🤔 Hydra.nixos.org is now sometimes throwing eval errors like

{UNKNOWN}: aggregate job ‘tested’ failed with the error: nixos.tests.installer.swraid.x86_64-linux: error: cannot connect to socket at '/nix/var/nix/gc-socket/socket': No such file or directory
 at /nix/store/j96zrgb48aiwxysh1vhq1zp1zv4da29r-hydra-perl-deps/lib/perl5/site_perl/5.34.1/Catalyst/Model/DBIC/Schema.pm line 526

In reply to @terwiz:matrix.org
Ouch. Well, the server is in AWS and NixOS was not on the premade image list (surprise, surprise). and the server is up and running already. This setup was/is just fine for doing manual builds, and those are still happening, so can't just reinstall everything.
With dockerized hydra I suppose the main issue was the /nix/store setup. Also, the nixos/hydra docker image is actually ubuntu + nix and not nixos. So I was also finding it difficult to configure, as all the instructions are for setting etc/nixos/configuration.nix and you don't have that in ubuntu + nix. That's also why I asked for pointers to docs, I expect the same issue without the docker...

32mP:88|F:1|R:0|T:0] Events: 2909 
The following jobs failed:
+--------------------------------------+------------------------------+
| Job ID                               | Test File                    |
+--------------------------------------+------------------------------+
| 3446CDB2-7E2E-11ED-A687-F3FB89213249 | Hydra/Controller/User/ldap.t |
+--------------------------------------+------------------------------+

                                   Yath Result Summary
-----------------------------------------------------------------------------------------
     Fail Count: 1
     File Count: 89
Assertion Count: 1368
      Wall Time: 165.97 seconds
       CPU Time: 999.09 seconds (usr: 1.65s | sys: 0.27s | cusr: 839.02s | csys: 158.15s)
      CPU Usage: 601%
    -->  Result: FAILED  <--

In reply to @grahamc:nixos.org
anyone else find it generally annoying that hydra only shows the first piece of a remote's hostname?