| 12 Sep 2025 |
 Vladimír Čunát | Hydra the SW is coupled to the Nix version it's using. (though I don't know how much) | 17:28:06 |
 K900 | Fairly heavily | 17:29:05 |
| K900 | But also RAM is not an issue | 17:29:05 |
| K900 | Generally | 17:29:06 |
| K900 | (and eval is generally not an issue most of the time) | 17:29:30 |
| Vladimír Čunát | We don't seem to be really running into RAM shortage, though there's not a huge margin sometimes. (zram serves as a buffer) | 17:29:34 |
| Vladimír Čunát | * We don't seem to be really running into RAM shortage on evaluations, though there's not a huge margin sometimes. (zram serves as a buffer) | 17:29:50 |
| Vladimír Čunát | * We don't seem to be really running into RAM shortage on Hydra evaluations, though there's not a huge margin sometimes. (zram serves as a buffer) | 17:29:56 |
 @wolfgangwalther:matrix.org | CI's Eval also got faster with 2.30, although I don't know how much that was related to RAM usage and hitting swap otherwise. | 17:31:51 |
 Philip Taron (UTC-8) | What's the unit in these Hydra metrics for "allocations"? https://hydra.nixos.org/job/nixpkgs/trunk/metrics/metric/nix-env.qa.allocations | 18:03:45 |
| Philip Taron (UTC-8) | Number of mallocs? | 18:04:00 |
| K900 | Yes, but also it's not really a relevant metric | 18:04:57 |
| Philip Taron (UTC-8) | As in, when it goes up or goes down, not really an impact on stuff? | 18:05:28 |
| K900 | Yes | 18:06:09 |
