You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@spark.apache.org by "Kent Yao (Jira)" <ji...@apache.org> on 2025/04/09 06:44:00 UTC
[jira] [Resolved] (SPARK-51512) Filter out null MapStatus when cleaning up shuffle data with ExternalShuffleService
[ https://issues.apache.org/jira/browse/SPARK-51512?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Kent Yao resolved SPARK-51512.
------------------------------
Fix Version/s: 4.0.0
Resolution: Fixed
Issue resolved by pull request 50277
[https://github.com/apache/spark/pull/50277]
> Filter out null MapStatus when cleaning up shuffle data with ExternalShuffleService
> -----------------------------------------------------------------------------------
>
> Key: SPARK-51512
> URL: https://issues.apache.org/jira/browse/SPARK-51512
> Project: Spark
> Issue Type: Task
> Components: Spark Core
> Affects Versions: 4.0.0
> Reporter: Wan Kun
> Assignee: Wan Kun
> Priority: Minor
> Labels: pull-request-available
> Fix For: 4.0.0
>
>
> When the application crashes unexpectedly, the registered map statuses may contain null values, therefore we should skip these null values when cleaning up shuffle data with ExternalShuffleService.
> {code:java}
> 25/03/14 15:41:35 ERROR ContextCleaner: Error cleaning shuffle 4
> org.apache.spark.SparkException: Exception thrown in awaitResult:
> at org.apache.spark.util.SparkThreadUtils$.awaitResult(SparkThreadUtils.scala:53)
> at org.apache.spark.util.ThreadUtils$.awaitResult(ThreadUtils.scala:342)
> at org.apache.spark.rpc.RpcTimeout.awaitResult(RpcTimeout.scala:75)
> at org.apache.spark.rpc.RpcEndpointRef.askSync(RpcEndpointRef.scala:101)
> at org.apache.spark.rpc.RpcEndpointRef.askSync(RpcEndpointRef.scala:85)
> at org.apache.spark.storage.BlockManagerMaster.removeShuffle(BlockManagerMaster.scala:204)
> at org.apache.spark.shuffle.sort.io.LocalDiskShuffleDriverComponents.removeShuffle(LocalDiskShuffleDriverComponents.java:47)
> at org.apache.spark.ContextCleaner.doCleanupShuffle(ContextCleaner.scala:241)
> at org.apache.spark.ContextCleaner.$anonfun$keepCleaning$3(ContextCleaner.scala:203)
> at org.apache.spark.ContextCleaner.$anonfun$keepCleaning$3$adapted(ContextCleaner.scala:196)
> at scala.Option.foreach(Option.scala:437)
> at org.apache.spark.ContextCleaner.$anonfun$keepCleaning$1(ContextCleaner.scala:196)
> at org.apache.spark.util.Utils$.tryOrStopSparkContext(Utils.scala:1383)
> at org.apache.spark.ContextCleaner.org$apache$spark$ContextCleaner$$keepCleaning(ContextCleaner.scala:190)
> at org.apache.spark.ContextCleaner$$anon$1.run(ContextCleaner.scala:80)
> Caused by: java.lang.NullPointerException: Cannot invoke "org.apache.spark.scheduler.MapStatus.location()" because "mapStatus" is null
> at org.apache.spark.storage.BlockManagerMasterEndpoint.$anonfun$removeShuffle$3(BlockManagerMasterEndpoint.scala:423)
> at scala.collection.ArrayOps$.foreach$extension(ArrayOps.scala:1323)
> at org.apache.spark.storage.BlockManagerMasterEndpoint.$anonfun$removeShuffle$2(BlockManagerMasterEndpoint.scala:421)
> at org.apache.spark.storage.BlockManagerMasterEndpoint.$anonfun$removeShuffle$2$adapted(BlockManagerMasterEndpoint.scala:420)
> at org.apache.spark.ShuffleStatus.$anonfun$withMapStatuses$1(MapOutputTracker.scala:435)
> at org.apache.spark.ShuffleStatus.withReadLock(MapOutputTracker.scala:72)
> at org.apache.spark.ShuffleStatus.withMapStatuses(MapOutputTracker.scala:435)
> at org.apache.spark.storage.BlockManagerMasterEndpoint.$anonfun$removeShuffle$1(BlockManagerMasterEndpoint.scala:420)
> at org.apache.spark.storage.BlockManagerMasterEndpoint.$anonfun$removeShuffle$1$adapted(BlockManagerMasterEndpoint.scala:419)
> at scala.Option.foreach(Option.scala:437)
> at org.apache.spark.storage.BlockManagerMasterEndpoint.org$apache$spark$storage$BlockManagerMasterEndpoint$$removeShuffle(BlockManagerMasterEndpoint.scala:419)
> at org.apache.spark.storage.BlockManagerMasterEndpoint$$anonfun$receiveAndReply$1.applyOrElse(BlockManagerMasterEndpoint.scala:201)
> at org.apache.spark.rpc.netty.Inbox.$anonfun$process$1(Inbox.scala:104)
> at org.apache.spark.rpc.netty.Inbox.safelyCall(Inbox.scala:216)
> at org.apache.spark.rpc.netty.Inbox.process(Inbox.scala:101)
> at org.apache.spark.rpc.netty.MessageLoop.org$apache$spark$rpc$netty$MessageLoop$$receiveLoop(MessageLoop.scala:76)
> at org.apache.spark.rpc.netty.MessageLoop$$anon$1.run(MessageLoop.scala:42)
> at java.base/java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1144)
> at java.base/java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:642)
> at java.base/java.lang.Thread.run(Thread.java:1583)
> {code}
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@spark.apache.org
For additional commands, e-mail: issues-help@spark.apache.org