11:15 <pkira> Hi mupuf, I have a question regarding CIBugLog

11:39 <mupuf> pkira: hi :)

11:39 <mupuf> pkira: I'm all earsa

11:39 <mupuf> ears*

11:40 <pkira> Hi! Let me just paste message I have prepared earlier

11:40 <pkira> So, I face issue where I would like to implement support for custom closed source test framework. This framework produces pytest results and custom logfile that includes dmesg. I would like to merge pytest results and dmesg, the problem is that then I cannot use pytest format as it does not support dmesg. I see some solutions.

11:40 <pkira> * Formating these results into piglit format which is already supported in CBL

11:40 <pkira> * Creating json where keys maps to the exactly same "keys" in CBL, for example: {tests: {name: name, stdout: stdout, dmesg: dmesg…}

11:41 <pkira> So, What do you think about that, would be it good to support just generic json format in CBL?

11:43 <pkira> mupuf Ok, I see I just need to use someone nick to "mention" them. First time using IRC

11:43 <mupuf> pkira: one could argue that piglit is generic json already :D But honestly, the code is meant to make it easy to add new formats, so if you have an idea of a format that would be easier to generate than piglit, I'm fine with merging support for it

11:44 <mupuf> hehe, sorry, I was just slow at composing my answer

11:46 <pkira> True, piglit results look like generic json

11:46 <mupuf> and you get vizualization tools for it too :)

11:47 <mupuf> either the ones from piglit itself, or the ones that Adrinael and tsa are maintaining

11:47 <pkira> That's another reason to use piglit results

11:48 <pkira> Thank you for your answer, it's useful, now I have to discuss about it with my teammates

11:48 <mupuf> piglit as a format is a pretty good transport mechanism

11:48 <pkira> Yeah, just the name itself doesn't sound generic - piglit

11:49 <mupuf> ah ah, yeah

11:49 <pkira> Everybody know what "json" is, I just got to know that piglit produces json a few days ago

11:50 <pkira> Ok, anyway, thank you for your help :D

11:50 <mupuf> had you all your results in a pytest format, adding an importer for that in cibuglog would have been nice... but if you need dmesg too...

11:51 <pkira> That's true, I already have implemented some kind of support of pytest in CBL, but I just asked to add dmesg to it... so.. yeah

11:51 <pkira> I just got asked*

11:51 <mupuf> Ack!

11:52 <mupuf> pkira: please write issues on the project with all the sharp corners you hit while putting cibuglog in production

11:52 <mupuf> and, of course, MRs are welcomed :p

11:53 <pkira> Yeah, sure, I will :D

11:53 <pkira> Thank you!

11:54 <pkira> So, I guess: till the next time

11:55 <mupuf> pkira: good luck, and have a good weekend

11:55 <pkira> mupuf: Thanks, you too!

13:53 <emersion> Mithrandir: ping for the VM

18:07 <Mithrandir> emersion: in progress, disk size will be somewhat interesting (which means downtime during maintenance, rather than a failover), but I think that might be ok, since it's broadly email and DNS?

18:07 <Mithrandir> (as in, it's literally running debootstrap now)

19:15 <Mithrandir> emersion: can you see if you can ssh into gabe2.freedesktop.org? It's a fairly blank VM, so need a bunch of stuff set up on it. I was thinking of replacing gabe with it, but if we just want mailman on it, that's fine (but then we should rename it to something more reasonable)

19:29 <emersion> Mithrandir: thanks! i can't login it seems, my pubkey for FDO root is

19:29 <emersion> ssh-ed25519 AAAAC3NzaC1lZDI1NTE5AAAAIKhSqHAPksA82NU1bNN98ytZqKV1Cl+46pSGn8PMoo0s contact@emersion.fr

19:29 <emersion> (you can check it on old gabe)

19:29 <Mithrandir> emersion: sorry, try again?

19:30 <emersion> still a no it seems

19:30 <emersion> (asks for a password)

19:30 <emersion> ah, hm

19:30 <Mithrandir> as emersionR?

19:30 <emersion> on gabe my usernme is emersionR

19:30 <emersion> yea

19:30 <emersion> oh

19:30 <Mithrandir> you seem to be trying a different user name, from the logs.

19:30 <emersion> nevermind, works!

19:31 <Mithrandir> cool

19:31 <emersion> i copied a · by mistake

19:31 <Mithrandir> happy to hear thoughts about splitting out mailman vs not; IP reputation is a thing, so we might want to have this replace gabe, but that also means copying over the other misc things from gabe, so…

19:31 <emersion> daniels, do we want to migrate everything to the new VM, or just mailman?

19:32 <emersion> maybe we can just pick a new name, and decide later to decomission gabe when it's not longer used

19:32 <Mithrandir> if that happens before the heat death of the universe.

19:32 <emersion> ahah

19:32 <Mithrandir> it's serving DNS, for instance. :-)

19:33 <Mithrandir> (which we probably shouldn't be hosting ourself, but meh, it is.)

19:33 <emersion> i feel like there'll always be "that last service we haven't migrated yet"

19:33 <Mithrandir> yup

19:33 <Mithrandir> at least absent a push to migrate "the rest"

19:33 <emersion> we also have some cgit stuff? or maybe just some archives of it

19:33 <Mithrandir> I guess we'll do that when i386 isn't supported any more.

19:33 <Mithrandir> cgit should be on molly, not gabe, I thought?

20:31 <daniels> I'm pretty sure the only things left on gabe are DNS and mail?

20:31 <daniels> cgit is indeed on molly

20:32 <emersion> okay, just saw files lying around somewhere

20:32 <emersion> when searching for stuff to delete

20:47 <Mithrandir> daniels: there's a /srv/bod.x.org and members.x.org, but neither point at gabe, so…

22:03 <rodrigovivi> does anyone have any clue about: an igt build container failure:

22:03 <rodrigovivi> STEP 2: RUN apt-get update

22:03 <rodrigovivi> 35error running container: error creating container for [/bin/sh -c apt-get update]: time="2022-06-23T22:01:54Z" level=warning msg="signal: killed"

22:03 <rodrigovivi> in a gitlab fork