| *** brinzhang_ has quit IRC | 00:13 | |
| *** _mlavalle_1 has quit IRC | 00:43 | |
| *** hamalq has quit IRC | 01:27 | |
| ianw | Process: 5316 ExecStart=/usr/bin/dockerd -H fd:// --containerd=/run/containerd/containerd.sock [0;1;31m(code=exited, status=1/FAILURE)[0m | 01:45 |
|---|---|---|
| ianw | one of thsoe days | 01:45 |
| ianw | https://f50bb2dc448e3365b416-87b3d049f8e641e201153a314e104d20.ssl.cf5.rackcdn.com/786814/3/check/dib-nodepool-functional-openstack-centos-8-stream-src/d50f847/docker/ | 01:46 |
| ianw | how this ended up with one file called "Got" is weird | 01:46 |
| ianw | i wonder if this relates to the ipv6 thing with zuul as well, docker has pushed some change | 01:47 |
| openstackgerrit | Ian Wienand proposed zuul/zuul-jobs master: collect-container-logs: don't copy on failure https://review.opendev.org/c/zuul/zuul-jobs/+/787019 | 02:10 |
| *** elod has joined #opendev | 02:38 | |
| *** ykarel has joined #opendev | 03:33 | |
| ianw | ok, i've seena buch of these docker failures now | 04:02 |
| ianw | nothing in the logs really points to what it is | 04:02 |
| *** vishalmanchanda has joined #opendev | 04:23 | |
| *** gothicserpent has quit IRC | 04:48 | |
| *** zimmerry_ has joined #opendev | 04:48 | |
| *** zimmerry has quit IRC | 04:50 | |
| *** gothicserpent has joined #opendev | 04:53 | |
| *** ysandeep|away is now known as ysandeep | 05:24 | |
| *** marios has joined #opendev | 05:48 | |
| *** slaweq has quit IRC | 05:55 | |
| *** ralonsoh has joined #opendev | 06:03 | |
| *** amoralej|off is now known as amoralej | 06:03 | |
| *** artom has quit IRC | 06:04 | |
| *** artom has joined #opendev | 06:04 | |
| *** sboyron has joined #opendev | 06:09 | |
| *** artom has quit IRC | 06:10 | |
| *** artom has joined #opendev | 06:11 | |
| *** artom has quit IRC | 06:12 | |
| *** artom has joined #opendev | 06:15 | |
| *** artom has quit IRC | 06:16 | |
| *** artom has joined #opendev | 06:16 | |
| *** artom has quit IRC | 06:17 | |
| *** artom has joined #opendev | 06:17 | |
| ianw | ok, i have proposed https://review.opendev.org/c/zuul/nodepool/+/787065 which gets things further by not failing the job when nodepool logs aren't available; but i still can't see why docker is failing to install | 06:24 |
| *** eolivare has joined #opendev | 06:28 | |
| *** brinzhang has joined #opendev | 06:29 | |
| *** lpetrut has joined #opendev | 06:31 | |
| *** artom has quit IRC | 06:37 | |
| *** artom has joined #opendev | 06:37 | |
| *** fressi has quit IRC | 06:42 | |
| *** artom has quit IRC | 06:47 | |
| *** artom has joined #opendev | 06:48 | |
| *** fressi has joined #opendev | 06:56 | |
| *** dmellado has quit IRC | 07:03 | |
| *** andrewbonney has joined #opendev | 07:14 | |
| *** artom has quit IRC | 07:16 | |
| *** artom has joined #opendev | 07:17 | |
| *** artom has quit IRC | 07:18 | |
| *** artom has joined #opendev | 07:18 | |
| *** artom has quit IRC | 07:19 | |
| *** artom has joined #opendev | 07:20 | |
| *** artom has quit IRC | 07:23 | |
| *** artom has joined #opendev | 07:24 | |
| *** rpittau|afk is now known as rpittau | 07:35 | |
| *** slaweq_ is now known as slaweq | 07:36 | |
| openstackgerrit | Daniel Blixt proposed zuul/zuul-jobs master: WIP: Make build-sshkey handling windows compatible https://review.opendev.org/c/zuul/zuul-jobs/+/780662 | 07:44 |
| *** tosky has joined #opendev | 07:45 | |
| openstackgerrit | Daniel Blixt proposed zuul/zuul-jobs master: Make build-sshkey handling windows compatible https://review.opendev.org/c/zuul/zuul-jobs/+/780662 | 07:45 |
| *** jpena|off is now known as jpena | 07:58 | |
| *** mgoddard has quit IRC | 08:08 | |
| *** fressi has quit IRC | 08:24 | |
| *** marios has quit IRC | 08:25 | |
| *** marios has joined #opendev | 08:26 | |
| *** fressi has joined #opendev | 08:29 | |
| *** artom has quit IRC | 08:48 | |
| *** artom has joined #opendev | 08:49 | |
| *** ysandeep is now known as ysandeep|afk | 09:14 | |
| openstackgerrit | Daniel Blixt proposed zuul/zuul-jobs master: Make build-sshkey handling windows compatible https://review.opendev.org/c/zuul/zuul-jobs/+/780662 | 09:23 |
| *** ykarel is now known as ykarel|lunch | 09:40 | |
| *** vishalmanchanda has quit IRC | 09:52 | |
| *** ysandeep|afk is now known as ysandeep | 10:03 | |
| *** rpittau is now known as rpittau|bbl | 10:04 | |
| *** fressi has quit IRC | 10:20 | |
| *** dtantsur|afk is now known as dtantsur | 10:23 | |
| *** fressi has joined #opendev | 10:34 | |
| *** ykarel|lunch is now known as ykarel | 10:47 | |
| *** vishalmanchanda has joined #opendev | 10:49 | |
| *** fressi has quit IRC | 10:49 | |
| *** fressi has joined #opendev | 10:52 | |
| openstackgerrit | Daniel Blixt proposed zuul/zuul-jobs master: Make build-sshkey handling windows compatible https://review.opendev.org/c/zuul/zuul-jobs/+/780662 | 11:09 |
| *** fressi has quit IRC | 11:19 | |
| *** fressi has joined #opendev | 11:22 | |
| *** jpena is now known as jpena|lunch | 11:31 | |
| *** lpetrut has quit IRC | 11:45 | |
| *** artom has quit IRC | 11:48 | |
| *** lpetrut has joined #opendev | 11:55 | |
| *** TiagoPrimini has quit IRC | 11:57 | |
| *** ykarel_ has joined #opendev | 12:20 | |
| *** HeroicHitesh has joined #opendev | 12:21 | |
| *** amoralej is now known as amoralej|lunch | 12:21 | |
| *** ykarel has quit IRC | 12:21 | |
| *** jpena|lunch is now known as jpena | 12:33 | |
| *** rpittau|bbl is now known as rpittau | 12:37 | |
| *** lamt has joined #opendev | 12:40 | |
| *** fressi has quit IRC | 12:44 | |
| *** HeroicHitesh has quit IRC | 12:50 | |
| *** HeroicHitesh has joined #opendev | 13:00 | |
| *** amoralej|lunch is now known as amoralej | 13:05 | |
| *** whoami-rajat_ has joined #opendev | 13:06 | |
| *** HeroicHitesh has quit IRC | 13:10 | |
| *** roman_g has joined #opendev | 13:11 | |
| *** whoami-rajat_ is now known as whoami-rajat | 13:15 | |
| *** HeroicHitesh has joined #opendev | 13:15 | |
| *** ykarel__ has joined #opendev | 13:16 | |
| *** ykarel_ has quit IRC | 13:18 | |
| *** ricolin has joined #opendev | 13:19 | |
| *** rfayan has joined #opendev | 13:36 | |
| *** vishalmanchanda has quit IRC | 13:39 | |
| *** roman_g has quit IRC | 13:45 | |
| *** vishalmanchanda has joined #opendev | 13:47 | |
| *** artom has joined #opendev | 13:49 | |
| *** lpetrut has quit IRC | 14:17 | |
| *** rfayan has quit IRC | 14:37 | |
| *** lassimus_ has joined #opendev | 14:46 | |
| *** lassimus_ has quit IRC | 14:47 | |
| *** lassimus has joined #opendev | 14:47 | |
| clarkb | fungi: ianw: first thought is are any of our production system upgraded to latest docker? and if so do any of them exhibit the same problem? | 14:50 |
| clarkb | I see ianw found the moby issue for failures when ipv6 is disabled on the host. But we shouldn't disable ipv6 on any of our hosts so I doubt that is related | 14:50 |
| clarkb | docker 20.10.6 released just over a week ago | 14:52 |
| clarkb | but these errors didn't start happening until more recently. Is it possible there is a ubuntu kernel + docker + ssh-keyscan interaction? | 14:52 |
| fungi | well the "breaks when ipv6 disabled" bug could still be a clue to a behavior change | 14:59 |
| fungi | if things have suddenly started preferring ipv6 for some subset of situations, and we have baked in assumptions that ipv4 will be tried, i could see running into similar problems | 14:59 |
| *** amoralej is now known as amoralej|off | 15:04 | |
| clarkb | fungi: ianw: looking at syslog for the failed nodepool job I see that docker.io appears to be installed, we then remove it then install docker-ce | 15:08 |
| clarkb | I wonder if docker.io is leaving behind things that break docker-ce | 15:08 |
| fungi | ooh, good catch | 15:10 |
| fungi | do we have competing tasks for installing docker? | 15:11 |
| fungi | is one installed from ubuntu and the other from the external docker package repository? | 15:11 |
| clarkb | yes I think docker.io is ubuntu and docker-ce is the upstream package | 15:17 |
| fungi | makes sense, though the only thing the docker.io package should leave behind is data and configuration. if we explicitly purge the package before installing docker-ce then any lingering data and configuration should also be cleaned up | 15:18 |
| fungi | (the difference between apt remove and apt purge) | 15:19 |
| clarkb | right, but maybe that would delete containers too | 15:19 |
| fungi | yes, it probably would | 15:19 |
| clarkb | I don't understand why docker.io is installed though | 15:19 |
| clarkb | trying to figure that out | 15:19 |
| clarkb | https://zuul.opendev.org/t/zuul/build/4c8186da9fe043bf83d63926cbb3fc84/console#1/0/56/ubuntu-bionic is where we uninstall docker.io | 15:21 |
| clarkb | the next task installs docker-ce | 15:22 |
| clarkb | found it | 15:22 |
| clarkb | it is bindep | 15:22 |
| fungi | hah, so we have docker.io in bindep? | 15:22 |
| fungi | er, in bindep.txt | 15:23 |
| clarkb | yup | 15:23 |
| clarkb | let me push up a change that removes that entry and see if the job is happier | 15:23 |
| *** dmellado has joined #opendev | 15:26 | |
| clarkb | fungi: does uninstall remove unit files or just disable them? the error complains about lack of socket activation stuff | 15:30 |
| clarkb | I wonder if there is an assumed unit behavior in the new package and the old one is leaving a different assumption behind | 15:30 |
| fungi | i think it would remove the old systemd units | 15:32 |
| fungi | but it might leave them in a disabled state instead, i'm not positive | 15:32 |
| fungi | purge should definitely remove them | 15:32 |
| clarkb | looks like my test of focal failed so same behavior there | 15:32 |
| clarkb | I think that may be it because we have't hit any retries yet | 15:41 |
| clarkb | that being the docker.io install prior to docker-ce | 15:41 |
| *** ykarel__ is now known as ykarel | 15:41 | |
| *** dmellado has quit IRC | 15:48 | |
| *** dmellado has joined #opendev | 15:52 | |
| *** ykarel is now known as ykarel|away | 15:56 | |
| clarkb | fungi: fyi https://review.opendev.org/c/opendev/system-config/+/778116 has passed testing and when I checked the screenshots I saw zuul status things. Maybe you want to check it over too and we can start working to land that then your jeepyb changes? | 16:01 |
| *** lpetrut has joined #opendev | 16:07 | |
| fungi | ahh cool, yep | 16:10 |
| *** dmellado has quit IRC | 16:10 | |
| *** dmellado_ has joined #opendev | 16:10 | |
| *** dmellado_ is now known as dmellado | 16:11 | |
| *** lpetrut has quit IRC | 16:20 | |
| *** HeroicHitesh has quit IRC | 16:26 | |
| *** ykarel|away has quit IRC | 16:39 | |
| corvus | clarkb: what wm do you use? | 16:39 |
| clarkb | corvus: xmonad | 16:41 |
| clarkb | most people complain about the config because it is written in haskell but I've always enjoyed that aspect of it :) | 16:41 |
| clarkb | (you also need a ghc toolchain which isn't small) | 16:41 |
| clarkb | i3 has become the more mainstream equivalent. dwm, ratpoison, and awesome are all sort of similar | 16:43 |
| corvus | thx; i think i tried all of them like 8 years ago; it's probably time for a redo :) | 16:43 |
| corvus | i also tried qtile (written/configured in python); it was very new and not quite there yet. but apparently it's still around. | 16:44 |
| clarkb | I would probably start with i3 if you aren't drawn to haskell. I keep meaning to try it too myself | 16:44 |
| *** hamalq has joined #opendev | 16:45 | |
| corvus | arch wiki to the rescue: https://wiki.archlinux.org/index.php/Comparison_of_tiling_window_managers | 16:45 |
| clarkb | corvus: I also use it with xfce. To do that I have to put an xfwm4 in my path ahead of the actual xfwm4 that is actuall xmonad. It works, there is at least one library that makes it work better too (it respects things like panel space) | 16:47 |
| fungi | i still swear by ratpoison, fwiw, i prefer the keybindings | 16:47 |
| clarkb | I wish xfce made it easier to switch out window managers but this has worked ok | 16:47 |
| *** hamalq has quit IRC | 16:47 | |
| clarkb | but that gets me all the nice things from a desktop like battery management, network management, *-agent, etc without doing it all myself | 16:48 |
| corvus | clarkb: oh that's helpful (i also use xfce), thx | 16:48 |
| *** hamalq has joined #opendev | 16:48 | |
| corvus | fungi: okay, i'll put it on the short list too :) | 16:48 |
| corvus | i love that all the tiling managers from 8 years ago are still around; time has done nothing to winnow the field :) | 16:48 |
| fungi | yeah, i don't use any desktop environment or login manager, just ratpoison called from .xinitrc and a basic rp config to precreate some layout i want and spawn terminals | 16:50 |
| fungi | very barebones. i'm not a fan of draggable windows nor window decorations, so all decorationless other than a one-pixel outline around the window with active keyboard focus | 16:50 |
| JayF | Meanwhile, I spent an hour this weekend getting GNOME 40 installed on my gentoo ;) | 16:51 |
| JayF | I do think that if I ever get a linux machine for work, I might look at a more barebones window manager, just to cut down on distractions. But gimme all the eye candy/decorations/etc for my fun PC :D | 16:51 |
| *** fressi has joined #opendev | 16:55 | |
| clarkb | JayF: I found it was particularly helpful on laptops where input devices don't work so well beyond the keyboard | 16:56 |
| JayF | Heh. I haven't run linux on a laptop in years, but there's a pilot program for linux work laptops at my employer, and I'm hoping to get one soon :D | 16:56 |
| corvus | 2021 year of the linux laptop | 16:57 |
| *** rpittau is now known as rpittau|afk | 16:59 | |
| openstackgerrit | Albin Vass proposed zuul/zuul-jobs master: WIP: Roles to snapshot and cleanup image builds for digitalocean https://review.opendev.org/c/zuul/zuul-jobs/+/786757 | 17:02 |
| *** jpena is now known as jpena|off | 17:08 | |
| fungi | one of the reasons i like keyboard-driven window management is that i have some graphical systems where a keyboard (or chording keyer) is the only input device, so the keynav utility winds up standing in as a pointer only when a pointer is absolutely required | 17:11 |
| *** marios is now known as marios|out | 17:11 | |
| fungi | though on that note, keynav *is* a really cool tool | 17:12 |
| *** marios|out has quit IRC | 17:13 | |
| JayF | Interesting. My only graphical linux system is the desktop I use (and dual boot with windows), all my other boxes are console only and managed over ssh only. | 17:15 |
| openstackgerrit | Gonéri Le Bouder proposed zuul/zuul-jobs master: ensure-docker: ensure docker.socket is stopped https://review.opendev.org/c/zuul/zuul-jobs/+/787271 | 17:19 |
| *** ysandeep is now known as ysandeep|away | 17:20 | |
| *** eolivare has quit IRC | 17:22 | |
| elod | fungi clarkb : thanks for the merge (meta-config patch)! I'll drop a mail to ML about starting the delete of ocata-eol tagged branches and after some time (~1 week) I'll do some "live" testing on some branch (whose repository's acl was changed in the patch) | 17:28 |
| elod | does this sound OK for you? ^^^ | 17:28 |
| clarkb | ya I think we were hoping for some live testing before we started merging more of those changes | 17:29 |
| clarkb | just to avoid any subsequent updates if they become necessary | 17:29 |
| fungi | elod: yeah, should be fine. i suggest starting slowly and double-checking things worked the way you expect. then if there are no complaints we can work on updating the other acls too | 17:29 |
| elod | sure, cool | 17:29 |
| elod | thanks! | 17:30 |
| *** ralonsoh has quit IRC | 17:31 | |
| elod | JayF: some warning in advance ;) ^^^ | 17:31 |
| JayF | elod: Thank you :) Courtesy of our previous discussion we now have a simple runbook to follow to reconfigure CI if a retired branch breaks it | 17:32 |
| elod | JayF: ++ | 17:32 |
| smcginnis | Probably doesn't have an impact on us, but just so folks know - https://grafana.com/blog/2021/04/20/grafana-loki-tempo-relicensing-to-agplv3/ | 17:35 |
| clarkb | smcginnis: should be fine, we deploy the upstream grafana images so we dn't even have modifications to publish (which we would sort of do automatically anyway) | 17:36 |
| *** brinzhang has quit IRC | 17:49 | |
| *** dtantsur is now known as dtantsur|afk | 17:52 | |
| *** klonn has joined #opendev | 17:54 | |
| *** fressi has quit IRC | 18:00 | |
| *** iurygregory has quit IRC | 18:15 | |
| *** sboyron has quit IRC | 18:24 | |
| *** hrw has quit IRC | 18:25 | |
| *** hrw has joined #opendev | 18:27 | |
| *** iurygregory has joined #opendev | 18:33 | |
| openstackgerrit | Albin Vass proposed zuul/zuul-jobs master: WIP: Roles to snapshot and cleanup image builds for digitalocean https://review.opendev.org/c/zuul/zuul-jobs/+/786757 | 18:35 |
| *** andrewbonney has quit IRC | 18:40 | |
| *** mlavalle has joined #opendev | 19:07 | |
| *** vishalmanchanda has quit IRC | 19:09 | |
| openstackgerrit | Matt McEuen proposed openstack/project-config master: Add ACL rule for deleting treasuremap branches https://review.opendev.org/c/openstack/project-config/+/787282 | 19:10 |
| openstackgerrit | Gonéri Le Bouder proposed zuul/zuul-jobs master: ensure-docker: ensure docker.socket is stopped https://review.opendev.org/c/zuul/zuul-jobs/+/787271 | 19:15 |
| *** slaweq has quit IRC | 19:16 | |
| *** slaweq has joined #opendev | 19:18 | |
| *** lbragstad has quit IRC | 19:26 | |
| openstackgerrit | Merged openstack/project-config master: Add ACL rule for deleting treasuremap branches https://review.opendev.org/c/openstack/project-config/+/787282 | 19:29 |
| zigo | fungi: Hi there! I saw that https://review.opendev.org/c/openstack/project-config/+/783613 was merged, does it mean I already can use Bullseye ? | 19:39 |
| *** slaweq has quit IRC | 19:42 | |
| fungi | zigo: turns out we merged it prematurely, we should have waited for a new dib release with the dependency of that change included, and then an updated nodepool container image with that dib release included. ianw went ahead and tagged a new dib version yesterday, but found a new problem with docker which is blocking us from merging the subsequent change to update the nodepool image | 19:43 |
| zigo | fungi: Though is the dib change for Bullseye ok? The docker thingy is unrelated, right? | 19:44 |
| fungi | we think we've finally got a handle on what's changed in docker | 19:44 |
| fungi | yeah, unrelated, just blocking us from upgrading dib on our nodepool builders | 19:44 |
| fungi | so we won't be able to build proper bullseye node images until that's fixed, hopefully in the next few hours | 19:44 |
| zigo | fungi: Allright, good news then, tomorrow should be ok for me... :P | 19:45 |
| fungi | looking that way, yes | 19:45 |
| zigo | fungi: I still got that https://review.opendev.org/c/openstack/puppet-openstack_extras/+/786503 patch to merge before I can try a WIP patch on puppet-openstack-integration, so I have to wait anyways... | 19:46 |
| fungi | there's still an outstanding change we need to merge after we have nodes to start building amd64 and arm64 python wheels for bullseye, but that's just a performance booster for pip install of things on those platforms | 19:46 |
| zigo | fungi: Did you hear about extrepo? It's great, it's kind of like PPAs for Debian, kind of ... | 19:46 |
| zigo | It's a completely independent initiative from a single DD though ... | 19:47 |
| zigo | PPAs are hosted by canonical, extrepo is just a list of available (but authenticated) random Debian repos hosted wherever one wishes... | 19:48 |
| zigo | The good thing is about auth: Extrepo is taking care of the GPG stuff, so you don't need to trust any random GPG key on the net. | 19:49 |
| zigo | As for me: | 19:49 |
| zigo | extrepo openstack_wallaby | 19:49 |
| zigo | DONE ! :) | 19:49 |
| fungi | very cool! | 19:49 |
| zigo | extrepo enable openstack_wallaby | 19:49 |
| zigo | (that is) | 19:49 |
| *** lbragstad has joined #opendev | 19:49 | |
| fungi | we're not running any debian servers currently, but i bet that could be useful in some of the roles in zuul-jobs | 19:49 |
| zigo | https://salsa.debian.org/extrepo-team/extrepo-data/-/tree/master/repos/debian <--- List of external repositories. You just need a merge request to add yourself. | 19:50 |
| openstackgerrit | Gonéri Le Bouder proposed zuul/zuul-jobs master: ensure-docker: ensure docker.socket is stopped https://review.opendev.org/c/zuul/zuul-jobs/+/787271 | 19:50 |
| fungi | thankfully we know better than to fetch signing keys from keyservers to configure package repositories anyway, and just bake the key material into our orchestration | 19:50 |
| zigo | Yeah... | 19:51 |
| zigo | wget https://example.com/key.asc -O - | apt-key add <---- To me, this is the equivalent of "curl | sudo bash" :) | 19:51 |
| fungi | yep | 19:52 |
| fungi | not just from a security perspective, but also because it's additional network access which increases the chances a job will break on random connection failures | 19:52 |
| openstackgerrit | Merged opendev/system-config master: Remove Jenkins related documentation fragments https://review.opendev.org/c/opendev/system-config/+/786796 | 19:57 |
| corvus | fungi, clarkb, mordred: https://fishshell.com/ i have not used the shell, but the web page is great | 20:13 |
| slittle1 | https://review.opendev.org/c/starlingx/integ/+/785891 seems to be stuck | 20:14 |
| slittle1 | neither recheck nor WF 0/+1 have worked | 20:15 |
| corvus | slittle1: can you elaborate on 'stuck'? do you mean zuul has not enqueued the change into gate? | 20:16 |
| mordred | corvus: it's written in reasonably modern C++ | 20:16 |
| mordred | corvus: https://github.com/fish-shell/fish-shell/blob/master/src/builtin_commandline.cpp#L127 | 20:16 |
| fungi | slittle1: its parent is outdated, so the change cannot merge in its current state. gerrit doesn't do a great job of indicating that | 20:17 |
| corvus | slittle1: it looks like that patch may be based on an outdated version of its parent | 20:17 |
| corvus | yeah that :) | 20:17 |
| corvus | a rebase on current branch tip should fix | 20:17 |
| corvus | mordred: neat :) | 20:18 |
| fungi | slittle1: look at the relation chain, and if you follow the link to the "merged" parent you'll see it takes you to patchset 1 of 785814 when patchset 2 is what merged | 20:18 |
| fungi | basically the git parent of 785891 is a commit which will never appear in that branch, so the commit can't be attached to the target branch without a rebase | 20:19 |
| *** lbragstad has quit IRC | 20:20 | |
| slittle1 | So shouldn't 785891 be marked as merge-conflict or needs-rebase or something ? | 20:30 |
| fungi | slittle1: the old gerrit ui used to show an indicator that the change had an "outdated" parent | 20:33 |
| fungi | it's not evident in the new ui where it displays that, if it does | 20:34 |
| fungi | i agree it's a bit confusing, i don't know if it's on the gerrit developers' radar as something which needs to be improved | 20:34 |
| fungi | zuul doesn't even attempt to evaluate the change, because the gerrit api indicates the change cannot merge | 20:35 |
| *** slaweq has joined #opendev | 20:35 | |
| *** lbragstad has joined #opendev | 20:36 | |
| fungi | the way developers usually avoid that in a dependent series of commits is to edit earlier commits with an interactive rebase, so that the commits following it will get updated accordingly, and then the whole series re-pushed (at which point gerrit will create new patchsets for any of the commits in that series which have been updated) | 20:38 |
| fungi | i find https://pypi.org/project/git-restack/ very useful for such cases | 20:39 |
| fungi | the other way i've seen it dealt with is to fetch the change you want to edit, and then cherry-pick the subsequent changes from gerrit in order onto it, before pushing | 20:41 |
| fungi | git-review's -x option is useful for that method | 20:41 |
| openstackgerrit | Gonéri Le Bouder proposed zuul/zuul-jobs master: ensure-docker: ensure docker.socket is stopped https://review.opendev.org/c/zuul/zuul-jobs/+/787271 | 20:44 |
| *** slaweq has quit IRC | 20:52 | |
| fungi | slittle1: i'm actually somewhat confused, since https://review.opendev.org/Documentation/user-review-ui.html#related-changes-tab says (and shows a screenshot for) "If an ancestor change is marked with an orange dot it means that the currently viewed patch set depends on a outdated patch set of the ancestor change." | 20:55 |
| fungi | i wonder if something is preventing those icons from displaying in our current theme | 20:55 |
| corvus | fungi: maybe that changes when ancestor is merged? | 20:55 |
| fungi | oh, that might | 20:56 |
| fungi | be it | 20:56 |
| fungi | (merged) was showing for it where i would normally have expected the icon to be displayed instead | 20:57 |
| fungi | so perhaps it reuses the same field | 20:57 |
| corvus | fungi: merged is orange; is that different? | 20:57 |
| fungi | yeah, perhaps that's what they were trying to convey. not very colorblindness-friendly | 20:57 |
| fungi | maybe it would have also displayed a tooltip if i'd thought to try hovering over it | 21:00 |
| *** odyssey4me has quit IRC | 21:05 | |
| *** odyssey4me has joined #opendev | 21:06 | |
| ianw | fungi / zigo : hopefully i can babysit the nodepool change in today if docker is behaving :) | 21:30 |
| *** roman_g has joined #opendev | 21:41 | |
| openstackgerrit | Monty Taylor proposed zuul/zuul-jobs master: Remove use of --generator=run-pod/v1 https://review.opendev.org/c/zuul/zuul-jobs/+/787291 | 21:43 |
| openstackgerrit | Monty Taylor proposed zuul/zuul-jobs master: Remove use of --generator=run-pod/v1 https://review.opendev.org/c/zuul/zuul-jobs/+/787291 | 21:43 |
| *** whoami-rajat has quit IRC | 21:46 | |
| *** roman_g has quit IRC | 21:54 | |
| openstackgerrit | Ian Wienand proposed opendev/system-config master: OSU OSL : change upload format to RAW https://review.opendev.org/c/opendev/system-config/+/787293 | 21:58 |
| openstackgerrit | Gonéri Le Bouder proposed zuul/zuul-jobs master: ensure-docker: ensure docker.socket is stopped https://review.opendev.org/c/zuul/zuul-jobs/+/787271 | 22:00 |
| openstackgerrit | Gonéri Le Bouder proposed zuul/zuul-jobs master: ensure-docker: ensure docker.socket is stopped https://review.opendev.org/c/zuul/zuul-jobs/+/787271 | 22:04 |
| *** bodgix has quit IRC | 22:07 | |
| *** bodgix_ has joined #opendev | 22:07 | |
| ianw | Ramereth asked us to upload raw to OSU OSL, which i proposed, but also pointed out https://github.com/osuosl/packer-templates/blob/master/bin/upload.sh#L48-L56 | 22:10 |
| ianw | --property hw_qemu_guest_agent=yes \ | 22:10 |
| ianw | --property os_require_quiesce=yes | 22:10 |
| ianw | i don't think we run the qemu guest agent | 22:11 |
| clarkb | I don't think our images have a qemu guest agent? | 22:11 |
| ianw | and then the quiesce argument i think is related, in that if snapshotting, migrating etc. it sends a quiesce command via the guest agent | 22:11 |
| ianw | anyway, i'm not sure they're important arguments for us, but just something pointed out | 22:12 |
| ianw | raw images should put much less load on their ceph | 22:13 |
| ianw | kevinz: ^ not sure if this is something that would be helpful in linaro as well; not something i'd thought of before. we just default to .qcow2 but doesn't have to be | 22:13 |
| *** DSpider has joined #opendev | 22:15 | |
| clarkb | ianw: one thing to keep in mind is that this will likely double the disk space needs for our builder | 22:16 |
| clarkb | if we can do raw for linaro too that may be a good thing to keep builder disk needs down | 22:17 |
| clarkb | (we do fstrim the raw images so they end up roughly equivalent to qcow2 in disk use iirc) | 22:17 |
| mordred | clarkb, ianw : I agree - we do not do the qemu guest agent in our instances | 22:21 |
| mordred | and that's quite on purpose - we'd rather the instances just get killed and we can make new ones. having a provider try to "nicely" manage test nodes would be rather yuck | 22:22 |
| mordred | which is to say - yeah, we should not supply either of those parameters | 22:24 |
| openstackgerrit | Monty Taylor proposed zuul/zuul-jobs master: Remove use of --generator=run-pod/v1 https://review.opendev.org/c/zuul/zuul-jobs/+/787291 | 22:30 |
| openstackgerrit | Clark Boylan proposed opendev/git-review master: Install PBR explicitly https://review.opendev.org/c/opendev/git-review/+/787297 | 22:46 |
| clarkb | I suspect we're going to need a story for solving ^ soon. Going to see if this works | 22:46 |
| clarkb | note that is preventing the --no-thin patch from landing | 22:46 |
| openstackgerrit | Steve Baker proposed openstack/diskimage-builder master: Ensure redhat efi packages are reinstalled during finalise https://review.opendev.org/c/openstack/diskimage-builder/+/786804 | 22:58 |
| openstackgerrit | Clark Boylan proposed opendev/git-review master: Install PBR explicitly https://review.opendev.org/c/opendev/git-review/+/787297 | 23:15 |
| clarkb | fungi: ^ thta might make packagers have a fit, but I think that is a reasonable workaround | 23:15 |
| ianw | "Unexpected API Error. Please report this at http://bugs.launchpad.net/nova/ and attach the Nova API log if possible." | 23:16 |
| ianw | a new failure for the nodepool func job :/ | 23:17 |
| *** tosky has quit IRC | 23:17 | |
| clarkb | fungi: looks like walmart has responded. I'm not sure what the proper fix is on the osa side? is it to pre clone then tell osa to fetch from there? | 23:34 |
| clarkb | I don't want to tell them the wrong thing | 23:35 |
| openstackgerrit | Merged openstack/project-config master: nodepool: make ARM64 config names consistent https://review.opendev.org/c/openstack/project-config/+/786357 | 23:38 |
| openstackgerrit | Merged opendev/system-config master: OSU OSL : change upload format to RAW https://review.opendev.org/c/opendev/system-config/+/787293 | 23:38 |
| fungi | clarkb: the last one to have this problem got help in #openstack-ansible... i suggested they follow up there on on the openstack-discuss ml | 23:39 |
| fungi | but if they really want to have us play proxy for a discussion with the ansible team, i guess we can try to get suggestions for them? | 23:39 |
| clarkb | well I'm not sure they want us to play proxy, they just responded to the email you sent them a week ago | 23:40 |
| openstackgerrit | Tristan Cacqueray proposed zuul/zuul-jobs master: Remove use of --generator=run-pod/v1 for oc https://review.opendev.org/c/zuul/zuul-jobs/+/787300 | 23:41 |
| clarkb | oh maybe it isn't a response they are just reaching out direclt | 23:41 |
| clarkb | y | 23:41 |
| clarkb | I wonder why they used incident | 23:41 |
| clarkb | maybe we aren't clear enough about hte prupose of that list | 23:41 |
| clarkb | fungi: maybe you want to respond to them with the details you already sent to them :/ | 23:41 |
| fungi | i cc'd service-incident on the message i sent them about blocking their access, because i didn't want to broadcast their ip address info or otherwise publicly shame them | 23:41 |
| fungi | so that's probably why they replied there | 23:42 |
| *** DSpider has quit IRC | 23:44 | |
| clarkb | I'm tryingto find that email | 23:45 |
| openstackgerrit | Monty Taylor proposed zuul/zuul-jobs master: ensure-docker: ensure docker.socket is stopped https://review.opendev.org/c/zuul/zuul-jobs/+/787271 | 23:45 |
| clarkb | but searching walmart doesn't seem tp bring it up | 23:45 |
| fungi | i sent it to "Chris Undernehr" | 23:46 |
| fungi | at the recommendation of some foundation account reps | 23:47 |
| clarkb | aha that found it | 23:47 |
| clarkb | ya I think your initial email seems like a reasonable response | 23:48 |
| fungi | and yeah looking at the message they sent us, i have a feeling they never saw what i sent to their department head last thursday | 23:48 |
| clarkb | so I guess let the incident email through then respond with what you sent previously? Do you want to do that or should I? | 23:48 |
| fungi | i can get to it shortly but have to switch rooms/computers | 23:49 |
| clarkb | cool, thanks (I don't think its a huge rush since we did reach out on thursday, but getting a response to them should hopeflly help them resolve things) | 23:49 |
| fungi | yep | 23:50 |
Generated by irclog2html.py 2.17.2 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!