| 26 Jun 2026 |
 Mic92 | hexa (signing key rotation when): I won't be around much today. I just finished the backport of the nix temproot gc fix and found a small race for uploads on the hydra side: https://github.com/NixOS/infra/pull/1104 | 07:47:19 |
| Mic92 | The nix fix very much looks like what we are seeing in the logs | 07:48:16 |
 dramforever | beware of appendPatch problems on darwin with unsandboxed builds https://github.com/NixOS/nix/issues/14751 | 07:50:51 |
| dramforever | some nix tests don't like the build dir path having a + in it | 07:51:16 |
| dramforever | maybe also sandboxed build with lix i'm not sure | 07:51:36 |
| Mic92 | dramforever: we build all machines in CI, so if it's an issue, we should see it? | 07:52:44 |
|  ccicnce113424 joined the room. | 08:13:10 |
 Vladimír Čunát | Compression should happen on individual builders now (with the new queue-runner). While it's possible that the aarch64-linux machine compresses slower, the length here is suspicious. | 08:14:33 |
| Vladimír Čunát | * Compression should happen on individual builders now (with the new queue-runner). While it's possible that the aarch64-linux machine compresses faster, the length here is suspicious. | 08:20:20 |
| Vladimír Čunát | Ah, the machine is 10x overloaded again, so maybe that's slowing down the compression. | 08:21:57 |
 Sergei Zimmerman (xokdvium) | In reply to @dramforever:matrix.org
beware of appendPatch problems on darwin with unsandboxed builds https://github.com/NixOS/nix/issues/14751 Those were fixed with build dir changes | 08:23:06 |
| Sergei Zimmerman (xokdvium) | drv name is no longer in the build dir name | 08:23:23 |
 Grimmauld (any/all) | more openmp things? Is there a list of things it is currently attempting to build eating all the cpu? | 08:23:54 |
| Vladimír Čunát | It's honestly tiring to chase individual packages not respecting such constraints, as we have way too many in nixpkgs. | 08:25:06 |
 jappie | you can view a list of jobs running on each machine here: https://hydra.nixos.org/machines | 08:25:08 |
 Yorusaka Miyabi | sleepy-brown: idle | 08:25:46 |
| Yorusaka Miyabi | 🤔 | 08:25:48 |
| Vladimír Čunát | But right now it's all python stuff again, apparently. | 08:25:54 |
| Grimmauld (any/all) | A global OMP_NUM_THREADS = 1 default may make sense after all... | 08:25:58 |
| Grimmauld (any/all) | Anything pulling numpy/scipy is potentially doing OMP | 08:26:10 |
| Vladimír Čunát | That one is reserved for big-parallel build.s | 08:26:17 |
| Vladimír Čunát | * That one is reserved for big-parallel builds. | 08:26:25 |
| Vladimír Čunát | * But right now it's all python processes again, apparently. | 08:27:41 |
 hexa |
Evaluation running since: 5h ago
| 08:50:04 |
| hexa | hrm hrm | 08:50:07 |
| Vladimír Čunát | No, it's not really running. | 08:50:52 |
| hexa | hydra 1061238 3.5 1.9 5161216 5115028 ? S 03:20 11:33 /nix/store/6plwsm6pkq79yjv4xvy8csk2pd4hzr67-perl-5.42.0/bin/perl /nix/store/sbz0i4y97gslxvb2yx7qyvwfy743wvkl-hydra-0.1.20260625.39515ec/bin/.hydra-eval-jobset-wrapped nixpkgs unstable
| 08:51:04 |
| Vladimír Čunát | https://github.com/NixOS/hydra/issues/1812 | 08:51:13 |
| Vladimír Čunát | It might be just an UI issue. | 08:51:20 |
| hexa | but the process is actually running? | 08:51:44 |
