feat(ci): use self-hosted runner for faster main build

[lidel]

The goal is to make it faster, this PR aims to test if using self-hosted runner speeds up release build.

If it is ok, we can decide to set CI_BUILD_RUNS_ON on this repo instead.

cc @galargh for sanity check + if we need to safelist this repo somewhere for this to work (i tried to set CI_BUILD_RUNS_ON but https://github.com/ipfs/distributions/actions/runs/5799417224/job/15719370344 was stuck waiting for free runner and i cancelled it).

[galargh]

cc @galargh for sanity check + if we need to safelist this repo somewhere for this to work (i tried to set CI_BUILD_RUNS_ON but https://github.com/ipfs/distributions/actions/runs/5799417224/job/15719370344 was stuck waiting for free runner and i cancelled it).

Yes, we do! I'll do that first thing tomorrow morning.

[galargh]

We have access to self-hosted runners here now 🥳 but we do need to make some adjustments before we update the CI_BUILD_RUNS_ON var. E.g. https://github.com/ipfs/distributions/actions/runs/5836552866/job/15830374575 - we'll need to update the ipfs setup action so that it works on self-hosted. I'll get back to it next week.

Edit: I fixed the setup ipfs action. We'll see how far it takes us.

Edit2: Looks like we're getting stuck at/after comparing versions with /ipfs/QmQAqFeyrqvLR1m8TWugeQojhzTKp1ASxTicM3sPru1ULF/kubo/versions

[github-actions]

This change produced no new differences in built artifacts.

[lidel]

@galargh rebased, is the issue gone?

[galargh]

Seems so, I think we're good to go 😁 I'll take the vars in a sec.

[galargh]

We can set CI_BUILD_RUNS_ON var to ["self-hosted", "linux", "x64", "2xlarge"] again.

After we merge it, that var will also affect the persist job.

Edit I set the variable.

[github-actions]

This change produced no new differences in built artifacts.

[github-actions]

This change produced no new differences in built artifacts.