!RROtHmAaQIkiJzJZZE:nixos.org

NixOS Infrastructure

483 Members
Next Infra call: 2024-07-11, 18:00 CEST (UTC+2) | Infra operational issues backlog: https://github.com/orgs/NixOS/projects/52 | See #infra-alerts:nixos.org for real time alerts from Prometheus.151 Servers

Load older messages


SenderMessageTime
20 Jun 2026
@hexa:lossy.networkhexain that we dont' schedule new jobs beyond 70% cpu pressure10:11:20
@k900:0upti.meK900 I mean like, instead of relying on big-parallel and friends 10:11:30
@k900:0upti.meK900We can probably have better heuristics10:11:36
@hexa:lossy.networkhexaoh, you mean mixed load?10:11:38
@k900:0upti.meK900 I think mixed load is what organically falls out if we start tracking per-job costs 10:12:04
@raitobezarius:matrix.orgraitobezariusRedacted or Malformed Event10:13:49
@hexa:lossy.networkhexayes, it would.10:14:49
@hexa:lossy.networkhexathe scheduler supports custom scheduling functions10:15:01
@hexa:lossy.networkhexahttps://github.com/NixOS/hydra/blob/master/nixos-modules/queue-runner-module.nix#L47-L6310:15:29
@c0d3h01:matrix.orgHarshal joined the room.12:36:12
@k900:0upti.meK900Did someone restart Hydra again13:08:41
@k900:0upti.meK900We will never get a kernel at this rate13:08:43
@hexa:lossy.networkhexa
● hydra-queue-runner-dev.service - Hydra Queue Runner main service
     Loaded: loaded (/etc/systemd/system/hydra-queue-runner-dev.service; enabled; preset: ignored)
     Active: active (running) since Sat 2026-06-20 08:35:31 UTC; 4h 33min ago
13:09:06
@hexa:lossy.networkhexathe builds are still running13:11:47
@hexa:lossy.networkhexathe queue-runner is still doing13:12:00
@k900:0upti.meK900 https://hydra.nixos.org/build/331682971#tabs-buildsteps 13:12:57
@hexa:lossy.networkhexayeah, it did release-26.05 6.18.36 first13:13:19
@hexa:lossy.networkhexatogether with the browsers13:13:26
@hexa:lossy.networkhexaonce that was done it had a free slot for the unstable-small 6.183613:14:26
@k900:0upti.meK900The 1h40 one was running ~3 hours ago13:14:39
@hexa:lossy.networkhexahm13:14:53
@hexa:lossy.networkhexathe build is still running13:16:51
@hexa:lossy.networkhexaI'm not touching anything until it is done13:17:02
@joerg:thalheim.ioMic92 hexa (signing key rotation when): looks like the accept look just exits when the number of connections exceeds a value. 13:17:46
@joerg:thalheim.ioMic92and we have 1024 FDs maximum13:17:53
@hexa:lossy.networkhexalmao13:18:14
@hexa:lossy.networkhexa
# ls /proc/3496116/fd/ | wc -l
401
13:18:52
@joerg:thalheim.ioMic92Deploying a fix now13:21:35
@joerg:thalheim.ioMic92 * hexa (signing key rotation when): looks like the accept loop just exits when the number of connections exceeds a value. 13:21:44
@joerg:thalheim.ioMic92 hexa (signing key rotation when): if you want to wait, can you deploy? 13:22:50

Show newer messages


Back to Room ListRoom Version: 6