[
https://issues.apache.org/jira/browse/SPARK-19655?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15873130#comment-15873130
]
Herman van Hovell commented on SPARK-19655:
-------------------------------------------
We current only push filters and columns down into the data source. Other
things, like aggregation, are not supported.
What you are seeing is that spark pushes the filter and a dummy field (count
doesn't require anything else) into data source, the aggregation is still done
on the spark side.
> select count(*) , requests 1 for each row
> -----------------------------------------
>
> Key: SPARK-19655
> URL: https://issues.apache.org/jira/browse/SPARK-19655
> Project: Spark
> Issue Type: Bug
> Components: SQL
> Affects Versions: 2.1.0
> Reporter: hosein
> Priority: Minor
>
> when I want query select count( * ) by JDBC and monitor queries in database
> side, I see spark requests: select 1 for destination table
> it means 1 for each row and it is not optimized
--
This message was sent by Atlassian JIRA
(v6.3.15#6346)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
