attilapiros commented on a change in pull request #25299: [SPARK-27651][Core]
Avoid the network when shuffle blocks are fetched from the same host
URL: https://github.com/apache/spark/pull/25299#discussion_r319966632
##########
File path: core/src/main/scala/org/apache/spark/network/BlockDataManager.scala
##########
@@ -22,16 +22,22 @@ import scala.reflect.ClassTag
import org.apache.spark.TaskContext
import org.apache.spark.network.buffer.ManagedBuffer
import org.apache.spark.network.client.StreamCallbackWithID
-import org.apache.spark.storage.{BlockId, StorageLevel}
+import org.apache.spark.storage.{BlockId, ShuffleBlockId, StorageLevel}
private[spark]
trait BlockDataManager {
+ /**
+ * Interface to get host-local shuffle block data. Throws an exception if
the block cannot be
Review comment:
Disk persisted host-local RDD blocks are already supported for details see
https://github.com/apache/spark/pull/25299.
I think using the exact type `ShuffleBlockId` here is better as it is
checked by compile time and describes the intended usage of this method better.
As SPARK-27651 is already merged I won't expect any upcoming change here
regarding host local block access.
----------------------------------------------------------------
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.
For queries about this service, please contact Infrastructure at:
[email protected]
With regards,
Apache Git Services
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
