Hydra | 390 Members | |
| 108 Servers |
| Sender | Message | Time |
|---|---|---|
| 7 Jun 2022 | ||
| (Mind you I'm also a first-time Hydra user, so there might be gaps in my understanding, but so far everything is working in my end.) | 08:10:58 | |
But because the flake completely specifies its inputs, ISTM that you don't need the separate spec.json to declare which inputs you want in NIX_PATH or whatever | 08:11:53 | |
If you have private repos, then the hydra user on your server needs access via SSH. I've just added a special GitLab user and SSH key pair on the server in /var/lib/hydra (I think that was the path). | 08:12:35 | |
In reply to @jackdk:matrix.orgYeah, that sounds about right. I've never used spec.json; feels like hydraJobs make that obsolete for flake projects. | 08:13:16 | |
Yeah it looks like spec.json does everything flakes do, but worse. Consider https://github.com/qfpl/reflex-workshop/blob/master/spec.json : it says "go evaluate release.nix from input reflex-workshop, setting <nixpkgs> and <reflex-workshop> in NIX_PATH. The attrset that the evaluation returns defines the jobs to actually run. | 08:16:50 | |
| Ah yes. Just use the flake output type and enjoy :) | 08:17:27 | |
| Sadly the NixOS tests I've written fail in my Hydra instance and I don't know why. I'll wait and see if anyone answers in the NixOS channel first, then re-post here in full detail. In essence all my tests use a
| 08:19:20 | |
| Locally they're running just fine, so I'm at a loss. It's a flake project with everything pinned down and supposedly reproducible. | 08:19:46 | |
In reply to @roberthensing:matrix.orgit came through ✅ | 13:49:28 | |
| 8 Jun 2022 | ||
| 19:22:46 | ||
| 9 Jun 2022 | ||
| 23:04:16 | ||
| 10 Jun 2022 | ||
I'm trying to create a minimal example for my Hydra/PostgreSQL problem mentioned above now, mainly just using a postgresql_10 service and waiting for the unit to come up. This is what usually seems to fail on my Hydra, but not locally. | 10:31:10 | |
| Hum, of course the minimal example works as expected. The only thing I can even think of is the pin. | 10:58:25 | |
But then it stresses me out that the same nix build runs fine locally. | 10:58:45 | |
| If anyone has any ideas as to what could lead to a NixOS test running locally but not on a Hydra instance (the last thing that comes to my mind would be performance differences), I'd be really happy. I'll put my example on github next week, hopefully a smarter person than me can see what's going on then. The gist of it is that I'm starting the postgres service, wait for it to be up, then change the user password. This works fine. But then doing anything after, like
| 14:17:48 | |
| This works just fine locally though, and I'm at a loss. | 14:18:01 | |