| *** hashar has joined #zuul | 06:51 | |
| *** saneax has joined #zuul | 06:59 | |
| avass | AJaeger: I don't think we can filter by checking the distribution in the suite | 07:05 |
| avass | AJaeger: So that would have to be solved. Maybe by adding a 'release' field for each item or what fungi suggested: grouping them by 'ubuntu-bionic', 'ubuntu-xenial' etc instead of just 'ubuntu' | 07:07 |
| *** asaleh_ has joined #zuul | 07:07 | |
| AJaeger | avass: I didn't really follow that discussion, do you have a good idea on how to update? | 07:09 |
| openstackgerrit | Merged zuul/zuul-jobs master: Fix typo: Deprecrated https://review.opendev.org/731439 | 07:11 |
| avass | AJaeger: I added a comment, currently we're filtering by checking if the suite includes the release of the distribution but that doesn't have to be the case | 07:11 |
| avass | AJaeger: and if it does we add the configuration to sources.list | 07:11 |
| *** jamesmcarthur_ has quit IRC | 07:12 | |
| * AJaeger hopes somebody has a good idea | 07:13 | |
| *** jamesmcarthur has joined #zuul | 07:13 | |
| avass | AJaeger: it's convention to add the release to the suites name so it will probably work until someone wants to add a repository that doesn't follow the convention :) | 07:14 |
| *** jcapitao has joined #zuul | 07:14 | |
| AJaeger | bad luck ;) | 07:17 |
| *** jamesmcarthur has quit IRC | 07:19 | |
| *** jpena|off is now known as jpena | 07:35 | |
| *** tosky has joined #zuul | 07:44 | |
| openstackgerrit | Lida Liu proposed zuul/zuul master: WIP: Add commit id to Change for mqtt reporter https://review.opendev.org/722478 | 07:48 |
| zbr | avass: AJaeger: can you help with https://review.opendev.org/#/c/702679/ ? | 07:52 |
| AJaeger | zbr: such linter changes should get some more review, so let's ask later today again, please | 07:53 |
| AJaeger | mordred, corvus, clarkb ^ | 07:53 |
| zbr | maybe i should split changes out of it to merge them before reconfiguring the linter? | 07:54 |
| *** jamesmcarthur has joined #zuul | 07:54 | |
| AJaeger | zbr: going to 4.3 without any noqas etc., then fixing the bugs, then adding noqa would be perfect | 07:59 |
| zbr | AJaeger: makes no sense what you say, cannot bump a linter w/o skipping its new rules. | 08:00 |
| zbr | you either fix them first, or ignore them. | 08:00 |
| AJaeger | zbr: ignore them in the first change | 08:01 |
| *** jamesmcarthur has quit IRC | 08:05 | |
| *** nils has joined #zuul | 08:06 | |
| openstackgerrit | Sorin Sbarnea (zbr) proposed zuul/zuul-jobs master: Bump ansible-lint to 4.3.0 https://review.opendev.org/702679 | 08:25 |
| openstackgerrit | Jan Kubovy proposed zuul/zuul master: Connect merger to Zookeeper https://review.opendev.org/716221 | 08:39 |
| openstackgerrit | Jan Kubovy proposed zuul/zuul master: Connect executor to Zookeeper https://review.opendev.org/716262 | 08:39 |
| openstackgerrit | Jan Kubovy proposed zuul/zuul master: Connect fingergw to Zookeeper https://review.opendev.org/716875 | 08:39 |
| openstackgerrit | Sorin Sbarnea (zbr) proposed zuul/zuul-jobs master: Enable linting of test-playbooks https://review.opendev.org/731471 | 08:45 |
| openstackgerrit | Sorin Sbarnea (zbr) proposed zuul/zuul-jobs master: Bump ansible-lint to 4.3.0 https://review.opendev.org/702679 | 08:52 |
| *** rpittau|afk is now known as rpittau | 08:53 | |
| openstackgerrit | Jan Kubovy proposed zuul/zuul master: Enhance ZK CA script and doku https://review.opendev.org/731475 | 09:10 |
| openstackgerrit | Jan Kubovy proposed zuul/zuul master: Enhance ZK CA script and doku https://review.opendev.org/731475 | 09:11 |
| openstackgerrit | Jan Kubovy proposed zuul/zuul master: Enhance ZK CA script and doku https://review.opendev.org/731475 | 09:18 |
| *** sugaar has quit IRC | 09:22 | |
| *** sugaar has joined #zuul | 09:30 | |
| openstackgerrit | Lida Liu proposed zuul/zuul master: WIP: Add commit id to Change for mqtt reporter https://review.opendev.org/722478 | 09:34 |
| *** ysandeep is now known as ysandeep|lunch | 09:35 | |
| openstackgerrit | Sorin Sbarnea (zbr) proposed zuul/zuul-jobs master: Bump ansible-lint to 4.3.0 https://review.opendev.org/702679 | 09:45 |
| openstackgerrit | Albin Vass proposed zuul/zuul master: Use absolute path to toolsdir for zk-ca.sh config https://review.opendev.org/731484 | 09:49 |
| openstackgerrit | Albin Vass proposed zuul/zuul master: Use Absolute path in zk-ca.sh for openssl config https://review.opendev.org/731484 | 09:51 |
| openstackgerrit | Albin Vass proposed zuul/zuul master: Use Absolute path in zk-ca.sh for openssl config https://review.opendev.org/731484 | 09:52 |
| *** tosky_ has joined #zuul | 09:57 | |
| *** tosky is now known as Guest48633 | 09:57 | |
| *** tosky_ is now known as tosky | 09:57 | |
| openstackgerrit | Sorin Sbarnea (zbr) proposed zuul/zuul-jobs master: Bump ansible-lint to 4.3.0 https://review.opendev.org/702679 | 10:09 |
| *** rpittau is now known as rpittau|bbl | 10:18 | |
| *** ysandeep|lunch is now known as ysandeep | 10:18 | |
| *** jamesmcarthur has joined #zuul | 10:39 | |
| openstackgerrit | Lida Liu proposed zuul/zuul master: WIP: Add commit id to Change for mqtt reporter https://review.opendev.org/722478 | 10:46 |
| *** jamesmcarthur has quit IRC | 10:48 | |
| *** sshnaidm|off has quit IRC | 10:55 | |
| *** jcapitao is now known as jcapitao_lunch | 10:57 | |
| *** fbo|off is now known as fbo | 11:05 | |
| openstackgerrit | Tristan Cacqueray proposed zuul/zuul-jobs master: shake-build: add shake build system job https://review.opendev.org/731365 | 11:10 |
| zbr | avass: AJaeger: ansible-lint bumping ^ is ready for rreview, w/o any new skips. | 11:15 |
| *** jamesmcarthur has joined #zuul | 11:18 | |
| *** jamesmcarthur_ has joined #zuul | 11:19 | |
| *** jamesmcarthur has quit IRC | 11:23 | |
| *** jamesmcarthur_ has quit IRC | 11:28 | |
| tobiash | corvus, clarkb: does opendev still offload the javascript to apache/nginx? I just noticed that the opendev-promote-javascript-deployment job fails in the promote pipeline. Maybe that's the cause of not having the timezone selector in opendev | 11:35 |
| *** jpena is now known as jpena|lunch | 11:39 | |
| *** rpittau|bbl is now known as rpittau | 12:05 | |
| *** jcapitao_lunch is now known as jcapitao | 12:13 | |
| *** rlandy has joined #zuul | 12:16 | |
| openstackgerrit | Oleksandr Kozachenko proposed zuul/zuul-jobs master: Add the namespace in wait-for-pods role https://review.opendev.org/731292 | 12:44 |
| openstackgerrit | Oleksandr Kozachenko proposed zuul/zuul-jobs master: Add namespace in the collect-k8s-logs role https://review.opendev.org/731319 | 12:44 |
| *** jpena|lunch is now known as jpena | 12:44 | |
| *** Goneri has joined #zuul | 12:46 | |
| openstackgerrit | Tristan Cacqueray proposed zuul/zuul-jobs master: shake-build: add shake build system job https://review.opendev.org/731365 | 12:48 |
| openstackgerrit | Lida Liu proposed zuul/zuul master: WIP: Add commit id to Change for mqtt reporter https://review.opendev.org/722478 | 12:49 |
| *** harrymichal has joined #zuul | 12:49 | |
| *** rlandy is now known as rlandy|mtg | 13:03 | |
| openstackgerrit | Oleksandr Kozachenko proposed zuul/zuul-jobs master: Add addons args in ensure-kubernetes role https://review.opendev.org/731523 | 13:06 |
| *** sanjayu_ has joined #zuul | 13:08 | |
| openstackgerrit | Oleksandr Kozachenko proposed zuul/zuul-jobs master: Add the namespace in wait-for-pods role https://review.opendev.org/731292 | 13:09 |
| openstackgerrit | Oleksandr Kozachenko proposed zuul/zuul-jobs master: Add namespace in the collect-k8s-logs role https://review.opendev.org/731319 | 13:09 |
| *** saneax has quit IRC | 13:10 | |
| openstackgerrit | Albin Vass proposed zuul/zuul master: zookeeper-howto: update to match zk-ca.sh https://review.opendev.org/731526 | 13:25 |
| openstackgerrit | Merged zuul/zuul-jobs master: shake-build: add shake build system job https://review.opendev.org/731365 | 13:28 |
| openstackgerrit | Albin Vass proposed zuul/zuul-jobs master: upload-artifactory: fix variable misspelling https://review.opendev.org/731535 | 13:37 |
| openstackgerrit | Albin Vass proposed zuul/zuul-jobs master: upload-artifactory: fix variable misspelling https://review.opendev.org/731535 | 13:38 |
| avass | AJaeger: that ^ should be fine right? The documentation says 'instances' while the code was using 'instance' and I hardly believe someone other than us uses that yet | 13:39 |
| avass | AJaeger: also, I noticed I probably changed that role enough so I don't think that actually needs a trusted context | 13:39 |
| avass | AJaeger: it was using lookups earlier so I think I just got it stuck in my head that it needed a trusted context :) | 13:42 |
| Open10K8S | Hi team. | 13:47 |
| Open10K8S | I made 3 patchsets on zuul-jobs repo. | 13:47 |
| Open10K8S | I used them on https://review.opendev.org/#/c/731262/ | 13:47 |
| *** sgw has joined #zuul | 13:47 | |
| Open10K8S | please check them | 13:47 |
| Open10K8S | AJaeger: updated the readme | 13:47 |
| AJaeger | Open10K8S: thanks, will review later... | 13:48 |
| Open10K8S | AJaeger: ok, thank you. | 13:48 |
| Open10K8S | will wait. | 13:48 |
| Open10K8S | Also made another one related to the ensure-kubernetes role | 13:48 |
| Open10K8S | One PS is waiting them to merge | 13:48 |
| mordred | Open10K8S: re: adding the namespace argument to the kubectl command - does that work with the nodepool k8s-namespace driver? (I don't lknow if these roles work with that or not) | 13:54 |
| clarkb | tobiash: I dont think so. We deploy from the containers now but it is possible that our apache is still serving old files? | 13:54 |
| clarkb | mordred: ^ see tobiash's not about opemdev's dashboard missing the new timezone selector | 13:55 |
| mordred | clarkb: maybe we haven't restarted our zuul-web container? | 13:55 |
| mordred | (since we don't do static offload anymore, we don't get insta-deploys of javascript) | 13:55 |
| clarkb | mordred: corvus did that yesterday aiui | 13:57 |
| mordred | ah - hrm | 13:57 |
| mordred | clarkb: 9e200492a0f1 bae28a4c5273 "/usr/bin/dumb-init …" 12 days ago Up 15 hours zuul-web_web_1 | 13:58 |
| mordred | clarkb: the container was definitely restarted 15 hours ago - but it says created 12 days ago | 13:59 |
| clarkb | hrm maybe we havent built and published a new web image for some reason | 13:59 |
| mordred | nope - we have new images | 13:59 |
| mordred | on the host are: | 13:59 |
| mordred | zuul/zuul-web latest 2b24f510e073 8 hours ago 455MB | 13:59 |
| mordred | zuul/zuul-web <none> bae28a4c5273 13 days ago 471MB | 14:00 |
| mordred | we're running bae28a4c5273 | 14:00 |
| fungi | so haven't retrieved a new image? | 14:00 |
| fungi | oh, on the host | 14:00 |
| clarkb | fungi: looks like we have | 14:00 |
| fungi | so we retrieved it but restarted on the old one | 14:00 |
| mordred | yeah - but something about the restart did not cause a new image to be used | 14:00 |
| mordred | yah | 14:00 |
| clarkb | wekeep the latest and the one we are running which matches mordreds list | 14:00 |
| mordred | want me to try restarting zuul-web again? | 14:01 |
| clarkb | mordred: ya maybe be sure to use docker compose stop then docker compose up -d to dpible chwck the process ansible uses is working? | 14:02 |
| mordred | clarkb: k. looks yesterday pull and restart was run | 14:02 |
| mordred | so maybe restart doesn't use a new image | 14:03 |
| mordred | actually - why don't I try restart again real quick - just to verify the restart behavior | 14:03 |
| clarkb | or will only if the previous pull got an update | 14:03 |
| mordred | yeah. restart does not seem to pick a new image | 14:03 |
| clarkb | we are always pulling the latest image in the background | 14:03 |
| clarkb | could be that then | 14:04 |
| mordred | (which is good to know - I guess maybe that's useful behavior sometimes " I want to restart this container but I don't want to use a new image" | 14:04 |
| mordred | doing down and up now | 14:04 |
| mordred | we are now running the new image | 14:04 |
| corvus | mordred, clarkb: aha, thanks for the education :) | 14:07 |
| mordred | corvus: we're learning things! :) | 14:07 |
| corvus | we should also look into tobiash's observation about the js promote job being borked (even though we're not using it) | 14:08 |
| mordred | I guess now that I think about it it makes sense. restart is "restart this existing container" and "down / up" is "delete this container and make a new one from the same spec" | 14:08 |
| mordred | corvus: ++ | 14:08 |
| openstackgerrit | Lida Liu proposed zuul/zuul master: WIP: Add commit id to Change for mqtt reporter https://review.opendev.org/722478 | 14:10 |
| avass | mordred: no restart doesn't pick up any new config | 14:16 |
| avass | mordred: actually, it doesn't recreate the container | 14:16 |
| avass | mordred: and 'up' doesn't do anything if nothing needs doing :) | 14:17 |
| *** rpittau is now known as rpittau|brb | 14:18 | |
| avass | mordred: I believe 'up' only recreates the container if there's a new image available or the docker-compose file has a relevant update | 14:20 |
| openstackgerrit | Albin Vass proposed zuul/zuul-jobs master: upload-artifactory: fix variable misspelling https://review.opendev.org/731535 | 14:26 |
| *** rpittau|brb is now known as rpittau | 14:29 | |
| avass | (just to confirm that's how it works) | 14:31 |
| mordred | avass: yah. I agree with you. oh - so we really couyld have just done docker-compose up -d in this case without the explicit down | 14:34 |
| avass | mordred: yeah | 14:34 |
| avass | mordred: and you can bring up specific services while you can't bring down specific services | 14:35 |
| avass | mordred: with 'down' that is, of course you can still kill them | 14:35 |
| avass | mordred: running 'up' twice shouldn't normally do anything the second time | 14:36 |
| SpamapS | Hey everyone, congrats on turning the corner toward Zuul 4! | 14:53 |
| fungi | thanks SpamapS! | 14:53 |
| corvus | Zuul 4 all | 14:54 |
| avass | corvus: can you take a quick look at this by the way: https://review.opendev.org/#/c/731526/1 | 14:56 |
| avass | corvus: I found that the documentation wasn't matching what zk-ca.sh was doing when getting ready for zuul 4 :) | 14:56 |
| corvus | avass: oh good catch, thanks! | 14:56 |
| corvus | tristanC: ^ | 14:57 |
| avass | corvus: the other part was this: https://review.opendev.org/731484 but I'm not sure if it was a specific version of openssl or why it didn't like the relative path | 14:59 |
| *** jcapitao is now known as jcapitao_afk | 15:14 | |
| Open10K8S | mordred: Regarding to nodepool k8s-namespace, | 15:23 |
| Open10K8S | mordred: | 15:24 |
| Open10K8S | good question | 15:24 |
| Open10K8S | mordred: how about this? | 15:24 |
| Open10K8S | when the namespace variable is not specified, then eliminate the -n clause | 15:24 |
| Open10K8S | so we can allow the k8s-namespace also | 15:24 |
| Open10K8S | in the kubeconfig context | 15:25 |
| *** jamesmcarthur has joined #zuul | 15:25 | |
| openstackgerrit | Albin Vass proposed zuul/zuul-jobs master: upload-artifactory: documentation fix https://review.opendev.org/731170 | 15:28 |
| avass | AJaeger: ^ :) | 15:29 |
| corvus | Open10K8S, mordred: i don't think the collect-kubernetes-logs role was originally intended for use with nodepool namespaces (instead, i think it was designed for a job-deployed k8s), but i think perhaps it could be (if used in a trusted playbook), so i think that approach may be worth taking | 15:30 |
| *** jamesmcarthur has quit IRC | 15:30 | |
| clarkb | corvus: ya it was originally written to help debug job deployed k8s services | 15:31 |
| zbr | clarkb: corvus: avass guidance on https://review.opendev.org/#/c/702679/13 ? | 15:33 |
| avass | zbr: ++, but depending on how verbose that is we might want to lower that later | 15:34 |
| zbr | avass: yep, i love that approach. verbose for a while until we get used to it. | 15:35 |
| zbr | once we gain more trust, we can lower it. | 15:35 |
| corvus | i don't understand why that should be in the job | 15:36 |
| corvus | if you want to debug it, just run it locally | 15:36 |
| corvus | avass: i -2d it to stop the merge (it seemed like tristanC and zbr wanted more feedback) | 15:38 |
| corvus | my vote is really a -1 | 15:38 |
| zbr | no worry, i can remove the -v* | 15:38 |
| corvus | zbr: you could consider an env variable to make it easier to add the -vvv locally | 15:39 |
| corvus | i changed my -2 to a C-1,W-1 to make it easier to push through a new patchset without the -vvv | 15:40 |
| zbr | okey | 15:40 |
| *** harrymichal has quit IRC | 15:40 | |
| *** harrymichal has joined #zuul | 15:40 | |
| openstackgerrit | Lida Liu proposed zuul/zuul master: WIP: Add commit id to Change for mqtt reporter https://review.opendev.org/722478 | 15:44 |
| *** harrymichal has quit IRC | 15:47 | |
| openstackgerrit | Sorin Sbarnea (zbr) proposed zuul/zuul-jobs master: Bump ansible-lint to 4.3.0 https://review.opendev.org/702679 | 15:50 |
| openstackgerrit | Merged zuul/zuul-jobs master: upload-artifactory: documentation fix https://review.opendev.org/731170 | 15:50 |
| *** jcapitao_afk is now known as jcapitao | 15:53 | |
| openstackgerrit | Merged zuul/zuul master: zookeeper-howto: update to match zk-ca.sh https://review.opendev.org/731526 | 15:57 |
| fungi | corvus: not urgent, but when you get a moment i've suggested another alternative on 731246 | 15:57 |
| *** jamesmcarthur has joined #zuul | 16:01 | |
| AJaeger | corvus: when you have some quiet time, could you review the configure-os-mirror change with fresh eyes, please? https://review.opendev.org/#/c/677578/ avass and myself got it passing all tests - and now it needs a good review and answering the comment that avass left based on the discussion from last night. | 16:02 |
| *** ysandeep is now known as ysandeep|afk | 16:02 | |
| corvus | fungi: i think you nailed it | 16:03 |
| corvus | AJaeger: will do | 16:04 |
| fungi | cool, will push that up then. thanks! | 16:04 |
| *** bhagyashris|ruck is now known as bhagyashris | 16:05 | |
| *** hashar has quit IRC | 16:05 | |
| openstackgerrit | Jeremy Stanley proposed zuul/zuul master: Docs: CPD between non-shared-queue changes https://review.opendev.org/731246 | 16:06 |
| *** rpittau is now known as rpittau|afk | 16:08 | |
| openstackgerrit | Oleksandr Kozachenko proposed zuul/zuul-jobs master: Add the namespace in wait-for-pods role https://review.opendev.org/731292 | 16:13 |
| openstackgerrit | Oleksandr Kozachenko proposed zuul/zuul-jobs master: Add namespace in the collect-k8s-logs role https://review.opendev.org/731319 | 16:13 |
| openstackgerrit | Oleksandr Kozachenko proposed zuul/zuul-jobs master: Add namespace in the collect-k8s-logs role https://review.opendev.org/731319 | 16:14 |
| openstackgerrit | Oleksandr Kozachenko proposed zuul/zuul-jobs master: Add addons args in ensure-kubernetes role https://review.opendev.org/731523 | 16:16 |
| *** dpawlik has quit IRC | 16:19 | |
| *** nils has quit IRC | 16:25 | |
| *** harrymichal has joined #zuul | 16:40 | |
| AJaeger | corvus: is 702679 now good? | 16:40 |
| corvus | lgtm | 16:41 |
| *** jcapitao has quit IRC | 16:43 | |
| *** asaleh_ has quit IRC | 16:44 | |
| *** guillaumec has joined #zuul | 16:47 | |
| openstackgerrit | Oleksandr Kozachenko proposed zuul/zuul-jobs master: Add the namespace in wait-for-pods role https://review.opendev.org/731292 | 16:52 |
| openstackgerrit | Oleksandr Kozachenko proposed zuul/zuul-jobs master: Add namespace in the collect-k8s-logs role https://review.opendev.org/731319 | 16:52 |
| *** jpena is now known as jpena|off | 16:53 | |
| *** harrymichal has quit IRC | 16:53 | |
| *** harrymichal has joined #zuul | 16:54 | |
| openstackgerrit | Merged zuul/zuul-jobs master: Bump ansible-lint to 4.3.0 https://review.opendev.org/702679 | 16:55 |
| *** ysandeep|afk is now known as ysandeep | 16:57 | |
| *** rlandy|mtg is now known as rlandy | 17:01 | |
| *** jamesmcarthur has quit IRC | 17:03 | |
| openstackgerrit | Merged zuul/zuul master: Use Absolute path in zk-ca.sh for openssl config https://review.opendev.org/731484 | 17:06 |
| openstackgerrit | Merged zuul/zuul master: requirements: add explicit reference to dateutil https://review.opendev.org/730654 | 17:08 |
| *** ysandeep is now known as ysandeep|away | 17:13 | |
| AJaeger | zuul-jobs-maint, please review these small changes https://review.opendev.org/730720 https://review.opendev.org/731535 https://review.opendev.org/730733 | 17:15 |
| avass | corvus: sorry, no problem | 17:15 |
| *** olaph has quit IRC | 17:20 | |
| openstackgerrit | Sorin Sbarnea (zbr) proposed zuul/zuul-jobs master: Enable linting of test-playbooks https://review.opendev.org/731471 | 17:27 |
| corvus | AJaeger, zbr: can we be more conservative with ansible-lint? | 17:41 |
| corvus | i feel like we're spending too much time on linting rules -- especially on ansible-lint which we very nearly stopped using completely because of how many useless rules it has | 17:42 |
| corvus | for example, the "zj_" loop var is a great idea and will help us avoid errors. the whole "don't use the git command" thing is just a red herring for us because it's so often wrong. | 17:43 |
| zbr | corvus: i hope not, most of them are useful, and having decent coding standards would keep code maintainable. | 17:43 |
| zbr | it is very easy to add stuff to whitelist | 17:44 |
| zbr | same as with flake8 | 17:44 |
| avass | zbr, corvus: I think the problem is that in the context in zuul a lot of the normal ansible rules don't really apply | 17:44 |
| corvus | yes, so if we're going to use linters at all (i think we'd be better off without them, but i'm compromising here :) let's use them where they provide actual value: catching bugs. | 17:46 |
| zbr | i am not so sure about the different context, i do not see a reason to have two quality standards | 17:46 |
| fungi | my understanding was that ansible-lint was intended to be the rule enforcer for ansible roles published to galaxy, while it seems unlikely most of our zuul-jobs roles would ever benefit from that | 17:46 |
| corvus | fungi: indeed, and the last time we had this conversation, that's why we decided its use in zuul-jobs should be minimized | 17:46 |
| zbr | as I said, skiplist works fine for stuff we do not agree woth. | 17:46 |
| corvus | i have to admit, i'm frustrated that we're having it again without the previous context :( | 17:46 |
| corvus | it only reinforces my belief that linters are 10% useful and 90% a waste of valuable developer time arguing about whitespace | 17:47 |
| openstackgerrit | Merged zuul/zuul-jobs master: Rename test install role to ensure- https://review.opendev.org/730720 | 17:47 |
| zbr | it would be more useful to see your feedback on ansible-lint issue tracker | 17:48 |
| corvus | zbr: i think adding most of those 3xx rules in 731471 to the whitelist would probably be a good start | 17:48 |
| *** cloudnull has quit IRC | 17:48 | |
| zbr | one of the benefit is that linter avoids wasting time debating whitespace during code reviews, it does save time, not take time. | 17:49 |
| corvus | zbr: on the contrary, that does not seem to be the result | 17:49 |
| corvus | zbr: we have a different approach in the zuul project: we just don't nitpick whitespace | 17:49 |
| corvus | zbr: we don't let it bother us | 17:49 |
| zbr | so we do have low quality standards? | 17:49 |
| *** cloudnull has joined #zuul | 17:49 | |
| corvus | zbr: whitespace is not quality | 17:49 |
| corvus | zbr: we promote a culture of not nitpicking | 17:50 |
| avass | AJaeger: ^ ;) | 17:50 |
| corvus | it results in people being happier | 17:50 |
| zbr | i bet to defer that the whitespace on jinja is important, the same kind of importance as whitespace on python code | 17:50 |
| corvus | i think you know what i mean, let's not argue about that | 17:51 |
| corvus | the point i'm trying to convey is that linting rules that catch bugs are welcome, and linting rules that are frequently erroneous or are mere stylistic preferences should be disabled | 17:52 |
| corvus | and that, when it does come to style, we should avoid nitpicking (yes -- sometimes style is important; but we should be careful about knowing when it is and isn't) | 17:53 |
| zbr | corvus: huh, glad to hear that. Yes, a buggy one should clearly be whitelisted. | 17:53 |
| zbr | one of the reason some of you hated the linter is that had few buggy rules, most of them where fixed in newer versions. | 17:54 |
| zbr | so less false-positives | 17:54 |
| corvus | zbr: so for a start, must of those 3xx rules we hit frequently enough they should probably be disabled | 17:54 |
| zbr | corvus: here is the fun bit: we did lint only rules/ which was bad, as test-playbooks/ is ansible too | 17:55 |
| zbr | the rules/ is ok and does not need changes. | 17:55 |
| zbr | roles/ | 17:55 |
| corvus | well, we deemed it more important | 17:55 |
| openstackgerrit | Merged zuul/zuul master: Docs: CPD between non-shared-queue changes https://review.opendev.org/731246 | 17:55 |
| zbr | corvus: true, but longer term we should normalize, which is what i did. | 17:56 |
| *** jamesmcarthur has joined #zuul | 17:56 | |
| zbr | roles/ is quite in good shape from this point of view | 17:56 |
| zbr | corvus: if you want I am willing to consider whitelisting E602 which is probably one of the most controversial one | 17:57 |
| corvus | looking at that patch, it is approximately 50% #noqa lines -- that's a red flag | 17:57 |
| avass | I think we should keep the pipefail rule though, that seems like it could catch errors | 17:57 |
| zbr | i will split it by nature of fix, it would be much easier to review. | 17:57 |
| zbr | avass++ true | 17:58 |
| corvus | every instance of "#noqa 306" in that patch is correct as written | 17:59 |
| zbr | for example 303/305: whitelisting them would be a regression in quality because at this moment roles/ does respect them, is only test-playbooks/ that does not. | 17:59 |
| corvus | so 306 caught zero bugs there | 18:00 |
| zbr | maybe in our case, but what about newly proposed code? code where OP would have missed to use a module that is idempotent and used shell instead? | 18:01 |
| zbr | even ansible itself does raise warnings on stuff like this | 18:01 |
| zbr | that is why we have: args: warn: false | 18:01 |
| zbr | which is supposed to be respected by linter too (i think) | 18:02 |
| tristanC | zbr: why not using args: warn: false then? | 18:02 |
| avass | hmm, I guess we only pipes with grep and anything else is just done in ansible | 18:02 |
| zbr | tristanC: that is a very good point. | 18:02 |
| zbr | i will look at warn: false bit, if it does not work I will propose a patch for the linter. | 18:03 |
| corvus | frankly, i'm not worried about having two quality standards for "real" and "test" code | 18:05 |
| corvus | all the ansible in test-playbooks is, by definition, tested. so the linter is unlikely to catch additional bugs. | 18:05 |
| tristanC | corvus: agreed | 18:06 |
| fungi | is ansible-lint catching actual bugs yamllint wouldn't? | 18:07 |
| *** jamesmcarthur_ has joined #zuul | 18:07 | |
| corvus | anyway, i've spent too long on linters for today; i'm supposed to be reviewing circular dependencies; i'm going to need to ignore irc a bit to get that done | 18:07 |
| fungi | i guess it checks that stuff you reference actually exists | 18:07 |
| *** fbo is now known as fbo|off | 18:07 | |
| openstackgerrit | Lida Liu proposed zuul/zuul master: WIP: Add commit id to Change for mqtt reporter https://review.opendev.org/722478 | 18:10 |
| *** jamesmcarthur has quit IRC | 18:10 | |
| openstackgerrit | Sorin Sbarnea (zbr) proposed zuul/zuul-jobs master: test-playbooks: improved syntax https://review.opendev.org/731591 | 18:12 |
| zbr | for example is able to detect a missing module (a typo in module name or argument may even make the reviewer miss it) | 18:13 |
| avass | fungi: https://zuul-ci.org/docs/zuul-jobs/policy.html#ansible-linting-rules :) | 18:13 |
| avass | fungi: I still have to finish the no-same-owner rule though | 18:13 |
| *** yolanda has quit IRC | 18:22 | |
| *** aspiers has quit IRC | 18:31 | |
| *** kmalloc has joined #zuul | 18:41 | |
| *** aspiers has joined #zuul | 18:43 | |
| Open10K8S | mordred: corvus: please check the PSs | 18:51 |
| Open10K8S | I tried to follow your advices :) | 18:51 |
| Open10K8S | https://review.opendev.org/#/c/731292/ | 18:51 |
| Open10K8S | https://review.opendev.org/#/c/731319/6 | 18:51 |
| Open10K8S | https://review.opendev.org/#/c/731523/ | 18:51 |
| *** aspiers has quit IRC | 18:57 | |
| avass | Open10K8S: commented on 731292 | 19:01 |
| Open10K8S | avass: just received | 19:01 |
| avass | Open10K8S: but I'd wait to see what the others think | 19:03 |
| *** jamesmcarthur_ has quit IRC | 19:04 | |
| openstackgerrit | Matthieu Huin proposed zuul/zuul master: web UI: add Autoholds Requests page https://review.opendev.org/729307 | 19:09 |
| avass | Open10K8S: same with 731523, it's not a -1 since it will probably not cause a problem, but I'd rather avoid setting a fact if it's not too much work :) | 19:11 |
| mhu | Hey there, I think the REST API How-to page is good to go: https://review.opendev.org/#/c/727785/ | 19:12 |
| Open10K8S | avass: ok, thank you avas | 19:12 |
| Open10K8S | I agree your idea, but we can use facts in case which we need to use several times the same fact | 19:12 |
| Open10K8S | so for the addons case, that would be good | 19:13 |
| Open10K8S | wait-for-pod case, i think it is convenient using the fact | 19:13 |
| Open10K8S | how about your idea? | 19:13 |
| mhu | If anybody has also time for reviews on https://review.opendev.org/#/q/topic:fix_rest_client I'd be grateful | 19:15 |
| *** harrymichal has quit IRC | 19:22 | |
| avass | Open10K8S: I do see the convenience of it and I like tristanC's way of using defaults for that: https://opendev.org/zuul/zuul-jobs/src/branch/master/roles/haskell-stack-test/defaults/main.yaml#L5 | 19:22 |
| avass | Open10K8S: that way the variable isn't set outside the role | 19:22 |
| Open10K8S | avass: looks good | 19:23 |
| *** jamesmcarthur has joined #zuul | 19:23 | |
| Open10K8S | ok then lets do like that | 19:23 |
| clarkb | installing zuul's deps in tox for unittesting isn't very quick. I think the reason is the node installation. Mostly an observation I had that we might be able to speed those jobs up a bit if we can make that faster | 19:23 |
| *** harrymichal has joined #zuul | 19:24 | |
| openstackgerrit | Sorin Sbarnea (zbr) proposed zuul/zuul-jobs master: Enable linting of test-playbooks https://review.opendev.org/731471 | 19:26 |
| avass | Open10K8S: have a look at the guidelines as well if you haven't already. We also want to namespace the variables (where it makes sense) to avoid something else accidentally overriding it: https://zuul-ci.org/docs/zuul-jobs/policy.html#role-variable-naming-policy | 19:27 |
| avass | but I'm not sure how hard we're enforcing that :) | 19:28 |
| openstackgerrit | Sorin Sbarnea (zbr) proposed zuul/zuul-jobs master: test-playbooks: avoid warnings with shell/command https://review.opendev.org/731605 | 19:47 |
| openstackgerrit | Albin Vass proposed zuul/zuul-jobs master: Update docs for os specific task guidelines https://review.opendev.org/731606 | 19:49 |
| openstackgerrit | Albin Vass proposed zuul/zuul-jobs master: Update guideline docs for os specific tasks https://review.opendev.org/731606 | 19:50 |
| openstackgerrit | Albin Vass proposed zuul/zuul-jobs master: Update guideline docs for os specific tasks https://review.opendev.org/731606 | 19:52 |
| openstackgerrit | Albin Vass proposed zuul/zuul-jobs master: Update guideline docs for os specific tasks https://review.opendev.org/731606 | 19:55 |
| avass | zuul-jobs-maint: I noticed the guidelines are different from what we're actually doing with multi os support. Can we update that? ^ | 20:04 |
| openstackgerrit | Albin Vass proposed zuul/zuul-jobs master: Update guideline docs for os specific tasks https://review.opendev.org/731606 | 20:06 |
| openstackgerrit | Albin Vass proposed zuul/zuul-jobs master: Update guideline docs for os specific tasks https://review.opendev.org/731606 | 20:08 |
| avass | sorry I missed removing some words. should be ready now :) | 20:08 |
| *** jamesmcarthur has quit IRC | 21:06 | |
| *** jamesmcarthur has joined #zuul | 21:07 | |
| *** jamesmcarthur has quit IRC | 21:12 | |
| *** jamesmcarthur has joined #zuul | 21:15 | |
| *** sanjayu_ has quit IRC | 21:31 | |
| openstackgerrit | Guillaume Chauvel proposed zuul/zuul master: Import user tutorials from Software Factory project blog https://review.opendev.org/728193 | 21:36 |
| openstackgerrit | Guillaume Chauvel proposed zuul/zuul master: Add tutorial tests https://review.opendev.org/728194 | 21:36 |
| *** threestrands has joined #zuul | 21:45 | |
| corvus | tobiash: all right, i left some questions on https://review.opendev.org/685354 | 21:56 |
| tobiash | corvus: thanks! I'll check tomorrow | 22:00 |
| corvus | AJaeger, avass: let's ask fungi to do the next review pass on https://review.opendev.org/677578, then tristanC. | 22:00 |
| *** jamesmcarthur has quit IRC | 22:04 | |
| *** jamesmcarthur has joined #zuul | 22:05 | |
| fungi | i bet i even know which change that is ;) | 22:05 |
| fungi | yep, exactly the one i was expecting | 22:05 |
| *** jamesmcarthur has quit IRC | 22:11 | |
| openstackgerrit | Guillaume Chauvel proposed zuul/zuul master: Import user tutorials from Software Factory project blog https://review.opendev.org/728193 | 22:21 |
| openstackgerrit | Guillaume Chauvel proposed zuul/zuul master: Add tutorial tests https://review.opendev.org/728194 | 22:21 |
| *** jamesmcarthur has joined #zuul | 22:21 | |
| *** rlandy has quit IRC | 22:38 | |
| *** armstrongs has joined #zuul | 22:42 | |
| armstrongs | Hey what happened in the end with the bitbucket driver did it ever get merged to master like the gitlab driver? | 22:43 |
| *** armstrongs has quit IRC | 22:55 | |
| *** sanjayu_ has joined #zuul | 22:59 | |
| mnaser | so i just ran into an issue with pip running in a constrained resource environment (in my case, an instance with 1gb of memory only). in fedora-32, /tmp is tmpfs so you only end up with 487M available in /tmp | 23:02 |
| mnaser | you end up running into: "Could not install packages due to an EnvironmentError: [Errno 28] No space left on device" during tox initial run with my assumption being that /tmp is running out of disk space | 23:02 |
| mnaser | would it make sense for us to define TMPDIR in our jobs to be {{ ansible_user }}/tmp or something | 23:02 |
| *** jamesmcarthur has quit IRC | 23:03 | |
| mnaser | because / has 50G and /tmp has only 500M | 23:03 |
| clarkb | mnaser: might be better to make /tmp not a tmpfs | 23:03 |
| mnaser | clarkb: right, is there a way to do that in dib then? because dib is building an image with /tmp as tmpfs | 23:03 |
| *** jamesmcarthur has joined #zuul | 23:03 | |
| clarkb | mnaser: yes, you can either edit fstab (and just remove the /tmp entry and it will end up in the same fs as /) or use the partitioning support to put it in its own fs | 23:04 |
| openstackgerrit | Oleksandr Kozachenko proposed zuul/zuul-jobs master: Add the namespace in wait-for-pods role https://review.opendev.org/731292 | 23:04 |
| openstackgerrit | Oleksandr Kozachenko proposed zuul/zuul-jobs master: Add namespace in the collect-k8s-logs role https://review.opendev.org/731319 | 23:04 |
| clarkb | ianw: ^ you may have thoughts on mnaser's fedora /tmp problem | 23:04 |
| mnaser | clarkb: cool well unforunately it seems like tmpfs is created by not fstab cause fstab only contains the root :( | 23:04 |
| mnaser | `LABEL=cloudimg-rootfs / ext4 defaults 0 1` is all i have in fstab | 23:05 |
| mnaser | so i think maybe something systemd-ish | 23:05 |
| mnaser | ` tmp.mount loaded active mounted Temporary Directory (/tmp)` | 23:05 |
| clarkb | ah disable that unit then I guess | 23:05 |
| clarkb | thats really weird to put in a unit | 23:05 |
| mnaser | this is fedora-32 fwiw | 23:05 |
| clarkb | but I probably don't understand why systemd needs to be more awesome and manage things like that | 23:06 |
| mnaser | so i dont think infra has ran into it yet, but also infra uses larger vms | 23:06 |
| clarkb | ya fedora tends to be a huge pain to deal with updating | 23:06 |
| mnaser | i guess the tricky / annoying thing is the fact that its nice to have if you have a lot of memory | 23:06 |
| clarkb | ianw somehow managed to get through it though | 23:06 |
| mnaser | but ugly if you dont | 23:06 |
| mnaser | (tmpfs that is) | 23:06 |
| clarkb | mnaser: ya an improvement to fedora might be to make that switch based on available memory | 23:07 |
| guilhermesp | i was wondering if –no-tmpfs would do the trick | 23:07 |
| guilhermesp | but that's a parameter we use with disk-image-create | 23:07 |
| *** jamesmcarthur has quit IRC | 23:07 | |
| clarkb | guilhermesp: ya dib itself will build in a tmpfs by default if you have enough memory | 23:07 |
| clarkb | this is to reduce disk writes | 23:07 |
| mnaser | guilhermesp: i think that --no-tmpfs is just for dib not to use tmpfs when it builds the image (cause i think that's a speed up) | 23:07 |
| *** jamesmcarthur has joined #zuul | 23:07 | |
| clarkb | yup | 23:07 |
| mnaser | i dont think this is something i can help bake into zuul/zuul-jobs eh? :\ | 23:08 |
| guilhermesp | ah yeah, similar to DIB_NO_TEMPFS, nvm | 23:08 |
| clarkb | mnaser: you probably can as long as the remounting works fine | 23:08 |
| mnaser | clarkb: i think maybe the only best thing to do is create a disable-tmpfs role that i can include in my base jobs which disables tmpfs if tmpfs size <= X | 23:08 |
| clarkb | mnaser: oh except things might need the contents of /tmp after boot | 23:08 |
| mnaser | ah crap, true | 23:08 |
| mnaser | yes, indeed | 23:09 |
| mnaser | i see a bunch of systemd stuff on this cleanly booted instance | 23:09 |
| mnaser | for systemd-logind and dbus-broker so they're certainly _not_ thing id go about forgetting | 23:09 |
| ianw | dib has a ram probe and i think basically doesn't use tmpfs | 23:10 |
| mnaser | ianw: right but we're talking about the final-generated image | 23:11 |
| ianw | as for fedora-32 mounting it tmp ... haven't got to 32 yet sorry :/ | 23:11 |
| mnaser | it sems like its been mounting as tmpfs since fedora 18 lol | 23:12 |
| mnaser | i guess its only an issue because its a 1g memory instance and infra's instances are big enough so that tmpfs is big enough to not cause issues | 23:12 |
| ianw | hrrmm, ok, yeah this is the first i've heard of it | 23:12 |
| guilhermesp | yeah that's basically the thing i guess.... like, fedora-minimal requires >1Gb >.< | 23:13 |
| clarkb | seems like if disabling the unit in the image build fixes it that is an easy enough change | 23:13 |
| mnaser | clarkb: i guess the thing is i was hoping for this to be a runtime option so a 16gb memory instance doesnt suffer because the 1g instance couldnt use tmpfs | 23:14 |
| mnaser | but i cant imagine a clean solution | 23:14 |
| clarkb | I mean how much are they going to suffer though | 23:14 |
| clarkb | its better if performance sensitive things opt into tmpfs instead I think | 23:14 |
| clarkb | we do that for zookeeper's data writes in zuul testing for example | 23:14 |
| ianw | https://www.freedesktop.org/wiki/Software/systemd/APIFileSystems/ does talk about it at the bottom | 23:15 |
| openstackgerrit | Oleksandr Kozachenko proposed zuul/zuul-jobs master: Add addons args in ensure-kubernetes role https://review.opendev.org/731523 | 23:16 |
| *** armstrongs has joined #zuul | 23:19 | |
| *** tosky has quit IRC | 23:20 | |
| mnaser | hmm | 23:21 |
| mnaser | i dont think you can set environment variables for zuul jobs, or can you | 23:21 |
| clarkb | mnaser: you can in the playbooks but not top level I odn't think | 23:22 |
| openstackgerrit | Oleksandr Kozachenko proposed zuul/zuul-jobs master: Add addons args in ensure-kubernetes role https://review.opendev.org/731523 | 23:24 |
| mnaser | hrm, fair enough | 23:24 |
| *** armstrongs has quit IRC | 23:28 | |
| mnaser | clarkb, ianw, guilhermesp: thanks for bouncing ideas around | 23:29 |
| mnaser | i think we'll suggest either using a bigger instance in terms of memory or overriing TMPDIR in the plays | 23:29 |
| *** rfolco|rover has quit IRC | 23:31 | |
| fungi | mnaser: clarkb: late to the party, but you could also create a swapfile, and then tempfs should page out to swap when needed | 23:35 |
| fungi | i merely skimmed, so someone might have already said that | 23:35 |
| fungi | however, if you're likely to be writing to disk anyway, then it's silly to use a tempfs for that at all | 23:36 |
| fungi | i agree | 23:36 |
| mnaser | fungi: yeah, i guess that's merely a workaround. i didn't know tmpfs will write to swapfile if it is full | 23:36 |
| fungi | and super-memory-constrained systems shouldn't be using one for anything more than small shared memory spaces | 23:37 |
| mnaser | fungi: which makes the idea of moving configure-swap role to zuul-jobs and adding an option of running it if system_memory < X interesting | 23:37 |
| mnaser | fungi: in this case its pip using TMPDIR and filling it up | 23:37 |
| clarkb | mnaser: probably doing compiles in there | 23:38 |
| mnaser | also that too | 23:38 |
| clarkb | mnaser: another workaround may be to ensure you have wheels available | 23:38 |
| fungi | right, i meant low-memory systems shouldn't be using tmpfs for things like /tmp | 23:38 |
| fungi | but that they're likely to still want them for stuff like pre-boot /run space and /dev/shm | 23:39 |
| mnaser | fungi: right but i dont think i have an easy way of identifying that in run-time, as i dont want to end up with lowmem and normalmem images | 23:39 |
| clarkb | mnaser: I think you force everything to use disk for temp | 23:39 |
| clarkb | then if something knows it is advantageous to operate on tmpfs it can opt into that | 23:39 |
| clarkb | that was my zookeeper example earlier | 23:39 |
| mnaser | right | 23:39 |
| mnaser | if opendev does that, i'll do it =P | 23:39 |
| fungi | yeah, honestly if you try to do a kernel build in /tmp, even 8gb memory isn't going to be enough | 23:40 |
| mnaser | yeah building wheels is hard because in this case it's hard for me to know all the possible different wheels used by my consumers | 23:40 |
| *** rfolco|rover has joined #zuul | 23:41 | |
| mnaser | would configuring swap increase tmpfs size | 23:41 |
| mnaser | conditionally creating swap seems like it might end up doing the same result, but yeah, it's annoying beacuse i'm trying to find the best possible outcome for all different scenarios, but there's none | 23:41 |
| clarkb | I wouldn't object to a change in our fedora builds to disable that unit | 23:43 |
| *** guillaumec has quit IRC | 23:44 | |
| *** rfolco|rover has quit IRC | 23:45 | |
| fungi | adding swap doesn't dynamically increase tmpfs size, that's usually something you set explicitly at mount time, but i believe you can increase it trivially with a mount -oremount --size=xxxx or whatever and not have to unmount it first | 23:46 |
| *** rfolco|rover has joined #zuul | 23:46 | |
| *** rfolco|rover has quit IRC | 23:50 | |
| *** rfolco|rover has joined #zuul | 23:50 | |
| mnaser | sounds too messy | 23:51 |
| *** rfolco|rover has quit IRC | 23:55 | |
Generated by irclog2html.py 2.17.2 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!