[jira] [Commented] (SPARK-11055) Use mixing hash-based and sort-based aggregation in TungstenAggregationIterator

["Apache Spark (JIRA)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/SPARK-11055?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14952368#comment-14952368 ] 

Apache Spark commented on SPARK-11055:
--------------------------------------

User 'viirya' has created a pull request for this issue:
https://github.com/apache/spark/pull/9067

> Use mixing hash-based and sort-based aggregation in TungstenAggregationIterator
> -------------------------------------------------------------------------------
>
>                 Key: SPARK-11055
>                 URL: https://issues.apache.org/jira/browse/SPARK-11055
>             Project: Spark
>          Issue Type: Improvement
>          Components: SQL
>            Reporter: Liang-Chi Hsieh
>
> In TungstenAggregationIterator we switch to sort-based aggregation when we can't allocate more memory for hashmap.
> However, using external sorter-based aggregation will write too much key-value pairs into disk. We should use mixing hash-based and sort-based aggregation to reduce the key-value pairs needed to write to disk.



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)

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