[
https://issues.apache.org/jira/browse/SPARK-10303?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Josh Rosen updated SPARK-10303:
---
Description:
See attached screenshot. For a job that uses SQLContext's JSON reader, the RDD
DAG visualization shows that Spark is using an inefficient version of Union
which chains together many UnionRDDs as opposed to the more-efficient form of
UnionRDD that accepts a larger list of child RDDs.
!screenshot-1.png!
was:
See attached screenshot. For a job that uses SQLContext's JSON reader, the RDD
DAG visualization shows that Spark is using an inefficient version of Union
which chains together many UnionRDDs as opposed to the more-efficient form of
UnionRDD that accepts a larger list of child RDDs.
> Spark SQL JSON Reader uses inefficient form of Union operation
> --
>
> Key: SPARK-10303
> URL: https://issues.apache.org/jira/browse/SPARK-10303
> Project: Spark
> Issue Type: Bug
> Components: SQL
>Affects Versions: 1.5.0
>Reporter: Josh Rosen
> Attachments: screenshot-1.png
>
>
> See attached screenshot. For a job that uses SQLContext's JSON reader, the
> RDD DAG visualization shows that Spark is using an inefficient version of
> Union which chains together many UnionRDDs as opposed to the more-efficient
> form of UnionRDD that accepts a larger list of child RDDs.
> !screenshot-1.png!
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
-
To unsubscribe, e-mail: issues-unsubscr...@spark.apache.org
For additional commands, e-mail: issues-h...@spark.apache.org
