[GitHub] taosaildrone opened a new pull request #23762: [SPARK-21492][SQL][WIP] Memory leak in SortMergeJoin

[GitBox <gi...@apache.org>]

taosaildrone opened a new pull request #23762: [SPARK-21492][SQL][WIP] Memory leak in SortMergeJoin
URL: https://github.com/apache/spark/pull/23762
 
 
   ## What changes were proposed in this pull request?
   
   If SortMergeJoinScanner doesn't consume the iterator from
   UnsafeExternalRowSorter entirely, the memory that
   UnsafeExternalSorter acquired from TaskMemoryManager will not
   be released. This leads to a memory leak, spills, and OOME. A
   page will be held per partition of the unused iterator.
   This patch will allow the SortMergeJoinScanner to explicitly close the iterators (for non-generated code)
   
   ## How was this patch tested?
   
   Manual testing and profiling with scripts in SPARK-21492 comments.
   
   

----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on GitHub and use the
URL above to go to the specific comment.
 
For queries about this service, please contact Infrastructure at:
users@infra.apache.org


With regards,
Apache Git Services

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