[ 
https://issues.apache.org/jira/browse/SPARK-32900?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
 ]

Tom van Bussel updated SPARK-32900:
-----------------------------------
    Description: In order to determine whether 
{{UnsafeExternalSorter.SpillableIterator}} has spilled already it checks 
whether {{upstream}} is an instance of {{UnsafeInMemorySorter.SortedIterator}}. 
When radix sorting is used (added by SPARK-14851) and there are NULLs in the 
input however, upstream will be an instance of 
{{UnsafeExternalSorter.ChainedIterator}} instead, but should still be spilled.  
(was: In order to determine whether {{UnsafeExternalSorter.SpillableIterator}} 
has spilled already it checks whether {{upstream}} is an instance of 
{{UnsafeInMemorySorter.SortedIterator}}. When radix sorting is used and there 
are NULLs in the input however, upstream will be an instance of 
{{UnsafeExternalSorter.ChainedIterator}} instead, but should still be spilled.)

> UnsafeExternalSorter.SpillableIterator cannot spill when there are NULLs in 
> the input and radix sorting is used.
> ----------------------------------------------------------------------------------------------------------------
>
>                 Key: SPARK-32900
>                 URL: https://issues.apache.org/jira/browse/SPARK-32900
>             Project: Spark
>          Issue Type: Bug
>          Components: Spark Core
>    Affects Versions: 2.4.7, 3.0.1
>            Reporter: Tom van Bussel
>            Priority: Major
>
> In order to determine whether {{UnsafeExternalSorter.SpillableIterator}} has 
> spilled already it checks whether {{upstream}} is an instance of 
> {{UnsafeInMemorySorter.SortedIterator}}. When radix sorting is used (added by 
> SPARK-14851) and there are NULLs in the input however, upstream will be an 
> instance of {{UnsafeExternalSorter.ChainedIterator}} instead, but should 
> still be spilled.



--
This message was sent by Atlassian Jira
(v8.3.4#803005)

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

Reply via email to