spark-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Apache Spark (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (SPARK-20414) avoid creating only 16 reducers when calling topByKey()
Date Thu, 20 Apr 2017 15:55:04 GMT

    [ https://issues.apache.org/jira/browse/SPARK-20414?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15976948#comment-15976948
] 

Apache Spark commented on SPARK-20414:
--------------------------------------

User 'yangyangyyy' has created a pull request for this issue:
https://github.com/apache/spark/pull/17697

> avoid creating only 16 reducers when calling topByKey()
> -------------------------------------------------------
>
>                 Key: SPARK-20414
>                 URL: https://issues.apache.org/jira/browse/SPARK-20414
>             Project: Spark
>          Issue Type: Improvement
>          Components: MLlib
>    Affects Versions: 1.5.2, 1.6.0, 1.6.1, 1.6.2, 1.6.3, 2.0.0, 2.0.1, 2.0.2, 2.1.0
>            Reporter: Yang Yang
>            Priority: Minor
>
> currently in the MLlib topByKey() function, it directly calls aggregateByKey(), which
by default uses very few partitions/reducers, in my experience I see only 16 reducers for
a 100GB input.
> the aggregateByKey() has an optional reducer count, adding this option to the top level
topByKey()



--
This message was sent by Atlassian JIRA
(v6.3.15#6346)

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


Mime
View raw message