[autoscaler] Default min and max workers to 0 for the head node type.

[DmitriGekhtman]

What is the problem?

Ray version and other system information (Python version, TensorFlow version, OS):

Most use-cases require a head node with a config distinct from the workers.
Having to explicitly state that there are no workers of head node type is confusing.
The default min and max workers for head node type should be 0. Head node type min and max workers should be removed from all cluster configs and from the helm chart.

Document the change and log a warning if necessary, as, strictly speaking, this change is backwards-incompatible.

Reproduction (REQUIRED)

Please provide a short code snippet (less than 50 lines if possible) that can be copy-pasted to reproduce the issue. The snippet should have no external library dependencies (i.e., use fake or mock data / environments):

If the code snippet cannot be run by itself, the issue will be closed with "needs-repro-script".

• I have verified my script runs in a clean environment and reproduces the issue.
• I have verified the issue also occurs with the latest wheels.

[richardliaw]

@DmitriGekhtman as per request from Zhe @zhe-thoughts , can you make sure this issue gets associated with the serverless team (maybe the milestone)?

[DmitriGekhtman]

Sure, milestone applied!

[AmeerHajAli]

Thanks @DmitriGekhtman for filing this.
I don't have context. But why do we have to make this change? won't this make it even harder to move away from head node or treating it in a special way?
Also, are we having a specific use case/user that we are optimizing for?
Why is this backward incompatible? can't we translate old yamls?