I’ve managed to finally set up Hydra and 2 jobsets for it that build my configs (I want to set up a cache once everything’s working). Everything is in a flake and working but after setting everything up I found out that Hydra doesn’t support propagating CI/CD results to Forgejo/Gitea/GitHub with flakes.
So I need to make a legacy jobset that will either load the jobset from my flake directly or will create jobs that will build the outputs of the flake.
The jobs are currently set as follows:
hydraJobs = {
nixosConfigurations = builtins.listToAttrs (map (name: {
name = name;
value = (import ./hosts.nix {inherit nixpkgs inputs outputs;}).${name}.config.system.build.toplevel;
}) (builtins.attrNames (import ./hosts.nix {inherit nixpkgs inputs outputs;})));
homeManagerConfigurations = builtins.listToAttrs (map (name: {
name = name;
value = (import ./homes.nix {inherit nixpkgs home-manager inputs outputs;}).${name}.activationPackage;
}) (builtins.attrNames (import ./homes.nix {inherit nixpkgs home-manager inputs outputs;})));
};
I’d appreciate any help creating these jobs as I wasn’t able to find much about how to work with Hydra and it’s jobs/jobsets.