Action times out

[astro]

The hosted runner: GitHub Actions 2 lost communication with the server. Anything in your workflow that terminates the runner process, starves it for CPU/Memory, or blocks its network access can cause this error.

https://github.com/astro/microvm.nix/actions/workflows/flakehub-publish-rolling.yml

[cole-h]

It's possible that evaluation of your flake is too expensive for the type of runner you have setup, seeing that "starves it for CPU/Memory" part of the error message. I just cloned your repo and ran the nix flake show command we run as a part of our evaluation checks in flakehub-push and got this:

$ command time nix flake show --all-systems --json --no-write-lock-file
[...snip...]
23.54user 3.15system 0:27.55elapsed 96%CPU (0avgtext+0avgdata 11787476maxresident)k
28058inputs+257outputs (86major+2945802minor)pagefaults 0swaps

Converting that maxresident value to GiB gives me:

$ qalc '11787476 KiB'
11787476 kibibytes ≈ 12.07037542 GB

And according to these GitHub docs, the Linux runners only have 7GiB:

(from https://docs.github.com/en/actions/using-github-hosted-runners/about-github-hosted-runners/about-github-hosted-runners#supported-runners-and-hardware-resources)

(I admit I don't really know if "resident" memory corresponds to "real" RAM usage, but that seems the most plausible explanation)