| *** rlandy|bbl is now known as rlandy|out | 01:51 | |
| *** pojadhav|sick is now known as pojadhav | 04:16 | |
| *** soniya29|ruck is now known as soniya29|ruck|afk | 04:32 | |
| *** ysandeep|out is now known as ysandeep | 04:37 | |
| *** raukadah is now known as chandankumar | 04:48 | |
| *** marios is now known as marios|ruck | 05:14 | |
| *** jpena|off is now known as jpena | 07:20 | |
| *** ysandeep is now known as ysandeep|lunch | 08:08 | |
| *** soniya29|ruck|afk is now known as soniya29|ruck | 08:21 | |
| *** marios|ruck is now known as marios | 08:26 | |
| *** ramishra_ is now known as ramishra | 09:38 | |
| *** ysandeep|lunch is now known as ysandeep | 10:18 | |
| *** rlandy|out is now known as rlandy | 10:39 | |
| *** dviroel|afk is now known as dviroel | 11:30 | |
| *** ravlew is now known as Guest2256 | 11:38 | |
| *** ravlew5 is now known as ravlew | 11:52 | |
| *** frenzyfriday is now known as frenzyfriday|food | 12:22 | |
| *** dasm|off is now known as dasm | 13:49 | |
| *** ysandeep is now known as ysandeep|out | 14:04 | |
| *** dviroel is now known as dviroel|lunch | 14:50 | |
| fungi | some strange bitrot in bindep's docs job: https://zuul.opendev.org/t/opendev/build/5e40df1ecaec41a7aa2b12f6dd001104/log/job-output.txt#851-853 | 14:58 |
|---|---|---|
| *** marios is now known as marios|out | 15:22 | |
| clarkb | fungi: I think testtools is only in test-requirements and not in the normal requirements or the docs requirements. I'm guessing some dependency was pulling it in previously but not anymore | 15:24 |
| fungi | oh, that would definitely explain it | 15:25 |
| clarkb | its trying to autodoc the tests | 15:25 |
| clarkb | I think maybe we don't need to autodoc the tests and that should avoid the problem | 15:25 |
| fungi | yeah, i wonder if we should just drop autodoc entirely there | 15:26 |
| fungi | https://docs.opendev.org/opendev/bindep/latest/ | 15:26 |
| fungi | i don't see where we're actually linking any autodoc-generated content | 15:26 |
| clarkb | ya either we add test requirements to the docs builds or we drop autodoc of the tests | 15:26 |
| clarkb | I think both are valid fixes. I lean towards dropping autodoc of tests as I doubt anyone uses that info | 15:27 |
| fungi | well, i mean drop the autodoc extension | 15:27 |
| fungi | i don't see where we've got any toc linking it | 15:27 |
| fungi | scratch that, it's in the contributing chapter | 15:27 |
| fungi | https://docs.opendev.org/opendev/bindep/latest/contributing.html#python-api | 15:28 |
| fungi | and we were previously autodoc'ing bindep.tests too | 15:29 |
| fungi | so yes, we can just add testtools to doc/requirements.txt i guess | 15:29 |
| clarkb | yes, but is there value to having a page that autodocs the tests? | 15:29 |
| clarkb | that seems unnecessary | 15:29 |
| fungi | it's in the contributor documentation, so i can see it being relevant there, but perhaps we can also expect contributors to just read the source | 15:30 |
| clarkb | yes exactly | 15:30 |
| clarkb | autodoc to me is useful for where you have third party integrations for things like libraries | 15:30 |
| clarkb | the internal unittests of an appliaction don't really need that imo | 15:31 |
| clarkb | but if we keep them updating the docs requirements is also easy | 15:32 |
| opendevreview | Jeremy Stanley proposed opendev/bindep master: Drop unit tests autodoc from contributor docs https://review.opendev.org/c/opendev/bindep/+/860274 | 15:33 |
| *** rlandy is now known as rlandy|biab | 15:40 | |
| opendevreview | Merged opendev/bindep master: Drop unit tests autodoc from contributor docs https://review.opendev.org/c/opendev/bindep/+/860274 | 16:06 |
| opendevreview | Merged opendev/bindep master: Add support for popos https://review.opendev.org/c/opendev/bindep/+/843444 | 16:12 |
| clarkb | fungi: I'll bring this up in the team meeting today, but yuo'ev been poking at mm3 with me so may want to get a head start. https://review.opendev.org/c/opendev/system-config/+/860157 is a "lightweight" fork of the upstream mm3 docker images. The only change I've made between the upstream and this change is to add lynx to the image. We have the opportunity to do a heavier fork and | 16:17 |
| clarkb | move some of the bind mounted workarounds in the role to the image. We can change uids and gids. We can port in the PR fix for uwsgi buffer sizes you linked to on the upstream mailing lsit and so on. I think we should probably make a decision on 1) whether or not we want to fork and 2) how fokred do we want to get :) | 16:17 |
| *** jpena is now known as jpena|off | 16:18 | |
| fungi | yes, i agree that's a good topic for the meeting | 16:18 |
| clarkb | I'm hoping that having a change up will help people see what is involved here. I am wary of building our own images from scratch instead of forking becuase there appears to be a lot of inside baseball knowledge for what versions of libraries and tools all fit together to make this work. | 16:18 |
| fungi | i do feel like trying not to fork farther than we need to would leave us the ability to reintegrate upstream images if maintenance of those resumes | 16:19 |
| clarkb | yup exactly | 16:19 |
| clarkb | I guess this really comes down to "do we want to maintain our own images long term or attempt to realign with upstream if it becomes more active again" | 16:20 |
| fungi | it's not been lying dormant for long enough yet that i'm expecting the work is abandoned | 16:20 |
| fungi | and the maintainer has been semi-around in the mm3 community fairly recently | 16:20 |
| clarkb | and if we start here we can shift to a bigger fork later with little effort | 16:20 |
| clarkb | going the other direction is potentially more problematic | 16:20 |
| *** dviroel|lunch is now known as dviroel | 16:21 | |
| *** rlandy|biab is now known as rlandy | 16:22 | |
| opendevreview | James E. Blair proposed zuul/zuul-jobs master: WIP: Add ensure-pip-localhost job https://review.opendev.org/c/zuul/zuul-jobs/+/860293 | 18:21 |
| opendevreview | James E. Blair proposed zuul/zuul-jobs master: More narrowly tailor the ensure-pip Debian workaround https://review.opendev.org/c/zuul/zuul-jobs/+/860294 | 18:21 |
| opendevreview | James E. Blair proposed zuul/zuul-jobs master: More narrowly tailor the ensure-pip Debian workaround https://review.opendev.org/c/zuul/zuul-jobs/+/860294 | 18:36 |
| opendevreview | James E. Blair proposed zuul/zuul-jobs master: Combine ensure-pip playbooks into a test role https://review.opendev.org/c/zuul/zuul-jobs/+/860296 | 18:46 |
| opendevreview | Merged zuul/zuul-jobs master: More narrowly tailor the ensure-pip Debian workaround https://review.opendev.org/c/zuul/zuul-jobs/+/860294 | 18:52 |
| frickler | RL9 builds are broken with some mirror issue it seems | 19:16 |
| clarkb | [MIRROR] systemd-udev-250-6.el9_0.1.x86_64.rpm: Interrupted by header callback: Server reports Content-Length: 107 but expected size is: 1547312 | 20:18 |
| clarkb | 4168812 is the size I get back from the rackspace mirror | 20:20 |
| clarkb | I wonder if it will work on the next round and previous ones had half snyced mirrors or something | 20:20 |
| *** rlandy is now known as rlandy|out | 20:26 | |
| *** dviroel_ is now known as dviroel|afk | 20:37 | |
| clarkb | no it looks like more recent builds are still failing. Maybe something wrong with a specific mirror? cc NeilHanlon | 20:55 |
| clarkb | I need to pop out for a bit but will try to look more when I return | 20:56 |
| *** dasm is now known as dasm|off | 22:32 | |
| clarkb | ok I ran `dnf install -y findutils util-linux sudo python3 NetworkManager` in a docker container using rockylinux:9 and that succeeded (it installed the same set of packages too) | 23:03 |
| clarkb | unfortunately dnf doesn't say what it was talking to that errored just that some MIRROR errored | 23:05 |
| clarkb | (I wonder if this is a dnf bug) | 23:05 |
| clarkb | the container image's yum.repos.d lists mirrorlist=https://mirrors.rockylinux.org/mirrorlist?arch=$basearch&repo=BaseOS-$releasever$rltype for the baseos mirror | 23:07 |
| clarkb | if I `curl 'https://mirrors.rockylinux.org/mirrorlist?arch=x86_64&repo=BaseOS-9'` I consistently get back the dfw mirror | 23:10 |
| clarkb | which makes sense | 23:10 |
| clarkb | and if I wget https://dfw.mirror.rackspace.com/rocky/9.0/BaseOS/x86_64/os/Packages/s/systemd-udev-250-6.el9_0.1.x86_64.rpm I get what appears to be a full size result back. I wonder if we aren't hitting some dib cache issue now | 23:11 |
| clarkb | like maybe it was broken and we cached the broken | 23:11 |
| clarkb | hrm I'm not seeing rockylinux content in /opt/dib_cache/yum | 23:13 |
| clarkb | fetching 404'd content results in 196 bytes which is larger than the 107 it is complaining about | 23:14 |
| clarkb | heh I havne't been the only one confused by dnf's lack of logging https://unix.stackexchange.com/questions/252538/dnf-how-to-show-which-mirror-url-has-been-chosen | 23:16 |
| clarkb | "FYI, DNF actually does not know the URLs. It just knows the URL of the "metalink" and the file name (and some other metadata) of the package." this might be where I get of the train and go do something else | 23:17 |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!