Reduce the data skew within some range in a running cluster

## Feature Request

### Describe your feature request related problem

We have region scatter to make the leader/region distribution balance before we import the data. But consider that as time goes by, the distribution of some table may be highly skewed. Even if we have scatter range scheduler which can reduce the skew of data distribution, it will conflict with the balance region scheduler which may cause the redundant scheduling.

### Describe the feature you'd like

A possible way is that utilizing the scatter range scheduler but exclude the range handled by scatter range scheduler in balance region scheduler.

## Score

- 3000

## Mentor(s)

- @lhy1024
- @rleungx

### Recommended Skills
- Go