| fungi | 2024-05-23 00:33:49.298 | > Selecting previously unselected package dns-root-data. | 01:23 |
|---|---|---|
| fungi | much more like it | 01:23 |
| fungi | it's finalizing the image versions now so should be uploading shortly | 01:23 |
| Clark[m] | fungi: what did you have to do to get a new build? Delete an old one? | 01:24 |
| fungi | no, for whatever reason it ignored my image-build command that time. when i ran it again after the building one became ready, it started building the new one just fine | 01:25 |
| fungi | but i do wonder whether it has to do with there being two ready images vs just one | 01:25 |
| fungi | anyway, it's checksumming now | 01:27 |
| fungi | should start uploading any moment | 01:27 |
| fungi | unfortunately the latest build landed on the other builder, so had to populate the cache all over again | 01:32 |
| opendevreview | Merged openstack/project-config master: Clean up unused labels from nl02 config https://review.opendev.org/c/openstack/project-config/+/920190 | 04:23 |
| frickler | o.k., some zuul-noble-jobs passing now, others need fixes, like bindep: "No package matching 'python-dev' is available", pretty sure we have a fix for that for bookworm already, too | 04:40 |
| opendevreview | Qihui Zhao proposed openstack/project-config master: to create a repo under cfn for a new-initiated sub-group cfn-security https://review.opendev.org/c/openstack/project-config/+/920264 | 08:41 |
| opendevreview | Merged openstack/project-config master: to create a repo under cfn for a new-initiated sub-group cfn-security https://review.opendev.org/c/openstack/project-config/+/920264 | 14:26 |
| clarkb | I've got a noble zuul-jobs fixup change in the works should haev it pushed shortly | 15:46 |
| clarkb | just currently battling with ubuntu pacakge sea4rch to determine why pkg-config is not in build essential anymore (or at least that is my assumption)_ | 15:46 |
| opendevreview | Clark Boylan proposed zuul/zuul-jobs master: Add ubuntu-noble testing https://review.opendev.org/c/zuul/zuul-jobs/+/920205 | 15:53 |
| opendevreview | Clark Boylan proposed zuul/zuul-jobs master: Ombibus set of fixups for Ubuntu Noble https://review.opendev.org/c/zuul/zuul-jobs/+/920299 | 15:53 |
| opendevreview | Clark Boylan proposed zuul/zuul-jobs master: Ombibus set of fixups for Ubuntu Noble https://review.opendev.org/c/zuul/zuul-jobs/+/920299 | 16:28 |
| opendevreview | Clark Boylan proposed zuul/zuul-jobs master: Add ubuntu-noble testing https://review.opendev.org/c/zuul/zuul-jobs/+/920205 | 16:28 |
| opendevreview | Clark Boylan proposed zuul/zuul-jobs master: Ombibus set of fixups for Ubuntu Noble https://review.opendev.org/c/zuul/zuul-jobs/+/920299 | 17:01 |
| opendevreview | Clark Boylan proposed zuul/zuul-jobs master: Add ubuntu-noble testing https://review.opendev.org/c/zuul/zuul-jobs/+/920205 | 17:01 |
| opendevreview | Clark Boylan proposed zuul/zuul-jobs master: Ombibus set of fixups for Ubuntu Noble https://review.opendev.org/c/zuul/zuul-jobs/+/920299 | 17:37 |
| opendevreview | Clark Boylan proposed zuul/zuul-jobs master: Add ubuntu-noble testing https://review.opendev.org/c/zuul/zuul-jobs/+/920205 | 17:37 |
| opendevreview | Clark Boylan proposed zuul/zuul-jobs master: Ombibus set of fixups for Ubuntu Noble https://review.opendev.org/c/zuul/zuul-jobs/+/920299 | 17:54 |
| opendevreview | Clark Boylan proposed zuul/zuul-jobs master: Add ubuntu-noble testing https://review.opendev.org/c/zuul/zuul-jobs/+/920205 | 17:54 |
| opendevreview | Merged zuul/zuul-jobs master: Ombibus set of fixups for Ubuntu Noble https://review.opendev.org/c/zuul/zuul-jobs/+/920299 | 18:54 |
| Clark[m] | fungi: any chance there are emails in the openstack discuss moderation queue? I got an email directly that hasn't shown up in my mailing list inbox | 18:58 |
| Clark[m] | Want to make sure there isn't some other issue | 18:59 |
| Clark[m] | Hrm actually it is in the archive. Maybe it is an issue on my side | 18:59 |
| fungi | Clark[m]: by default, mm3 tries to avoid sending you duplicate copies of messages, mainly when there's an address in to or cc which is also subscribed | 19:00 |
| fungi | check the archive to confirm, and if you want the duplicate copies i think it's a checkbox in your list subscription settings | 19:00 |
| Clark[m] | Oh neat. That breaks my filtering rules unfortunately. There is also the trove email that I don't think I got today | 19:01 |
| Clark[m] | The one I first got suspicious about is the dib one | 19:01 |
| fungi | the dib one did go through moderation | 19:01 |
| fungi | but i approved it hours ago and saw the list copy arrive | 19:01 |
| Clark[m] | Ok I don't think I got that one delivered I wonder what is up with my subscription delivery | 19:02 |
| Clark[m] | Oh wait sorry dib is explained by deduplication | 19:02 |
| Clark[m] | Trove is the one I see in the archive and not locally | 19:02 |
| fungi | if your mail client has the ability to dynamically set reply-to and/or mail-followup-to headers when you're replying on mailing lists, that can be a workaround to reduce the replies that end up cc'ing you | 19:02 |
| fungi | another possibility with the trove e-mail is that it ended up caught in a spam filter, or the server is being greylisted and the message just hasn't reached you yet, or could be delays in the outbound mta queue on the lists server | 19:03 |
| fungi | `exim4 -bp` on lists.o.o will show you the current deferral queue | 19:04 |
| fungi | the message ids can be matched to entries in /var/log/exim/mainlog or content and deferral log files in /var/spool/exim if you need details | 19:05 |
| Clark[m] | Ok I'm currently consuming lunch but will dig more | 19:05 |
| fungi | and yeah, i don't personally like that mm3 changed the default to deduplicate when mm2 defaulted to not deduplicating, but since it's per-user configurable i don't have much room to complain ;) | 19:06 |
| Clark[m] | I did an explicit search for the trove message and now it loads. Maybe this is merely a client caching issue | 19:07 |
| opendevreview | Merged zuul/zuul-jobs master: Add ubuntu-noble testing https://review.opendev.org/c/zuul/zuul-jobs/+/920205 | 19:08 |
| fungi | i also received it, arrived at lists.o.o at 18:57:56 and reached my mta at 19:05:44, so there's a 7m50s delay somewhere in there | 19:08 |
| Clark[m] | Ya sending openstack discuss list emails has about a 5-10 minute delay these days | 19:09 |
| Clark[m] | I guess I could've simply raced seeing it in the archive and getting it delivered. All seems well now and the deduplication is the only unexpected behavior which I now know about | 19:09 |
| fungi | the server isn't loaded at all (load average seems to be at/under 0.5 for a while), so it may just be that we need to tell mailman it can do more simultaneous deliveries or something | 19:10 |
| clarkb | makes sense | 19:22 |
| fungi | i remember seeing discussion on the users ml about it defaulting to a conservative limit in order to avoid overwhelming the mta on smaller systems | 19:30 |
| fungi | much more polite than mm2 which didn't bother, and would just chuck everything at exim the moment it was received | 19:31 |
| frickler | if someone has time to check why https://review.opendev.org/c/openstack/kolla-ansible/+/920294 is failing all jobs without logging results, that would be nice, otherwise I'll dig into zuul logs tomorrow, eoding now | 19:49 |
| fungi | frickler: FAILED! => {"msg": "The task includes an option with an undefined variable. [...] | 20:21 |
| fungi | frickler: i think ansible is choking on the commit message | 20:23 |
| corvus | should be fixed in https://review.opendev.org/c/zuul/zuul/+/920361 Fix missing unsafe marker in new build_refs var | 20:25 |
| fungi | thanks! i was just about to post the same link | 20:26 |
| clarkb | fungi: with the merging of the zuul-jobs update for noble I wonder if we should suggest that people start using noble? For example for python312 jobs? I'm about to push a change up for zuul to do that now and see what happens | 22:02 |
| clarkb | oh we need an ubuntu-noble nodeset in base jobs too | 22:04 |
| * clarkb does that | 22:04 | |
| opendevreview | Clark Boylan proposed opendev/base-jobs master: Add ubuntu-noble nodesets https://review.opendev.org/c/opendev/base-jobs/+/920363 | 22:05 |
| clarkb | gmann: ^ fyi we're getting close to making that an option for jobs. May be worth converting the existing py312 jobs in poenstack over to avoid building python from scratch | 22:18 |
| clarkb | looks like there may be a problem running the zuul py312 job though so need to figure that out | 22:18 |
| clarkb | ah its a libffi bindep problem | 22:19 |
| clarkb | fungi: if you have a quick moment for https://review.opendev.org/c/openstack/diskimage-builder/+/920215 I think that is an easy bash review. I just don't want to single core approve it because it is in simple-init which has a fairly large blast radius | 22:26 |
| fungi | clarkb: yes on switching to noble for 3.12 but only after the default ansible version increase or when overriding ansible with a newer version, right? | 22:50 |
| clarkb | fungi: ya though it can be done a per job basis. In the zuul case zuul's tenant is already flipped | 22:52 |
| fungi | clarkb: on 920215 do we care that it's introducing stray trailing whitespace on that line? also would it make more sense to just drop the else? | 22:52 |
| clarkb | I don't think we care about the trailing whitespace if the bash linter (if we even run one) in dib doesn't catch it. One raeson for the noop else is to keep ianw's comment there in a clear scope | 22:53 |
| clarkb | but we could drop the block if you feel strongly about it | 22:53 |
| fungi | nah, it's fine with me | 22:54 |
| clarkb | I do think we can add the noble nodesets nowish (they are noops until used) then announce a day for changing the default ansible version. Maybe early june? and then proceed from there | 22:54 |
| clarkb | really its up to openstack to decide if they want to swap out the node for py312 jobs and set the ansible version. Or if they want ot run on noble at all. Just calling out that it is an option now | 22:55 |
| clarkb | In zuul's case I expect the switch to make the job run about 10 minutes faster. | 22:55 |
| fungi | right, i think openstack is going to want to make that switch | 22:55 |
| ianw | clarkb: thanks :) not sure what i was thinking | 22:56 |
| fungi | right now their 3.12 jobs are merely advisory anyway: https://governance.openstack.org/tc/reference/runtimes/2024.2.html | 22:57 |
| clarkb | ianw: TheJulia fixed it :) | 22:57 |
| TheJulia | wait, what did I do? | 22:57 |
| TheJulia | I'm only supposed to break things! | 22:57 |
| clarkb | TheJulia: the continue -> true fix in dib | 22:57 |
| fungi | TheJulia: fixed all the bugs | 22:57 |
| TheJulia | woot! | 22:57 |
| * TheJulia dances | 22:58 | |
| fungi | we can go home now | 22:58 |
| TheJulia | nooooo!!! ;) | 22:58 |
| fungi | mission complete | 22:58 |
| clarkb | fungi: actually zuul is using default ansible version 8 too | 22:58 |
| clarkb | so maybe the problem is being able to run ansible commands on noble requires newer ansible | 22:58 |
| clarkb | but talkign to ansible nodes is fine with older ansible | 22:59 |
| clarkb | because the zuul jobs are definitely working | 22:59 |
| fungi | oh, that's convenient | 22:59 |
| clarkb | or maybe dansmith's ansible from the distro didn't get patched updates to make it work and upstream did so for 8? | 22:59 |
| fungi | so yeah, let's get the nodesets added, announce availability, suggest projects doing nested ansible that they may need a newer version | 23:00 |
| clarkb | or just hand wave around potential compatibility problems and if seen update to newer ansible | 23:00 |
| clarkb | but ya sounds like a plan | 23:00 |
| fungi | and in the openstack tenant we can switch the py312 jobs from pyenv to distro packaged | 23:00 |
| fungi | after some heads up | 23:01 |
| clarkb | the main upside is going to be removing the compile step which takes a couple minutes and getting a distro compiled python that sets a bunch more optimization flags than pyenv does by default. In theory it will make thinsg go faster | 23:02 |
| fungi | vroom vroom | 23:04 |
| opendevreview | Merged openstack/diskimage-builder master: simple-init: Swap continue for true https://review.opendev.org/c/openstack/diskimage-builder/+/920215 | 23:58 |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!