07:37 <MrCooper> haha, https://gitlab.freedesktop.org/mesa/mesa/-/issues/9504 confidential spam issue, exclusively for Mesa project members

07:40 <airlied> really trying to get those niches

07:58 <bochecha> hi everyone, I'm having trouble with the gitlab.freedesktop.org registry, how can I list the images that are on it, so I can use one in the Gitlab CI ? :-/

07:59 <MrCooper> e.g. https://gitlab.freedesktop.org/<namespace>/<project>/container_registry

07:59 <bochecha> in the gitlab-ci.yml file for libcangjie I have BUILD_IMAGE_REGISTRY: registry.freedesktop.org/cangjie/build-essential, but that URL is 404

08:00 <bochecha> aha !! :)

08:00 <bochecha> so in my case that's https://gitlab.freedesktop.org/cangjie/libcangjie/container_registry

08:02 <bochecha> just did the same for my second project (pycangjie) and there are no images for it, so now I get to build and push them, thank you :D

08:06 <MrCooper> no worries

08:06 <bochecha> huh, my login password doesn't work for this :-/

08:07 <bochecha> same one I use to login to gitlab in Firefox

08:12 <MrCooper> you mean for logging into the registry?

08:13 <MrCooper> you need to create an access token for that under https://gitlab.freedesktop.org/-/profile/personal_access_tokens

08:13 <MrCooper> or the corresponding project setting

08:18 <bochecha> thanks, did that

08:19 <bochecha> now I could login to the registry, and I'm trying to build the image and push it to the registry, but got a permission denied...

08:19 <bochecha> Got permission denied while trying to connect to the Docker daemon socket at unix:///var/run/docker.sock: Post "http://%2Fvar%2Frun%2Fdocker.sock/v1.24/build?buildargs=%7B%7D&cachefrom=%5B%5D&cgroupparent=&cpuperiod=0&cpuquota=0&cpusetcpus=&cpusetmems=&cpushares=0&dockerfile=Dockerfile&labels=%7B%7D&memory=0&memswap=0&networkmode=default&rm=1&shmsize=0&t=registry.freedesktop.org%2Fcangjie%2Fpycangjie&target=&ulimits=null&ver

08:19 <bochecha> sion=1": dial unix /var/run/docker.sock: connect: permission denied

09:44 <daniels> bochecha: that's a local issue - whatever you're running as doesn't have permission for that UNIX socket

10:17 <bochecha> daniels, ah, indeed, my bad 🤦‍♂️

13:31 <alatiera> maintenance ongoing on gst-5

13:31 <alatiera> did manage to migrate it from mbr to gpt on the fly!

13:57 <zmike> mupuf: https://gitlab.freedesktop.org/mesa/mesa/-/pipelines/956647

14:36 <mupuf> zmike: I'm away from home rn, will check in30 minutes

14:36 <mupuf> But seems to have fixed itself

14:36 <zmike> smh another vacation

14:36 <mupuf> Lol

14:37 <mupuf> If walking the dog is counted as such, then yeah ;)

14:38 <zmike> "if" ?

14:38 <zmike> better believe we'll be discussing this in your performance review friday

14:40 <mupuf> Affirmative, sir!

20:32 <glehmann> seems like there is an issue with marge timeouts before windows jobs can start

20:35 <airlied> has planet.fd.o stopped updating?

21:39 <zmike> jenatali: windows ci dead?

21:39 <alatiera_afk[m]> zmike: which one

21:39 <alatiera_afk[m]> one machine is down for maintainance

21:39 <jenatali> zmike: not me

21:39 <alatiera_afk[m]> but there's another one

21:40 <zmike> alatiera_afk[m]: https://gitlab.freedesktop.org/mesa/mesa/-/pipelines/957044

21:40 <zmike> jobs not starting

21:40 <zmike> and I'm suspecting it won't

21:40 <alatiera> zmike there's a backlog of jobs probably

21:40 <alatiera> the runner is up and well

21:40 <zmike> :/

21:41 <zmike> is this gstreamer again?

21:41 <zmike> feels like this is happening with some regularity

21:41 <alatiera> only mesa seems to be queuing windows job atm

21:41 <zmike> so private branch jobs are starving the marge jobs?

21:41 <alatiera> in gst we have like 3 windows builds per pipeline

21:43 <alatiera> so looks like there are 4 jobs queued, all of which are trying to build windows images

21:43 <zmike> 🤕

21:43 <alatiera> ah no there are 4 builds jobs as well

21:43 <alatiera> but the image builds hit hard

21:44 <zmike> can we limit image builds to 1 at a time or something?

21:44 <alatiera> they used to be mostly noops but it seems like soemthing has gone wrong lately

21:44 <alatiera> we could but that would block all build jobs

21:44 <alatiera> and they are not resource hungry, just take forever

21:44 <zmike> yeah...

21:44 <zmike> it's obviously not ideal that the marge jobs are blocking because of private image builds

21:46 <jenatali> Something is broken with image caching

21:46 <jenatali> Private builds are not pulling the main containers

21:47 <alatiera> yes I noticed that too

21:48 <alatiera> and it's eating a lot of storage

21:48 <alatiera> I cleaned it yday and overnight it was 1tb of images filled

21:48 <alatiera> but only some of them are different objects

21:48 <alatiera> a lot of them were reused shas

21:51 <zmike> eric_engestrom: btw I forgot to ask earlier: it seems like it should be possible to trigger a timeout on those radv jobs much sooner? cuz they lasted the full hour somehow

21:52 <jenatali> I'm on paternity leave at the moment so I can't dig into what's going on with the pipelines

21:54 <zmike> sounds like a learning opportunity for the little one