module loader: Reduce parallelism to 1/1 in all cases

[radeksimko]

This will cause indexing to take longer to finish on machines with >2 CPUs, but it should flatten the CPU spikes some users have reported in #733

It is worth noting that this puts a bit more pressure on resolving #5 / #150

[dbanck]

LGTM!

Were you able to identify the module loader as source of the high cpu usage from the profiling results?

[radeksimko]

Were you able to identify the module loader as source of the high cpu usage from the profiling results?

In broader sense yes, based on a provided CPU profile from a user I believe the main root cause is the command(s) we execute for each (initialized) module. I reckon the ultimate solution will be sourcing the schemas from a (future) Registry API, which should consume far less CPU than running terraform providers schema -json. That is however more involved solution for long term. I assume that "slowing down" the execution should reduce the aggregate CPU usage at any given time.

[github-actions]

I'm going to lock this pull request because it has been closed for 30 days ⏳. This helps our maintainers find and focus on the active issues.
If you have found a problem that seems related to this change, please open a new issue and complete the issue template so we can capture all the details necessary to investigate further.