| opendevreview | Ian Wienand proposed openstack/diskimage-builder master: Remove CentOS-8 jobs https://review.opendev.org/c/openstack/diskimage-builder/+/827196 | 00:01 |
|---|---|---|
| ianw | argh, clean still has it | 00:02 |
| ianw | glean even | 00:02 |
| *** dviroel|brb is now known as dviroel | 00:03 | |
| fungi | oh, yep | 00:04 |
| opendevreview | Ian Wienand proposed opendev/glean master: testing: drop centos-8, add 8-stream and 9-stream https://review.opendev.org/c/opendev/glean/+/827197 | 00:04 |
| opendevreview | Ian Wienand proposed openstack/diskimage-builder master: Remove CentOS-8 jobs https://review.opendev.org/c/openstack/diskimage-builder/+/827196 | 00:05 |
| *** dviroel is now known as dviroel_ | 00:19 | |
| *** dviroel_ is now known as dviroel | 00:19 | |
| ianw | hrm one of the nodepool container builds failed with WARNING: Retrying (Retry(total=4, connect=None, read=None, redirect=None, status=None)) after connection broken by 'NewConnectionError('<pip._vendor.urllib3.connection.HTTPSConnection object at 0x7fae54241fa0>: Failed to establish a new connection: [Errno 113] No route to host')': /simple/yappi/ | 00:29 |
| ianw | i'm guessing this came from the mirror | 00:29 |
| ianw | cloud: inmotion | 00:30 |
| ianw | i'll keep an eye ... it's going to be one of those days | 00:33 |
| *** rlandy|ruck is now known as rlandy|out | 00:44 | |
| clarkb | before we restart gerrit we should preserve the dump file that corvus generated earlier today | 01:02 |
| clarkb | luca is asking for it on the issue I filed. I don't think we should upload it immediately as I'm not sure what is included in that (would it have anything private?) but if we restart on new container it will be lost | 01:02 |
| clarkb | I think the file was at /tmp/dump in the container itself | 01:03 |
| clarkb | my ssh keys have expired tehmselves for the day and I need toget back to dinner | 01:04 |
| clarkb | but maybe someone can copy that off the container really quickly and stash it somewhere on the host | 01:04 |
| opendevreview | Merged openstack/project-config master: Stop building CentOS 8 https://review.opendev.org/c/openstack/project-config/+/827195 | 01:06 |
| clarkb | hrm I went ahead and loaded a key let see fi I can do this real quick | 01:07 |
| *** cloudnull2 is now known as cloudnull | 01:07 | |
| clarkb | ok thats done review02:~clarkb/dump has the file | 01:10 |
| clarkb | I think we can restart whenever is convenient now | 01:10 |
| fungi | thanks! | 01:10 |
| fungi | (now get back to dinner) | 01:10 |
| clarkb | yes I'm told I need to eat now. :) | 01:11 |
| fungi | i'm familiar with this | 01:11 |
| ianw | excellent. i'm happy to restart in about 2-3 hours, that seems a nice quiet time | 01:18 |
| opendevreview | Merged opendev/system-config master: grafana: update docs and make an import script https://review.opendev.org/c/opendev/system-config/+/826241 | 01:25 |
| fungi | ianw: i may still be around at that time if problems arise | 01:33 |
| ianw | ++ heh, hopefully neither :) | 01:35 |
| *** ysandeep is now known as ysandeep|away | 01:44 | |
| *** dviroel is now known as dviroel|out | 01:47 | |
| opendevreview | Merged opendev/glean master: testing: drop centos-8, add 8-stream and 9-stream https://review.opendev.org/c/opendev/glean/+/827197 | 02:28 |
| ianw | planning to do gerrit restart in about 10 minutes | 03:53 |
| ianw | "RepoDigests": [ | 04:06 |
| ianw | "opendevorg/gerrit@sha256:8776f6539f47917683d4e6763fa4cf3aee8da476166f44434def9bfb6940e611" | 04:06 |
| ianw | is what i just pulled, which matches https://hub.docker.com/layers/opendevorg/gerrit/3.4/images/sha256-8776f6539f47917683d4e6763fa4cf3aee8da476166f44434def9bfb6940e611?context=explore | 04:06 |
| ianw | which was pushed by 827153 | 04:06 |
| ianw | #status log restarted gerrit to get changes from 827153 | 04:11 |
| opendevstatus | ianw: finished logging | 04:11 |
| opendevreview | Merged openstack/diskimage-builder master: Remove CentOS-8 jobs https://review.opendev.org/c/openstack/diskimage-builder/+/827196 | 04:34 |
| opendevreview | Ian Wienand proposed openstack/diskimage-builder master: Cleanup more CentOS 8 bits https://review.opendev.org/c/openstack/diskimage-builder/+/827210 | 04:44 |
| opendevreview | Ian Wienand proposed openstack/diskimage-builder master: Remove contrib/setup-gate-mirrors.sh https://review.opendev.org/c/openstack/diskimage-builder/+/827211 | 04:44 |
| prometheanfire | new ssh? | 05:20 |
| prometheanfire | kex mismatch | 05:20 |
| ianw | prometheanfire: umm, none of that should have chagned | 05:37 |
| ianw | i have pushed changes via ssh since the update and didn't notice anything out of order | 05:39 |
| ianw | https://zuul.opendev.org/t/zuul/build/1ddea9a38a6641dbbc7c809bd4f1b008/log/zuul-info/inventory.yaml also failed in inmotion, a few hours ago though | 05:44 |
| ianw | i can't see anything wrong with the mirror, i just logged in an it's active and serving responses | 05:50 |
| ianw | might just be a glitch in the matrix | 05:50 |
| * prometheanfire shrugs | 05:51 | |
| *** marios is now known as marios|ruck | 06:04 | |
| *** ysandeep|away is now known as ysandeep | 07:04 | |
| *** mrunge_ is now known as mrunge | 07:35 | |
| opendevreview | Merged openstack/diskimage-builder master: centos: do not use $releasever in .repo files https://review.opendev.org/c/openstack/diskimage-builder/+/826244 | 07:43 |
| opendevreview | Merged openstack/diskimage-builder master: Switch 9-stream testing to use opendev mirrors https://review.opendev.org/c/openstack/diskimage-builder/+/821651 | 07:43 |
| opendevreview | Merged openstack/diskimage-builder master: Add 9-stream ARM64 testing https://review.opendev.org/c/openstack/diskimage-builder/+/821653 | 07:44 |
| *** amoralej|off is now known as amoralej | 07:52 | |
| *** jpena|off is now known as jpena | 08:36 | |
| amoralej | according to https://static.opendev.org/mirror/logs/rsync-mirrors/ rsync for centos-8-stream is not longer running since yesterday 06:47? | 08:55 |
| ianw | amoralej: ahh, thanks, i think i accidentaly still had a lock held from when i bumped its quota yesterday | 09:24 |
| opendevreview | Merged openstack/project-config master: Set centos-8 min-ready to 0 https://review.opendev.org/c/openstack/project-config/+/827183 | 09:28 |
| rpittau | hello everyone, we found a little issue with dib and debian bullseye, after building an image with the debian element it seems the network interfaces are not configured during boot :/ | 10:27 |
| rpittau | from what I can see the /usr/local/sbin/dhcp-all-interfaces.sh script is not executed at startup | 10:27 |
| *** marios|ruck is now known as marios | 10:41 | |
| *** marios is now known as marios|ruck | 10:41 | |
| amoralej | ianw, good, so you are removing the lock? | 10:41 |
| opendevreview | Alfredo Moralejo proposed zuul/zuul-jobs master: Change RDO train repository for Centos 8 stream https://review.opendev.org/c/zuul/zuul-jobs/+/827067 | 10:42 |
| *** rlandy is now known as rlandy|ruck | 11:14 | |
| *** dviroel|out is now known as dviroel | 11:27 | |
| dviroel | marios|ruck: ysandeep: please see https://review.rdoproject.org/r/c/rdo-jobs/+/38634/2#message-d942642017ca1572f4e0c4c6eb154d9edef7eb0c | 11:45 |
| dviroel | marios|ruck: wrong channel :P | 11:47 |
| marios|ruck | dviroel: ;) | 11:53 |
| *** ysandeep is now known as ysandeep|break | 12:04 | |
| *** ysandeep|break is now known as ysandeep | 12:39 | |
| *** rlandy|ruck is now known as rlandy|ruck|mtg | 13:04 | |
| *** sshnaidm is now known as sshnaidm|afk | 13:06 | |
| *** amoralej is now known as amoralej|lunch | 13:13 | |
| fungi | rpittau: is this one you built yourself, or you mean our debian-bullseye job nodes are no longer working? | 13:35 |
| rpittau | fungi: hi! This is one ipa ramdisk built using the debian DIB module | 13:35 |
| rpittau | we change to bullseye in december and since then we were not able to use dhcp on the images built with DIB | 13:37 |
| fungi | is that with or without glean? | 13:38 |
| fungi | also, you're aware there's an #openstack-dib channel, right? you might get better help there | 13:38 |
| rpittau | fungi: I know about that channel, I sent a message there too more than 1 month ago but I didn't get any answer | 13:39 |
| fungi | ahh, okay | 13:39 |
| fungi | so... the bullseye images we're booting use glean, maybe that's why we're able to use dhcp? | 13:40 |
| rpittau | we're building the image in the same way as we did with buster, the precise command is: disk-image-create -o /var/lib/ironic/httpboot/ipa debian ironic-python-agent-ramdisk | 13:41 |
| rpittau | fungi: that could be, are they built using a different DIB element ? | 13:42 |
| fungi | we build on the debian-minimal element, setting release: bullseye and then we add these elements: vm, simple-init, growroot (and some others which probably aren't relevant in this case) | 13:49 |
| fungi | rpittau: are you using the simple-init element? | 13:49 |
| fungi | i think that's the one responsible for our network setup | 13:49 |
| fungi | yeah, looking at its readme, that's the one that handles findinging a configdrive, calling glean, and so on | 13:50 |
| fungi | it's possible that the debian-minimal element on its own isn't complete with regards to network setup and nobody's noticed until now | 13:51 |
| rpittau | fungi: we use debian element that includes debian-minimal, according to the docs, but I don't see simple-init | 13:51 |
| rpittau | I could give that a try | 13:51 |
| fungi | worth a shot, but i have no idea whether it will be well-suited to a ramdisk boot since that's going to be different from a cloud vm in a few ways | 13:52 |
| rpittau | fungi: thanks, I'll work on that and see :) | 13:53 |
| fungi | anyway, if the buster images were booting on their own and bullseye aren't, then it's probably a bug which will need solving in the debian(-minimal) element in dib | 13:53 |
| rpittau | actually, now that I see the full command, we add simple-int in the end, so it doesn't look like it works :/ | 13:54 |
| *** sshnaidm|afk is now known as sshnaidm | 13:54 | |
| fungi | interesting. in that case if you can get the boot logs from glean on a working buster and broken bullseye image, maybe we could try to spot the difference | 13:55 |
| rpittau | one thing that I noticed is the network interfaces naming difference, on bullseye we don't have the old naming anymore (e.g. eth0) | 13:55 |
| rpittau | and DIB_NETWORK_INTERFACE_NAMES is set to eth0 eth1 | 13:56 |
| fungi | yeah, debian switched to bus-based interface numbering a while back. i mostly use sid though so am not sure which release it happened in | 13:57 |
| rpittau | seems bullseye, in buster I can still see the old naming | 13:57 |
| fungi | it may be possible to get them back with net.ifnames=0 | 13:59 |
| rpittau | that's maybe worth a try, can I add that option when building an image with DIB ? | 14:00 |
| fungi | DIB_BOOTLOADER_DEFAULT_CMDLINE envvar maybe? | 14:02 |
| rpittau | fungi: thanks, I have some to work on at least :) | 14:03 |
| fungi | https://opendev.org/openstack/kayobe/src/branch/master/releasenotes/notes/dib-consistent-network-device-naming-e0edec5d5fdf83e2.yaml | 14:03 |
| fungi | codesearch for net.ifnames turns up a number of projects fiddling with it in one way or another | 14:03 |
| rpittau | I'll have a look there too | 14:04 |
| fungi | also often coupled with biosdevname=0 | 14:04 |
| *** rlandy|ruck|mtg is now known as rlandy|ruck | 14:05 | |
| rpittau | I've also enabled DIB debugging message, let's see if we get some more info | 14:08 |
| *** amoralej|lunch is now known as amoralej | 14:13 | |
| opendevreview | dasm proposed zuul/zuul-jobs master: Pull correct upstream pip version https://review.opendev.org/c/zuul/zuul-jobs/+/827297 | 14:22 |
| lourot | o/ has anyone got time to land https://review.opendev.org/c/openstack/project-config/+/825089 ? thanks! | 14:38 |
| lourot | thanks fungi! | 14:40 |
| fungi | yw | 14:42 |
| *** bhagyashris_ is now known as bhagyashris | 14:42 | |
| opendevreview | Merged openstack/project-config master: Mirror charm-nova-compute-nvidia-vgpu to GitHub https://review.opendev.org/c/openstack/project-config/+/825089 | 14:47 |
| clarkb | fungi: rpittau: glean really expects to have a config drive to be able to do its job. I want to say the way ironic does things with its images is does dhcp all interfaces (which also requires you to know the interface names iirc) and then also configures glean. This works beacuse dhcp all interfaces writes out /e/n/i configs at image build time then if you haev a config drive present | 14:55 |
| clarkb | glean will write over the top of those on boot | 14:55 |
| clarkb | But glean won't do much of anything without a config drive iirc | 14:55 |
| rpittau | clarkb: thanks, the problem happened when swithcing from buster to bullseye and one main differnece is that the network interfaces names changed, so I'm trying now with what fungi suggested to switch to the old names | 15:00 |
| *** artom__ is now known as artom | 15:16 | |
| *** ysandeep is now known as ysandeep|dinner | 15:19 | |
| opendevreview | Aurelien Lourot proposed openstack/project-config master: Add dummy charm project https://review.opendev.org/c/openstack/project-config/+/827307 | 15:31 |
| clarkb | fungi: any reason to not recheck (and potentially land) https://review.opendev.org/c/opendev/system-config/+/825338 today? Then we can test the gitea links change further down the stack | 15:32 |
| clarkb | I mean it should be a complete noop but it udpates gerrit config minorly and is worth monitoring I suppose | 15:32 |
| opendevreview | Tristan Cacqueray proposed zuul/zuul-jobs master: Add packages support to ensure-cargo https://review.opendev.org/c/zuul/zuul-jobs/+/827130 | 15:39 |
| *** dviroel is now known as dviroel|lunch | 15:40 | |
| *** ysandeep|dinner is now known as ysandeep | 15:45 | |
| opendevreview | Aurelien Lourot proposed openstack/project-config master: Add dummy charm project https://review.opendev.org/c/openstack/project-config/+/827307 | 15:47 |
| lourot | clarkb, fungi ^ how exactly should I specify that the default branch is `main` instead of `master`? Should it be done directly below the `upstream: ...` line? | 15:48 |
| clarkb | lourot: its a dictionary so order of the keys doesn't matter. Just needs to be in there | 15:48 |
| clarkb | looks like the key is default-branch | 15:49 |
| lourot | nice thanks, giving a try at `default-branch: main` | 15:49 |
| fungi | cool! | 15:50 |
| opendevreview | Aurelien Lourot proposed openstack/project-config master: Add dummy charm project https://review.opendev.org/c/openstack/project-config/+/827307 | 15:50 |
| fungi | clarkb: no objection, i've rechecked 825338 now | 15:50 |
| clarkb | https://opendev.org/opendev/jeepyb/src/branch/master/jeepyb/cmd/manage_projects.py#L488 and https://opendev.org/opendev/system-config/src/branch/master/playbooks/roles/gitea-git-repos/library/gitea_create_repos.py#L132 are the main points where we handle that | 15:51 |
| clarkb | you'll also need to set a zuul setting somewhere that I don't remember to have zuul treat main as the fallback branch for branch resolution stuff | 15:51 |
| fungi | yeah, we can work that out once the repo creation is done, i expect | 15:55 |
| clarkb | this just occured to me, but we may need to test renaming of a gerrit project whose branch isn't master | 15:55 |
| clarkb | Not necessary to create a new project that way. But if we get to renaming projects that aren't default master I wonder if renaming them in gerrit will try to switch HEAD to master | 15:56 |
| clarkb | but we can take this one step at a time and learn as we go. No need to solve that before we do a new project | 15:56 |
| fungi | yep. some of that we might be able to just do in testinfra too | 15:58 |
| clarkb | ++ | 16:00 |
| clarkb | fungi: ianw: so now that gerrit updated we can test the signed tag pushing fix? | 16:13 |
| clarkb | And once we are happy with it here I'll push the merge of that fix into gerrit 3.5 | 16:14 |
| fungi | yeah. we can simply approve it and if any new uses of createTag have snuck in the change will fail on removing it from the allowed list in the linter script | 16:14 |
| clarkb | and warn the openstack release team to do a single release first? | 16:15 |
| fungi | right, once it merges | 16:15 |
| fungi | and deploys | 16:15 |
| clarkb | https://review.opendev.org/c/openstack/project-config/+/826335 has the necessary +2's do you want to +A or should I? | 16:15 |
| fungi | i can | 16:16 |
| clarkb | thanks! | 16:16 |
| fungi | approving will be faster than rechecking in that tenant | 16:17 |
| clarkb | lourot: fungi: looks like we need to update our ci checks to allow for no master branch in the upstream repo if the default branch exist and accept default branch as a key | 16:24 |
| clarkb | lourot: is that something you'd be willing to look at and update as part of your change? The nice thing with that is it will be self testing | 16:24 |
| opendevreview | Dariusz Smigiel proposed zuul/zuul-jobs master: Pull correct upstream pip version https://review.opendev.org/c/zuul/zuul-jobs/+/827297 | 16:34 |
| *** dviroel|lunch is now known as dviroel | 16:35 | |
| *** ysandeep is now known as ysandeep|out | 16:43 | |
| opendevreview | Merged openstack/project-config master: Remove unsigned tagging permission from projects https://review.opendev.org/c/openstack/project-config/+/826335 | 16:46 |
| *** marios|ruck is now known as marios|out | 16:46 | |
| opendevreview | Merged zuul/zuul-jobs master: Pull correct upstream pip version https://review.opendev.org/c/zuul/zuul-jobs/+/827297 | 16:57 |
| fungi | deploy for 826335 has finished. i'll let #openstack-release know | 17:07 |
| *** jpena is now known as jpena|off | 17:31 | |
| opendevreview | Merged opendev/system-config master: Fix mixed spaces and hard tabs in Gerrit config https://review.opendev.org/c/opendev/system-config/+/825338 | 17:48 |
| fungi | clarkb: ^ i guess we're clear to resume testing the patch now? | 17:52 |
| clarkb | fungi: ya I think we can recheck the other changes and tehn depends on should work | 17:53 |
| fungi | doing now | 17:53 |
| fungi | clarkb: should i set a new autohold for 825396 before rechecking it? | 17:54 |
| fungi | yeah, i guess i should, otherwise there's not much point to it | 17:54 |
| clarkb | fungi: I had an autohold but maybe it caught a different run? We should check and ya | 17:55 |
| fungi | i'll clean up the previous on | 17:55 |
| fungi | e | 17:55 |
| *** dviroel is now known as dviroel|biab | 17:55 | |
| clarkb | 0000000008 that one is the one I had before. not sure if it caught anything | 17:55 |
| clarkb | I've also got my hold of the gerrit node to test the tag acl which we can remoev as soon as someone pushes a tag successfully | 17:58 |
| fungi | there was also an earlier 0000000004 i'd set | 18:00 |
| fungi | any idea what we need to do to get zuul-client autohold to support the --change option correctly? | 18:00 |
| clarkb | fungi: I think you need to pass refs/changes/xy/abcxy and then it would work? So probably have the tool convert a lone number to that format? | 18:01 |
| clarkb | but I'm not completely sure of that | 18:01 |
| fungi | already tried --change=refs/changes/96/825396 but in the ref filter column autohold-list just says .* | 18:01 |
| fungi | same as when i did --change=825396 | 18:02 |
| fungi | maybe we need a newer zuul-client version? | 18:02 |
| fungi | Zuul-client version: 0.0.5.dev19 | 18:02 |
| fungi | looks like it may be up to dev21 now in master | 18:03 |
| clarkb | lookign at the log the 2 latest changes don't seem related to that | 18:04 |
| clarkb | I'm guessing it is just broken | 18:04 |
| clarkb | fungi: the client appears to send chagne through unmodified to the zuul api currently | 18:06 |
| fungi | yeah, i'm not seeing any proposed patches which might address it either | 18:06 |
| clarkb | and then the zuul api does ref_filter = project.source.getRefForChange(jbody['change']) | 18:07 |
| fungi | do i need to add ,rev? | 18:07 |
| clarkb | oh except the .* is set on an else case for ^ so either getRefForChange is returning .* or jbody(['change']) is none | 18:08 |
| clarkb | fungi: no it explicitly errors if you have a , in it | 18:08 |
| fungi | or /rev? | 18:08 |
| *** dviroel|biab is now known as dviroel | 18:08 | |
| fungi | no dice | 18:09 |
| clarkb | ok gerrit's getRefForChange can never return just .* which means we're hitting the else case for change being unset | 18:10 |
| clarkb | that implies to me the issue is entirely on the zuul-client side | 18:10 |
| clarkb | I'm not seeing how that could happen though | 18:12 |
| fungi | yeah, it literally passes "change": change in the json=args | 18:12 |
| clarkb | maybe something with argparse? | 18:13 |
| fungi | do we need to recast it as str? | 18:13 |
| fungi | oh, or int | 18:13 |
| fungi | what does the rest api want in that json blob? int or str? | 18:14 |
| fungi | you could be right and argparse is coercing it to the wrong thing | 18:14 |
| clarkb | looks like the zuul api coerces to string so either should be fine | 18:14 |
| fungi | so not that i guess | 18:15 |
| fungi | --ref=refs/changes/96/825396/3 seems to have put something other than .* there | 18:17 |
| clarkb | ya I'm stumped. Might need to add debugging in the lcient | 18:17 |
| fungi | --ref=825396 also | 18:17 |
| fungi | looks like --ref will just pass anything through | 18:17 |
| fungi | i wonder if that's a viable workaround | 18:17 |
| fungi | which of those would you expect to work? either? both? | 18:18 |
| fungi | i'll try with the simple one first | 18:19 |
| fungi | if this works, we might be able to just deprecate --change and alias it to --ref | 18:22 |
| clarkb | my keyboard is suddenly missing keys and doubing some entries. I need to fiurethat ut | 18:22 |
| fungi | have you tried turning it off and on again? | 18:22 |
| clarkb | athat is step 1 | 18:23 |
| clarkb | I semy numlo ight flapping too | 18:24 |
| clarkb | uplgging th eybard and plgig i back n dnt hlp | 18:25 |
| clarkb | tis i fun :) | 18:25 |
| clarkb | I am on a new usb bus and now it seems to work | 18:25 |
| clarkb | ugh I think that means my monitor or whatever bus the monitor is plugged into is going bad | 18:26 |
| fungi | yeah, i don't trust external hubs. my keyboard is important enough to win one of the few internal ports connecting directly to my motherboard | 18:30 |
| fungi | i've had too many bad experiences where a file transfer to a usb mass storage device caused me to no longer be able to type | 18:30 |
| clarkb | I think I've isolated it to a single port on the main device | 18:31 |
| clarkb | another port which I believe is on the same bus works fine | 18:31 |
| clarkb | oh I wonder. I've got the usb dac on possibly the same hub as the monitor hub and maybe that bus can't provide power to all the thigns at the same time | 18:32 |
| clarkb | let me try again after removing the dac | 18:32 |
| clarkb | ya quick testing seems ot indicate that is a likely cause | 18:33 |
| clarkb | "fun" | 18:33 |
| clarkb | anyway sorry to distract I just had a really weird experience with that is all :) | 18:34 |
| fungi | yeah, in fact this workstation has two stacks of usb ports on opposite sides of the motherboard. the stack where my keyboard is plugged in has nothing else connected, so i expect it's the only device on that bus at all other than probably an inbuilt hub which connects those ports | 18:34 |
| fungi | the other stack is where my pointer device and a powered hub are attached | 18:34 |
| fungi | since shuffling them around to this configuration i've had zero problems with the keyboard | 18:35 |
| *** amoralej is now known as amoralej|off | 18:35 | |
| clarkb | TIL this is something to be careful of | 18:38 |
| opendevreview | Tristan Cacqueray proposed zuul/zuul-jobs master: Update cargo-test job to support system packages https://review.opendev.org/c/zuul/zuul-jobs/+/827131 | 18:38 |
| clarkb | the power cable for the dac is now on its own bus I believe | 18:39 |
| fungi | yep. for me, there's no peripheral so important that it should get to preempt my keyboard. like, ever | 18:41 |
| fungi | neat, i missed that we're now getting warnings in comments from zuul. that's awesome | 18:44 |
| fungi | Warning: Job system-config-run-review-3.4 requires artifact(s) gerrit-3.4-container-image provided by build 7a2820c9b3934619a761a7a5092e0f5a (triggered by change 825337 on project opendev/system-config), but that build failed with result "FAILURE" | 18:45 |
| fungi | that one was from a few days ago, just hadn't spotted it until now | 18:46 |
| clarkb | ya that was what caued the previous buidls to not do what we needed tehm to do | 18:46 |
| fungi | i knew the feature was in the works but couldn't recall if it had been implemented. i guess it has | 18:46 |
| clarkb | I'm going to do a reboot now just to clear any power state sillyness gremlins that may be in my usb stack. Hopefully I come back :) | 18:46 |
| fungi | good call | 18:46 |
| clarkb | and back. Seems like everything is fine after reorganizing cables | 18:54 |
| fungi | no luck with --ref=825396 (it doesn't seem to have held the node). i'll try again now with --ref=refs/changes/96/825396/3 | 19:33 |
| ianw | fungi: it is making it into the hold request right? when you check with autohold-list, the values look right? | 19:56 |
| ianw | just we had some issues with quoting etc. when passing through arguments using the wrapper command | 19:56 |
| fungi | yeah, it makes the hold but sets the change filter to .* instead of whatever value you supply | 19:57 |
| fungi | and github becomes increasingly more gerritlike... if you click the "force-pushed" on a pr which had a push --force update, it shows you a diff of the changes between the old and new revisions | 20:01 |
| fungi | https://github.com/pypa/pip/pull/10869 | 20:02 |
| *** dviroel is now known as dviroel|afk | 20:35 | |
| fungi | oh hey, --ref=refs/changes/96/825396/3 worked! | 20:40 |
| fungi | clarkb: ianw: ^ | 20:40 |
| clarkb | good to know | 20:40 |
| fungi | maybe that can give us some clue as to how to fix --change in the client | 20:40 |
| clarkb | at least until we fix it | 20:41 |
| fungi | anyway, the held gerrit is at https://172.99.69.247/ now | 20:41 |
| clarkb | looks like there is a hash in the file diff gitea links. Also super cool we can depends on upstream like that | 20:42 |
| fungi | and the "gitea" link at https://172.99.69.247/c/x/test-project/+/3/1/file-3.txt looks right, yeah | 20:42 |
| clarkb | corvus: ^ fyi thats a really fun use of depends on in opendev | 20:48 |
| opendevreview | Ade Lee proposed zuul/zuul-jobs master: WIP/DNM - Test new version of python3 https://review.opendev.org/c/zuul/zuul-jobs/+/827366 | 20:49 |
| fungi | okay, announcement followup about the signed tags situation has been sent | 20:56 |
| corvus | clarkb: you should be sure to tell the upstream gerrit folks you did that :) | 20:57 |
| fungi | yeah, a comment on the change in their gerrit to that effect would be great | 20:58 |
| opendevreview | Tristan Cacqueray proposed zuul/zuul-jobs master: Remove the periodic-weekly pipeline https://review.opendev.org/c/zuul/zuul-jobs/+/827369 | 21:00 |
| clarkb | can do | 21:10 |
| clarkb | commented on https://gerrit-review.googlesource.com/c/gerrit/+/329279 | 21:15 |
| clarkb | ok I looked at forward porting the gerrit 3.4 fix for acl pushing to 3.5 but 3.5 already has at least 2 outstanding forward merge chagnes and I don't want to deal with stacking that properly. I'll just keep an eye on it to ensure it gets forward ported | 21:27 |
| fungi | probably good enough | 21:27 |
| fungi | as long as it happens before we upgrade, should be fine | 21:27 |
| clarkb | this is one reason why I'm a little concerned about updating to latest release in gerrit it seems to lag all the bug fixes. I guess we would just have to get aggressive about forward porting things ourselves if/when we get to the latest release | 21:27 |
| fungi | yep | 21:27 |
| opendevreview | James E. Blair proposed opendev/system-config master: Remove gearman from Zuul https://review.opendev.org/c/opendev/system-config/+/826791 | 21:52 |
| opendevreview | Ade Lee proposed zuul/zuul-jobs master: WIP/DNM - Test new version of python3 https://review.opendev.org/c/zuul/zuul-jobs/+/827366 | 22:15 |
| ianw | what's the feeling on what will happen if we merge the centos-8 node removal https://review.opendev.org/c/opendev/base-jobs/+/827181 with zuul reporting twice on it? | 22:45 |
| opendevreview | Ian Wienand proposed openstack/project-config master: Remove centos-8 https://review.opendev.org/c/openstack/project-config/+/827184 | 22:48 |
| fungi | ianw: i think openstack/openstack-ansible-tests is reporting from the openstack tenant with a configuration error because it can't evaluate the configuration there, but the change is being gated by the opendev tenant where it's passing, so it should merge modulo some other problem | 22:52 |
| fungi | i expect the end result to be that configuration for openstack/openstack-ansible-tests will be broken once that merges, but presumably we're resigned to that | 22:53 |
| fungi | odds are it was broken anyway since we mirrored red hat's removal of the packages | 22:53 |
| ianw | yep, the jobs are now all broken irrespective of the node being there or not, the only path out of that would be to somehow start pointing things to vault.centos.org | 22:54 |
| ianw | that seems like pointless work though | 22:54 |
| fungi | yes, probably more work than fixing the remaining tests | 22:55 |
| opendevreview | Merged opendev/base-jobs master: Remove centos-8 as it is EOL https://review.opendev.org/c/opendev/base-jobs/+/827181 | 22:59 |
| fungi | and there we go | 22:59 |
| ianw | https://review.opendev.org/c/openstack/project-config/+/827184 was unhappy until that merged, but is verifying itself now after a recheck | 23:07 |
| opendevreview | Merged opendev/system-config master: Remove gearman from Zuul https://review.opendev.org/c/opendev/system-config/+/826791 | 23:11 |
| fungi | remove all the things | 23:12 |
| ianw | fungi: https://review.opendev.org/c/zuul/nodepool/+/826773 and https://review.opendev.org/c/zuul/zuul/+/826772 are two more if you're in a removal mood :) | 23:16 |
| fungi | oh, right, we talked about that already. in they go! | 23:19 |
| opendevreview | Merged openstack/project-config master: Remove centos-8 https://review.opendev.org/c/openstack/project-config/+/827184 | 23:25 |
| ianw | when that applies i'll dig into fedora, that should work now | 23:41 |
| ianw | f25 | 23:41 |
| ianw | f35 even | 23:41 |
| ianw | and we can hopefully turf f34 | 23:41 |
| fungi | more things to destroy! er, i mean delete! | 23:42 |
| *** dviroel|afk is now known as dviroel | 23:44 | |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!