Clear separation between api.ClusterConfig and api.ClusterProvider

[errordeveloper]

Description

This refactors how we separate provider configuration and it's recievers
from cluster configuration. We no longer copy cluster configuration in
every places where we need it, instead we pass it explicitly every time.

Checklist

• Code compiles correctly (i.e make build)
• Added tests that cover your change (if possible)
• All tests passing (i.e. make test)
• Added/modified documentation as required (such as the README)
• Added yourself to the humans.txt file

[errordeveloper]

Few things to tidy up...

[errordeveloper]

We still keep a pointer here, but it's arguably more sensible, as there aren't many different kinds of calls to stack manager from the outside, so it's sensible to set it once. It basically gets handed off at this point, and the stack manager is coupled the cluster itself, there isn't a good reason to have it decoupled at this point, unlike the instance of the provider.

[richardcase]

This generally looks good to me, great work. Just a couple of comments.

[richardcase]

We may need some guidance for contributors when to split out / duplicate fields like we have done with region. I get why its done but its probably something we need to keep an eye on just in case there is a proliferation .

[errordeveloper]

Yes, I agree. I wanted to avoid Profile() here, but it's used somewhere and I didn't manage to get rid of it in a sensible manner. We need some guidance here indeed.

[richardcase]

LGTM!