feat: Custom Balancer

[metcoder95]

• Add histograms to workers - feat: add histogram to workers #619
• Expose workers lifecycle events - feat: add state properties to worker #620
• Expose pool lifecycle events - feat: pool events for workers #624
• Decouple handling of when new worker available and when a worker has finished a task
• Abstract event handling through AsyncResourcePool
• Decouple LB from scheduler (LB should return either worker or null)
• Implement LB branching
  • LB on new worker
  • LB once worker finished task
  • LB on new task

[metcoder95]

I'm doubtful about wether or not pass the workers on every call or make the balancer rely directly on the events.

By passing them on every call, balancer will always have the up-to-date list of workers, but might cause them to need to constantly adapt and lack possibly the state.

On the event based approach they get definitely more control; maybe keep both?

[metcoder95]

@ronag @jerome-benoit this should be ready for a review for landing; so far the balancer has been decoupled from an scheduler.

I'll update docs and more later this week

[jerome-benoit]

@ronag @jerome-benoit this should be ready for a review for landing; so far the balancer has been decoupled from an scheduler.

Great!

I do not think I will be able to do a review before next week ...

[metcoder95]

No rush! I'll proceed with the documentation this week

[ronag]

Suggested change

	if (task.abortSignal) {
	if (task.abortSignal?.aborted) {

?

[metcoder95]

We do that check already in upper layers, if it reaches here, it means the task is not aborted yet.

[ronag]

RSLGTM. A bit too much for me to get a good overview. Would be great with some docs and samples. I think sorting out the public API is higher prio than the implementation details.

[metcoder95]

SGTM, I'll start preparing either Tomorrow or later in the weekend