Hydra | 370 Members | |
| 110 Servers |
| Sender | Message | Time |
|---|---|---|
| 25 Sep 2024 | ||
| There was no build to begin with | 13:57:33 | |
| ah right, just saw that after asking, thx | 13:57:52 | |
| just for context i'm thinking of building a matrix bot that periodically fetches job pages for selected packages, and notifies subscribers in case of failures | 13:59:17 | |
| Make sure to rate-limit your bot. Maybe the nixpkgs-review-tools data is public? | 15:02:22 | |
| Would be great if we could not all scrape hydra 😁 | 15:02:39 | |
| I managed to get myself blocked for a short while by overloading hydra | 15:03:04 | |
| 26 Sep 2024 | ||
| 01:16:06 | ||
| 27 Sep 2024 | ||
| 00:18:20 | ||
| re: this: https://github.com/NixOS/hydra/blob/95003f2eb503f71979856470c7caea946f1ae7f0/doc/manual/src/notifications.md is this a) idk if hydra-dev is actually prod | 08:43:42 | |
| It's deployed but turned off | 08:44:00 | |
| And the reason it's turned off is because it takes up stupid amounts of RAM and no one has figured out why | 08:44:14 | |
| (or as far as I can tell attempted to) | 08:44:18 | |
| And that machine is already severely RAM starved | 08:44:45 | |
| Because it is running nixpkgs evals | 08:44:50 | |
| The other reason is that we're not aware of any use for it at hydra.nixos.org | 09:16:32 | |
| Honestly I don't think anyone would have cared about it | 09:17:22 | |
| If it didn't use half the RAM on the box | 09:17:26 | |
In reply to @vcunat:matrix.orgwouldn't it be useful for notifying maintainers of build failures, with mechanisms other than email? | 13:13:35 | |
| Probably. But I believe there's no code for that. Even this code was unreliable, sending mass notifications in some edge cases IIRC. | 13:46:04 | |
| IIRC in the last corner of my brain I also saved that those mails where put into the spam directory | 14:44:32 | |
| is a URL like this one supposed to take me to the last successful build, or the latest build? | 14:48:55 | |
| looking at e.g. https://hydra.nixos.org/job/nixpkgs/trunk/rippled.x86_64-linux/:
| 14:50:03 | |
| 16:09:06 | ||
| 16:27:32 | ||
| btw i think the something equivalent to the job endpoint does exist:
contrary to what the docs say, there does seem to be a whole api endpoint at a separate url than the webpage (see above) | 20:04:00 | |
| * btw i think that something functionally equivalent to a job endpoint does exist:
contrary to what the docs say, there does seem to be a whole api endpoint at a separate url than the webpage (see above) | 20:04:14 | |
| * btw i think that something functionally equivalent to a job endpoint does exist:
contrary to what the docs say, there does seem to be a whole api endpoint at a separate url than the webpage (see above) | 20:04:28 | |
| * btw i think that something functionally equivalent to a job endpoint does exist:
contrary to what the docs say, there does seem to be a whole api endpoint at a separate url than the webpage (see above) | 20:07:57 | |
| 28 Sep 2024 | ||
| 06:03:22 | ||
In reply to @asymmetric:matrix.dapp.org.uki guess what i'm asking is, can i rely on this endpoint sticking around, or is it part of some private/unstable API that will be removed? | 10:43:40 | |