| 28 Jul 2023 |
 chayleaf | oh thanks | 15:17:40 |
| chayleaf | and i am using the module but it isnt | 15:17:59 |
 ajs124 | which one? the nixpkgs one or the one from the hydra repo?
I seem to remember them having some differences, but idk if that would be one of them. | 15:19:21 |
| chayleaf | (neither was nix-serve for that matter) | 15:18:31 |
| chayleaf | maybe it's because list settings arent getting merged in whatever freeform type nix.settings is used | 15:20:53 |
| chayleaf | the nixpkgs one | 15:23:58 |
| chayleaf | oh nvm it did get set | 15:24:50 |
| chayleaf | so the issue is something else i guess | 15:24:59 |
| chayleaf | the actual non-cached error is Caught exception in Hydra::Controller::Build->build "[31;1merror:[0m cannot open connection to remote store '[35;1mdaemon[0m': [35;1m[31;1merror:[0m reading from file: [35;1mConnection reset by peer[0m[0m at /nix/store/971v8rm57fb2hpsgzx22df98vm87wqvm-hydra-2023-06-25/libexec/hydra/lib/Hydra/Controller/Build.pm line 83." btw, though I guess that was more or less to be expected | 15:28:11 |
 Rick (Mindavi) | In reply to @andreas.schraegle:helsinki-systems.de
which one? the nixpkgs one or the one from the hydra repo?
I seem to remember them having some differences, but idk if that would be one of them. If you know about any consider making an issue on the nixpkgs tracker | 15:28:48 |
| Rick (Mindavi) | I am running the nixpkgs one | 15:29:16 |
| ajs124 | In reply to @rick:matrix.ciphernetics.nl
If you know about any consider making an issue on the nixpkgs tracker it's probably just outdated information. there used to be a time where certain perl deps also were only part of the hydra flake instead of being in nixpkgs, but I think it's been some years since these issues. | 15:30:33 |
| chayleaf | the symptoms:
- I can see the queue (it says 0/4, but the builds are actually running)
- "running builds" is empty
- "machine status" is idle
- "latest builds" populates after build finish (that's probably fine)
- "latest steps" works perfectly fine too
- build page (like /build/4) returns 500 with the issue above
| 15:31:27 |
| chayleaf | * the symptoms:
- I can see the queue (it says 0/4, but the builds are actually running)
- "running builds" is empty
- "machine status" is idle
- "latest builds" populates after build finish (that's probably fine)
- "latest steps" works perfectly fine too
- build page (like /build/4) returns 500 with the error above
| 15:31:59 |
| ajs124 | does your nix-daemon log anything? | 15:32:32 |
| chayleaf | oh, error: error processing connection: user 'hydra-www' is not allowed to connect to the Nix daemon | 15:33:58 |
| chayleaf | also a bunch of connections are from hydra user | 15:34:31 |
| chayleaf | yeah it works after adding hydra-www to allowed-users | 15:39:42 |
|  @rimuru:gentoo.chat changed their profile picture. | 16:19:54 |
|  @loop:tchncs.de removed their profile picture. | 18:55:48 |
| @loop:tchncs.de removed their display name loop. | 18:57:02 |
| @loop:tchncs.de left the room. | 18:58:07 |
| 29 Jul 2023 |
|  @dylan:fdylan.co.uk joined the room. | 00:33:05 |
|  @miangraham:matrix.org left the room. | 06:19:18 |
 hacker1024 | Is the jobset checkinterval persistent across restarts? I want to make a jobset check weekly. | 12:35:35 |
| ajs124 | I think so, yes | 12:45:47 |
