himanshug opened a new pull request #8412: skip unnecessary aggregate(..) calls 
with LimitedBufferHashGrouper
URL: https://github.com/apache/incubator-druid/pull/8412
 
 
   ### Description
   
   This is a short patch removing unnecessary work done during execution of 
groupBy query with limit push down. When limit is pushed down to merge phase in 
the groupBy query processing , `aggregate(..)` is called for all aggregators 
whenever a new bucket is added to the underlying hash table. This is 
unnecessary when `sortHasNonGroupingFields=false` and newly added bucket did 
not find a place in the min-max-heap .  With `sortHasNonGroupingFields=false`, 
that row's position in overall ordering can never change so it is ok to skip 
the `aggregate(..)` calls. this improves performance for cases where you have 
large number of buckets with multiple complex aggregators.
   
   No behavior change is introduced in this PR.
   
   <hr>
   
   This PR has:
   - [X] been self-reviewed.
   - [X] been tested in a test Druid cluster.
   <hr>
   

----------------------------------------------------------------
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:
[email protected]


With regards,
Apache Git Services

---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Reply via email to