You are viewing a plain text version of this content. The canonical link for it is here.
Posted to reviews@spark.apache.org by GitBox <gi...@apache.org> on 2021/04/04 10:29:12 UTC

[GitHub] [spark] wangyum edited a comment on pull request #31024: [SPARK-33979][SQL] Reorder predicate

wangyum edited a comment on pull request #31024:
URL: https://github.com/apache/spark/pull/31024#issuecomment-812971948


   > Since this changes the TPCDS plan golden file, do we have some TPCDS benchmark results?
   
   In fact, it will not have significant improvement in TPC-DS query, because the filter conditions of the TPC-DS query are very simple.
   
   ---
   
   This is a real query from cluster. The order of filter conditions for this query is:
   1. LIKE ANY with 70 values
   2. In predicate with 702 values
   3. CAST(D.AUCT_END_DATE as DATE)>=current_date-65
   
   We will reorder the filter conditions with this optimizer:
   1. CAST(D.AUCT_END_DATE as DATE)>=current_date-65
   2. In predicate with 702 values
   3. LIKE ANY with 70 values
   
   
   
   Before this pr | After this pr
   -- | --
   More than 72 min | 13 min 
   
   
   
   ![image](https://user-images.githubusercontent.com/5399861/113498923-f32b9800-9543-11eb-8e8c-a89e22c7528f.png)
   


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

For queries about this service, please contact Infrastructure at:
users@infra.apache.org



---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscribe@spark.apache.org
For additional commands, e-mail: reviews-help@spark.apache.org