Guide: How to scale NKG for Reliability

[mpstefan]

As a potential user of NKG who operates a large environment
Or as an existing user of NKG who has grown their environment
I want a guide that explains how to scale NKG to reach a consistently higher availability than the default Helm deployment
So that I can continue to use NKG with my higher expectations for availability.

Acceptance

• TBD

[mpstefan]

Blocked by #637

[kate-osborn]

Talked with @bjee19 and we think this would be a good documentation story for him to work on.

[mpstefan]

Right now, one only needs to scale the number of replicas from our Helm chart to scale for reliability. This will likely change in some capacity once we have data and control plane separation. When we implement that feature, we will produce this kind of documentation.