> Please help me on this....let me know you need other info.
Are the ORC tables fully compacted? Looks like you're running a version of Hive-ACID, which does not perform well without compacting delta files. dfs -ls <path>; should tell you whether there are any delta_* files in the list. > | ACID table:true > | > | alias:tuning_dd_key > | Other than that, it does look like you have only 1 shuffle and if you did run this using Tez, I recommend using <https://github.com/apache/tez/tree/master/tez-tools/swimlanes> to find out the slowest task & find more information about it. You will get a diagram which looks like this <http://www.slideshare.net/t3rmin4t0r/tez8-ui-walkthrough/20> The longest bar of that is the slowest task. I have another version of it, which is unreleased yet (is a bit hard to explain) giving an image which looks like <http://people.apache.org/~gopalv/q21_suppliers_who_kept_orders_waiting.svg > <https://github.com/t3rmin4t0r/tez-swimlanes/blob/master/vertex.py> which is better at finding the 1-2 reducers skewing in a large DAG like q21 [1]. Cheers, Gopal [1] - <http://people.apache.org/~gopalv/tpch-plans/q21_suppliers_who_kept_orders_ waiting.svg>
