Github user squito commented on a diff in the pull request:

    https://github.com/apache/spark/pull/6648#discussion_r32026272
  
    --- Diff: core/src/main/scala/org/apache/spark/storage/BlockId.scala ---
    @@ -56,18 +56,27 @@ case class RDDBlockId(rddId: Int, splitIndex: Int) 
extends BlockId {
     // Format of the shuffle block ids (including data and index) should be 
kept in sync with
     // 
org.apache.spark.network.shuffle.ExternalShuffleBlockResolver#getBlockData().
     @DeveloperApi
    -case class ShuffleBlockId(shuffleId: Int, mapId: Int, reduceId: Int) 
extends BlockId {
    -  override def name: String = "shuffle_" + shuffleId + "_" + mapId + "_" + 
reduceId
    +case class ShuffleBlockId(shuffleId: Int, mapId: Int, reduceId: Int, 
stageAttemptId: Int)
    +    extends BlockId {
    +  override def name: String = {
    +    "shuffle_" + shuffleId + "_" + mapId + "_" + reduceId + "_" + 
stageAttemptId
    +  }
     }
     
     @DeveloperApi
    -case class ShuffleDataBlockId(shuffleId: Int, mapId: Int, reduceId: Int) 
extends BlockId {
    -  override def name: String = "shuffle_" + shuffleId + "_" + mapId + "_" + 
reduceId + ".data"
    +case class ShuffleDataBlockId(shuffleId: Int, mapId: Int, stageAttemptId: 
Int, reduceId: Int)
    +    extends BlockId {
    +  override def name: String = {
    +    "shuffle_" + shuffleId + "_" + mapId + "_" + reduceId + "_" + 
stageAttemptId + ".data"
    +  }
     }
     
     @DeveloperApi
    -case class ShuffleIndexBlockId(shuffleId: Int, mapId: Int, reduceId: Int) 
extends BlockId {
    -  override def name: String = "shuffle_" + shuffleId + "_" + mapId + "_" + 
reduceId + ".index"
    +case class ShuffleIndexBlockId(shuffleId: Int, mapId: Int, stageAttemptId: 
Int, reduceId: Int)
    +    extends BlockId {
    +  override def name: String = {
    +    "shuffle_" + shuffleId + "_" + mapId + "_" + reduceId + "_" + 
stageAttemptId + ".index"
    +  }
    --- End diff --
    
    These changes to the block ids are the "important" changes -- 
`DiskBlockManager` uses the blockId to determine the filename, and now it 
includes the `stageAttemptId`, so each one gets a different file


---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at [email protected] or file a JIRA ticket
with INFRA.
---

---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Reply via email to