Description
Hi @arunoda,
first of all thank you for creating this awesome package. You are doing a really great job!
I recently went to production and started evaluating load-balancing strategies and I settled for cluster because it is the cheapest and most efficient I could find.
When I was reading through the docs I first had some issues understanding how exactly cluster works with regards to "load balancing". I think my confusion came from the fact that the environment variable which allows an instance to handle ddp-traffix for other instances is called CLUSTER_BALANCER_URL. Also the textual descriptions refer to the server becoming a balancer by setting the before mentioned environment variable.
As far as I understand each instance automatically becomes a "balancer" when it uses the cluster package and there are instances which declare that they can take ddp-traffic from other instances.
So setting the CLUSTER_BALANCER_URL does not make a node a "balancer" but makes it a "worker" in the distributed balancing scheme.
So every instance is a "balancer", but not every instance has to be a "worker".
Does that make any sense?
If so, I would suggest renaming CLUSTER_BALANCER_URL and the respective js-api to CLUSTER_DDP_WORKER_URL.