## Development Task [tests](https://gist.github.com/disksing/a1bf78007b674da7e15b52a9e25eb3a5) show that when there are too many stores in a cluster, `balance-region-scheduler` could have performance issues. - [ ] Improve ruleChecker (waiting regions too long) - [ ] Improve fitRegion - [x] Improve `GetStoreRegionCount` - [ ] Improve the big loop of `BalanceRegionScheduler`
