MacOS CI repeated fails

[thomas-bc]

F´ Version
Affected Component

---

Problem Description

MacOS CI has been failing repeatedly for no apparent reason, timing out after 6 hours. See for example
https://github.com/nasa/fprime/actions/runs/7302456221/job/19901093748?pr=2454
https://github.com/nasa/fprime/actions/runs/7294730948/job/19901273821?pr=2451

[LeStarch]

This problem started around the time we corrected fprime-util to build with parallelism properly. This implies that the higher number of parallel jobs causes the issue.

As noted by @thomas-bc: actions/runner#1326 seems to be related. There is a lot of speculation that it could be an OOM or large memory usage that causes the issue. We know FPP (Jav)a uses much memory, as does gcc/g++ so running many parallel jobs could strain the memory usage.

I would recommend we do the following:

1. Specifically limit macOS jobs to a single job or small number of jobs
2. Add a timeout as suggested there to something more normal

Looking for build-parallelism issues with random numbers of jobs could be restricted to Ubuntu as a matter of practicality.