Hydra | 394 Members | |
| 111 Servers |
| Sender | Message | Time |
|---|---|---|
| 9 Feb 2022 | ||
| I'm already 20% through copy-pasting ldif files around | 16:12:01 | |
| I could abort if kanidm is easy :) | 16:12:09 | |
I was 5% tempted to make subs to create users and groups and instead I think I'll just have the test writer copy paste their own ldif around | 16:13:37 | |
In reply to @grahamc:nixos.orgGood opportunity to test kanidm, it looks simple, but only speaks LDAP to a certain degree | 16:37:30 | |
| There is an open pr that i did a first review on a few days ago | 16:38:18 | |
| Kanidm people are on matrix btw and pretty helpful | 16:40:00 | |
| https://matrix.to/#/!fzbFDgtoaOQXSSxOZc:gitter.im?via=matrix.org&via=gitter.im&via=lossy.network | 16:40:28 | |
| I'm already so far in to getting slapd going | 16:45:58 | |
| hydra-queue-runner might need some poking for connections handling. It seems that the max_db_connections as a result also ends up limiting the concurrency of the steps of a build. I had a build stalled for 30m with no new output, I suspect because I set the max_db_connections too low | 16:49:43 | |
In reply to @amanda:camnet.siteIn any case, that default should be looked at, judging from the sample config postgres's default is 100 | 17:32:52 | |
| das_j: https://github.com/NixOS/hydra/blob/13b6308581a36b52e92221273f533b3844469def/t/Hydra/Controller/User/ldap.t | 19:37:52 | |
Is it possible that the localStore check is broken? I can see the variable being referenced in templates but not in perl code: [% IF localStore %] | 23:45:25 | |
| 10 Feb 2022 | ||
| das_j: it is indeed broken: https://github.com/NixOS/hydra/commit/44ef743d847ee62188cc3e40ca6f93493499dd07 | 00:40:11 | |
| das_j: https://github.com/NixOS/hydra/pull/1151/commits/42a7f5283c68641f3c4f37283d3d3047bd0718bb | 02:25:28 | |
| I don't like perl still, but at least I can get something done and not feel too bad about it | 02:33:27 | |
In reply to @grahamc:nixos.orgAh great, that looks better than the old approach | 08:37:32 | |
| we should get the hydra docs published somewhere | 14:06:29 | |
grahamc (he/him): It'd be nice to serve it under $HYDRA_URL/docs so everybody can have the docs matching their version | 14:07:15 | |
| oh that'd be great | 14:07:24 | |
| something about this PR seems to be making the tests take 6h+, except there doesn't seem to be any log output for the last 5h50m | 14:09:40 | |
| 14:09:52 | |
| github actions were a mistake | 14:10:27 | |
| oh it fails in hydra too | 14:11:04 | |
| * ~github actions were a mistake~ I take it back, GHA | 14:11:18 | |
| maybe it has to do with this:
| 14:15:51 | |
races :( | 15:07:40 | |
| I'm using hydraUnstable from release-21.11 and some pages like /build/2 are awfully slow | 15:29:08 | |
| loading in like 15-25 seconds | 15:29:13 | |
| while the machine and the database are mostly idle | 15:29:33 | |
how do C++ feel about std::mutex ? https://github.com/NixOS/hydra/pull/1152 | 16:05:24 | |