Github user ravipesala commented on the issue:
https://github.com/apache/carbondata/pull/3066
> i think the query performance of carbon_solution is lower than
hive_solution's, because carbon_solution has more segment (insert generates a
segment and update generates more segment)
> Do we have some method to optimize this?
Since we are updating the existing data it creates extra files like delete
delta and incremental carbondata files It may degrade query performance a
little but when do the compaction it will get improved.
---
