[GitHub] [spark] ChenMichael commented on a change in pull request #34684: [SPARK-37442][SQL] InMemoryRelation statistics bug causing broadcast join failures with AQE enabled

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

ChenMichael commented on a change in pull request #34684:
URL: https://github.com/apache/spark/pull/34684#discussion_r756308545



##########
File path: sql/core/src/main/scala/org/apache/spark/sql/execution/columnar/InMemoryRelation.scala
##########
@@ -259,6 +259,12 @@ case class CachedRDDBuilder(
       rowCountStats.add(batch.numRows)
       batch
     }.persist(storageLevel)

Review comment:
       I thought `BlockManager.getStatus` was expensive. The docs are also indicating that it should mainly be used for testing. Is it ok to do so in `isCachedColumnBuffersLoaded`? 
   My concern with only checking one partition of the RDD is that it's still possible to run into this problem if the partition is small and finishes soon before the other partitions. Then, it is still possible to run into this problem. 




-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: reviews-unsubscribe@spark.apache.org

For queries about this service, please contact Infrastructure at:
users@infra.apache.org



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