[14:29:25]
<Aleks (he/him/il/lui)> working on it since like 4 days ...
[14:29:32]
<Aleks (he/him/il/lui)> idk, we're facing like 5 different absolutely nonsensical issue
[14:29:34]
<Aleks (he/him/il/lui)> like basically lxc/lxd is not reliable
[14:29:42]
<Aleks (he/him/il/lui)> or at least that installation ...
[14:29:45]
<Aleks (he/him/il/lui)> or could be a hardware issue idk
[14:29:48]
<Aleks (he/him/il/lui)> right now we should probably just redeploy on a completely different machine
[14:29:56]
<Aleks (he/him/il/lui)> actually yesterday night it seemed like the CI was sort of stable again ... but with only 1 worker enabled ... so I re-enabled 3 workers and then everything exploded again
[14:30:01]
<Aleks (he/him/il/lui)> today I tried purging lxd and reinstalling lxd from scratch on the machine but that doesnt seem to help ...
[14:30:56]
<Mateusz Szymański> How much strain is put on CI? Perhaps having 1 node working would be better for now?
[14:31:18]
<Mateusz Szymański> Have no experience with LXC so I can't be much of help with that :/
[14:31:35]
<Aleks (he/him/il/lui)> yeah but it's a bit ridiculous, there's like 64GB ram + 16 vCPU on the machine
[14:31:40]
<Aleks (he/him/il/lui)> it should be perfectly capable of running several workers in parallel
[14:31:54]
<Aleks (he/him/il/lui)> and it was for quite a time actually
[14:32:05]
<Aleks (he/him/il/lui)> soooo could be a kernel issue, a bunch of bugs in recent LXD version, idk ...
[14:32:15]
<Mateusz Szymański> > <@Alekswag:matrix.org> yeah but it's a bit ridiculous, there's like 64GB ram + 16 vCPU on the machine
I'm not saying it shouldn't run more nodes, just that band aid might be good enough for now
[14:32:28]
<Aleks (he/him/il/lui)> hmyeah
[14:32:45]
<Aleks (he/him/il/lui)> well now it's again miserably crashing trying to restore lxc snapshot : https://ci-apps-dev.yunohost.org/ci/job/5595
[14:33:03]
<Aleks (he/him/il/lui)> didnt see that one in like 3 days
[14:33:10]
<Aleks (he/him/il/lui)> and suddently it reappears
[14:44:58]
<Aleks (he/him/il/lui)> this is really a waste of time, we should just reinstall on another machine -_-
[16:08:51]
<Dante> Hi everyone 🙂
Since I worked on maintaining and implementing the config panel in mautrix_whatsapp app, I thought that it would be a good idea to include `yq` either as deps of the packages that use it (like every app that has a YAML file as config file) or as Yunohost core dependency since YAML seems to be pretty widespread.
For now the alternative is to use `sed` which works fine but is much slower to implement and it is much bug prone in my opinion, also much less intuitive. What are your thoughts about it? 😄