Github user kunal642 commented on the issue:
https://github.com/apache/carbondata/pull/2925
@xubo245 I dont think we are supporting thriftServer to be run through
IntelliJ on local. In cluster mode this jar would already be available as part
of spark dependency.
For SDK because it is dependent on all modules therefore httpclient is
required.
@manishgupta88 @kumarvishal09 @jackylk @ravipesala What are your opinions?---
