| 13 Sep 2021 |
 Vladimír Čunát | I'm not sure, maybe the do. Or some of them at least. | 16:19:16 |
| Vladimír Čunát | Well, I do know that the t4b builder does not 😇 but that's not a significant data point. | 16:20:52 |
| Vladimír Čunát | Thanks to that it doesn't seem to run nix affected by the bug. Perhaps the hash-named builders do follow some nixpkgs channel automatically 🤷 | 16:25:38 |
| Vladimír Čunát | * Thanks to that it doesn't seem to run nix affected by the bug. Perhaps the hash-named builders do follow some nixpkgs channel automatically 🤷 (when they get re-deployed?) | 16:25:58 |
 baloo | the one running over at packet/equinix metal are running on the spot market | 16:27:10 |
| baloo | so they get redeployed every day from my understanding | 16:27:27 |
| Vladimír Čunát | No... that wouldn't be good. Some builds take many hours (e.g. chromium). | 16:28:56 |
| Vladimír Čunát | On wendy it could be even more than one day, but the hash-based machines are faster. | 16:29:32 |
| Vladimír Čunát | * On wendy it could take even more than one day, but the hash-based machines are faster. | 16:29:46 |
| 14 Sep 2021 |
 @grahamc:nixos.org | they do get deployed every day, but they go in to a drain phase and wait for jobs to complete before shutting down | 14:32:25 |
| @grahamc:nixos.org | * they do get terminated and recreated every day, but they go in to a drain phase and wait for jobs to complete before shutting down | 14:33:13 |
 lukegb (he/him) | What config do they get recreated from? Some pinned flake somewhere? | 14:42:57 |
| lukegb (he/him) | Presumably not the nixos-org-configurations repo? | 14:43:08 |
 andi- | They are netbooted from an image that is on a netboot host. Not exactly sure which repo that lives in. | 15:25:33 |
|  moritz.hedtke joined the room. | 17:15:31 |
|  Andreas Schrägle joined the room. | 19:36:09 |
| Vladimír Čunát | I think the evaluator got stuck: https://hydra.nixos.org/evals | 21:11:42 |
| 15 Sep 2021 |
| lukegb (he/him) | In reply to @vcunat:matrix.org
I think the evaluator got stuck: https://hydra.nixos.org/evals hmm, staging-next just got an eval | 00:30:12 |
|  tomberek joined the room. | 01:01:30 |
| 17 Sep 2021 |
|  cleverca22 joined the room. | 06:01:14 |
| lukegb (he/him) | mac2-guest might be wedged again: https://hydra.nixos.org/build/153432095 | 19:25:18 |
 sterni | next evaluations should fix the tarball jobs, but we should keep an eye on them since they timed out a few times which worries me | 20:30:03 |
| 21 Sep 2021 |
 toonn | I was looking for a good Hydra evaluation to rebase a PR on to target staging but the listed evaluations are all pretty old. Should I using nixpkgs:staging-next instead? | 08:16:18 |
| Vladimír Čunát | Depends what you want from the commit. | 08:17:24 |
| Vladimír Čunát | staging-small jobset gives some basic binaries (and tests a few packages) | 08:17:53 |
| Vladimír Čunát | staging-next has full rebuilds, and at this moment is in quite a good state (not too many build regressions, rebuilds finished). | 08:18:45 |
| 26 Sep 2021 |
| lukegb (he/him) | a lot of the x86_64-darwin builders appear to be working on builds that haven't completed in 8 days, and some of them have been "sending inputs" for 7h+ | 16:01:01 |
| 28 Sep 2021 |
| lukegb (he/him) | https://cloudblogs.microsoft.com/opensource/2021/09/28/announcing-azure-credits-for-open-source-projects/ even more hydra builders? :P | 21:16:40 |
| 29 Sep 2021 |
| sterni | what do you think about a label for recurring PRs (staging-next, haskell-updates, ...)? These can be a bit of a pain to find if you want to use the PR view | 11:58:47 |
| sterni | opening the branch and clicking on the PR kinda works, but it hasn't displayed properly for me before | 11:59:09 |
