Github user jihoonson commented on the pull request:
https://github.com/apache/tajo/pull/1003#issuecomment-212704759
I did basic tests to verify query results, but there is one thing I'm
concerned with. That is the shuffle cost. This patch enables the parallel
execution of non-leaf EBs, thereby more shuffle operations can be executed
simultaneously. It can increase the shuffle cost especially for range shuffle
due to heavy index search.
So, we need to improve hash shuffle by reusing http connections and range
shuffle by redesigning the index structure.
However, I think the heavy shuffle costs should be addressed in other
Jiras. What do you think?
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at [email protected] or file a JIRA ticket
with INFRA.
---
