Github user ravipesala commented on the issue:
https://github.com/apache/carbondata/pull/2148
@jackylk But with sparks RPC endpoint framework it is dependent on spark,
so is it possible for carbon to have separate cluster without running sparks
executors?---
