| 23 Jan 2025 |
 hexa | nope | 15:31:03 |
| hexa | Jan 23 10:19:24 mimas hydra-queue-runner[2420170]: performing step ‘/nix/store/wi65qgy0plx7s0p32wxqgdh4g97pv2mf-ungoogled-chromium-unwrapped-132.0.6834.83.drv’ 1 times on ‘build@sleepy-brown.builder.nixos.org’ (needed by build 286341474 and 1 others)
Jan 23 10:39:16 mimas hydra-queue-runner[2420170]: performing step ‘/nix/store/a71z2063mlxyzf87vfc7a7qwldpgsw1f-ghc-9.6.4.drv’ 1 times on ‘build@sleepy-brown.builder.nixos.org’ (needed by build 286242596 and 0 others)
Jan 23 11:12:13 mimas hydra-queue-runner[2420170]: performing step ‘/nix/store/2hrdb8fwp6cc81zjxh3d514rgi1xwhjs-circt-1.101.0.drv’ 1 times on ‘build@sleepy-brown.builder.nixos.org’ (needed by build 286339038 and 0 others)
Jan 23 11:19:09 mimas hydra-queue-runner[2420170]: performing step ‘/nix/store/pcw2g2ja229wb7viw64affdbzg928dz2-nixpkgs-release-checks.drv’ 1 times on ‘build@sleepy-brown.builder.nixos.org’ (needed by build 286343041 and 1 others)
Jan 23 11:29:23 mimas hydra-queue-runner[2420170]: performing step ‘/nix/store/d35yzd99xgzshn2mmx45xxhpd8jxbas5-ghc-native-bignum-9.4.5.drv’ 1 times on ‘build@sleepy-brown.builder.nixos.org’ (needed by build 286090073 and 0 others)
Jan 23 11:45:39 mimas hydra-queue-runner[2420170]: performing step ‘/nix/store/g5fqvv4n0slvdf7yvlha0r8ljz8008pc-ungoogled-chromium-unwrapped-132.0.6834.83.drv’ 1 times on ‘build@sleepy-brown.builder.nixos.org’ (needed by build 286209520 and 1 others)
Jan 23 12:13:03 mimas hydra-queue-runner[2420170]: performing step ‘/nix/store/7d179amka4xdyqa4nvkajzhpkzggb4nh-ghc-native-bignum-9.6.4.drv’ 1 times on ‘build@sleepy-brown.builder.nixos.org’ (needed by build 286141208 and 0 others)
Jan 23 12:14:48 mimas hydra-queue-runner[2420170]: performing step ‘/nix/store/z8y8icsg1rphx86f70ihglbbld0rrb5y-ghc-9.4.6.drv’ 1 times on ‘build@sleepy-brown.builder.nixos.org’ (needed by build 286282418 and 0 others)
Jan 23 14:47:11 mimas hydra-queue-runner[2420170]: performing step ‘/nix/store/2bhsqbcr4ckyd5pbdhsshi47lcjwjs7s-nixpkgs-release-checks.drv’ 1 times on ‘build@sleepy-brown.builder.nixos.org’ (needed by build 286357790 and 1 others)
| 15:31:18 |
| hexa | that is like the only matches for sleepy-brown that I have and they line up perfectly well with the nix-daemon interaction | 15:31:36 |
| hexa |
build@sleepy-brown.builder.nixos.org x86_64-linux,i686-linux /var/lib/hydra/queue-runner/.ssh/id_ed25519 6 4 kvm,nixos-test,big-parallel big-parallel
| 15:31:45 |
| hexa | ehhhhhhh | 15:31:47 |
| hexa | the ssh hostkey is now big-parallel or what | 15:31:52 |
| hexa |
-rw-r--r-- 1 root root 2.9K Jan 23 12:03 /etc/nix/machines
| 15:32:03 |
| hexa | fixed | 15:32:24 |
| hexa | fixed the same issue for one of the arm builders | 16:25:53 |
 Vladimír Čunát | No. That was making the scheduler send only big-parallel jobs there. | 17:23:36 |
| Vladimír Čunát | The reason is that otherwise you can waste their slots by hour-running NixOS tests. | 17:24:01 |
| Vladimír Čunát | * The reason is that otherwise you can waste their slots by hour-running NixOS tests. (for example) | 17:24:08 |
| Vladimír Čunát | It's been there for many days. The config certainly did work. | 17:24:55 |
| Vladimír Čunát | I even posted about it here. | 17:26:20 |
| hexa | I don't get it … the field after the system features is the ssh hostkey | 17:29:49 |
| hexa | https://github.com/NixOS/hydra/blob/master/src/hydra-queue-runner/hydra-queue-runner.cc#L170-L175 | 17:30:50 |
| hexa | so foo,bar,big-parallel big-parallel will configure the sshHostKey instead | 17:31:09 |
| hexa | which apparently broke sending jobs that way | 17:31:19 |
| Vladimír Čunát | After "supported features" there's "mandatory features". | 17:31:28 |
| Vladimír Čunát | I do read it there. | 17:31:31 |
| hexa | ah, indeed | 17:31:39 |
| hexa | ok, so it only did big-parallel jobs and that is why it was idle | 17:31:54 |
| hexa | but that is not a great use of the resources 😕 | 17:32:14 |
| Vladimír Čunát | And were there non-big in queue in larger numbers? | 17:32:31 |
| hexa | in queue? maybe. runnable? some | 17:32:55 |
| hexa | restored the previous confirguration and added the appropriate number of dashes | 17:34:56 |
| Vladimír Čunát | I don't have a strong opinion on this, as I posted. | 17:35:19 |
| Vladimír Čunát | But the parallel-make stuff is irritating me for a long time. The inflexibility that's forcing us to do these two types of machines. | 17:36:43 |
| Vladimír Čunát | But it's a bigger problem for people who only have a single (big) machine. | 17:37:57 |
| Vladimír Čunát | It's still making latency worse even for Hydra. | 17:38:29 |
