daniels changed the topic of #freedesktop to: https://www.freedesktop.org infrastructure and online services || for questions about freedesktop.org projects, please see each project's contact || for discussions about specifications, please use https://gitlab.freedesktop.org/xdg or xdg@lists.freedesktop.org
dcunit3d has quit [Remote host closed the connection]
Haaninjo has joined #freedesktop
krastevm has joined #freedesktop
MajorBiscuit has joined #freedesktop
blatant has joined #freedesktop
bmodem has quit [Ping timeout: 480 seconds]
<Kayden>
bentiss: thanks for restarting mesa!25020 :)
<Kayden>
looks like it worked this time
<bentiss>
Kayden: yeah, it was taking too long because it was rebuilding the lava images
<Kayden>
ah
dcunit3d has joined #freedesktop
bmodem has joined #freedesktop
<bentiss>
daniels: so the "new-and-failing" cluster is completely dead. it really didn't like me trying to migrate it away from coreos, and is having network issues everywhere. So I'm going to wipe it, meaning that we won't be able to recover the data from s3.fd.o from the past month that are not in the old s3 (now currently in use)
<bentiss>
daniels: but that also means that we probably should send regular backups to GCS in case our ceph dies like here
<bentiss>
daniels: also I pushed yesterday an update on helm-gitlab-infra: fdo-infra.py is now at the root, and also has commands like `./fdo-infra.py ssh fdo-k3s-server-15` or `./fdo-infra.py console fdo-k3s-server-15`
<bentiss>
this just outputs what ssh command needs to be called, and I just call `$(./fdo-infra.py ssh fdo-k3s-server-15)`. No need to remember which user/port needs to be used, it's all handled by the script (using the customdata on equinix)
<bentiss>
I also have a handy "fdo-infra.py reinstall" command to refresh the server :)
<bentiss>
ouch. found out why the new cluster was not in a good shape (now that I wiped it): 2 nodes were having the same IP
<daniels>
bentiss: ahhh ugh, that's a real mess, but at least the mystery is solved
<daniels>
bentiss: don't worry about s3 - everything on there is designed to be ephemeral and recoverable so we should be ok :)
<daniels>
thankyou again
<bentiss>
yeah, mystery solved is really good. Compared to not understanding why coreos is failing, we can still move forward :)
<daniels>
\o/
<mupuf>
bentiss: wait, what's the resolution here? It wasn't coreos, it was just the whole cluster being broken?
<bentiss>
mupuf: so coreos was broken (random timeouts). So I tried to switch the new coreos cluster in live to debian. ANd when I did that, I messed up the fact that 2 nodes where having the same IP
<mupuf>
oh, that could explain some networking issues, indeed :D
<mupuf>
but was this mistake done during the switch from coreos to debian, or was it there for a while?
<bentiss>
mupuf: it was from yesterday evening
<bentiss>
(when s3 started to fail)
blatant has quit [Ping timeout: 480 seconds]
<mupuf>
ack :)
<mupuf>
so, the original network issue is still TBD
jkhsjdhjs has quit [Remote host closed the connection]
<mupuf>
maybe I need to start using a newer version of registry.freedesktop.org/freedesktop/ci-templates/x86_64/container-build-base:
<mupuf>
oh, wait! Of course, I see what the problem is
<mupuf>
sorry for the noise!
pulsar12 has joined #freedesktop
<bentiss>
mupuf: glad I helped :)
blatant has joined #freedesktop
thaller has quit [Ping timeout: 480 seconds]
psykose has quit [Remote host closed the connection]
psykose has joined #freedesktop
bmodem has quit [Ping timeout: 480 seconds]
cstub has joined #freedesktop
pulsar12 has quit [Read error: Connection reset by peer]
bmodem has joined #freedesktop
lack has quit [Read error: Connection reset by peer]
ximion has joined #freedesktop
lack has joined #freedesktop
cstub has quit []
ximion has quit [Quit: Detached from the Matrix]
mripard has quit [Remote host closed the connection]
ximion has joined #freedesktop
ximion has quit [Quit: Detached from the Matrix]
mripard has joined #freedesktop
mirai_ has quit [Remote host closed the connection]
mirai_ has joined #freedesktop
psukys has quit [Ping timeout: 480 seconds]
imm has joined #freedesktop
mripard has quit [Quit: mripard]
mripard has joined #freedesktop
<imm>
hi, I think I've found a bug in X. Can anyone point me towards where to report it? I can reliably make my entire desktop hang just by running 'scrot' with a breakpoint added in gdb, so that libX11 doesn't send a response that the server apparently expects?
imm has quit [Remote host closed the connection]
colemickens_ has joined #freedesktop
colemickens_ has left #freedesktop [#freedesktop]
<MrCooper>
for the entertainment of the channel, it's not a bug, it's a feature called server grabs
bmodem has quit [Ping timeout: 480 seconds]
<sam_>
lol
<DavidHeidelberg[m]>
mupuf: I think I cannot reasonably easily generate from the gfx-ci/linux repo, but I could set up latest cpio archive and just replace the modules.. not solution I love, but it'll fulfill purpose
<mupuf>
???
<mupuf>
The modules are stored in their own cpio
<mupuf>
If that's what you want, it should be simple to reproduce
<mupuf>
The issue is that you'll want to add firmware too
<mupuf>
Unlike arm, amd needs it
<DavidHeidelberg[m]>
I probably don't want to touch firmwares, also you bundle some boot2contrainer stuff there, which I would have to manually pull from boot2container
<mupuf>
Boot2container.cpio does not contain any module or fw
imm has joined #freedesktop
<mupuf>
We have linux_x86_64.{modules,firmware,headers,dtbs}.cpio.xz
<mupuf>
Same for all the other architectures
imm has quit [Remote host closed the connection]
<DavidHeidelberg[m]>
where is it created?
<mupuf>
b2c's makefile can generates these
<mupuf>
But their interface is unrelated to b2c
<DavidHeidelberg[m]>
as I'm grepping for cpio, I don't see anything in b2c repo
<DavidHeidelberg[m]>
*anything related to mod,fw,headers,dtbs
<mupuf>
make linux LINUX_SRC=/path/to/linux FEATURES=common,gpu,serial_consoles,...
<mupuf>
Are you looking at gfx-ci/boot2container or mupuf boot2container