Re: [PR] [SPARK-47678][CORE] Check `spark.shuffle.readHostLocalDisk` when reading shuffle blocks [spark]

["viirya (via GitHub)" <gi...@apache.org>]

viirya commented on code in PR #45803:
URL: https://github.com/apache/spark/pull/45803#discussion_r1548303144


##########
core/src/main/scala/org/apache/spark/storage/ShuffleBlockFetcherIterator.scala:
##########
@@ -417,7 +418,7 @@ final class ShuffleBlockFetcherIterator(
         numBlocksToFetch += mergedBlockInfos.size
         localBlocks ++= mergedBlockInfos.map(info => (info.blockId, info.mapIndex))
         localBlockBytes += mergedBlockInfos.map(_.size).sum
-      } else if (blockManager.hostLocalDirManager.isDefined &&
+      } else if (doLocalDiskRead && blockManager.hostLocalDirManager.isDefined &&

Review Comment:
   Hmm, if `SHUFFLE_HOST_LOCAL_DISK_READING_ENABLED` is disabled, I think `BlockManager` won't have `hostLocalDirManager` (i.e., it is `None`), except push-based shuffle is enabled.
   
   But I think push-based shuffle is only enabled under YARN mode. For K8s, it is disabled.
   
   How does it enters this branch if `blockManager.hostLocalDirManager.isDefined` is false?
   
   
   



-- 
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