[server] Add resource-based KV leader replica capacity control

[swuferhong]

Search before asking

• I searched in the issues and found nothing similar.

Motivation

Fluss currently relies mainly on max.bucket.num and max.partition.num to limit table metadata size. These limits do not directly protect the cluster from creating too many KV leader replicas, which can consume resident memory on TabletServers and increase OOM risk.

We need a coordinator-side capacity guard that can estimate the total KV leader replica capacity from TabletServer memory and reject new KV table/partition creation when the cluster no longer has enough capacity.

Solution

No response

Anything else?

No response

Willingness to contribute

• I'm willing to submit a PR!