[
https://issues.apache.org/jira/browse/SPARK-34788?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17304581#comment-17304581
]
wuyi commented on SPARK-34788:
------------------------------
[~dongjoon] I have pasted the error stack. It happened when we reading the
shuffle data. The cause only says the file is not found, but we believe it's
highly possible due to the disk full issue after investigation.
> Spark throws FileNotFoundException instead of IOException when disk is full
> ---------------------------------------------------------------------------
>
> Key: SPARK-34788
> URL: https://issues.apache.org/jira/browse/SPARK-34788
> Project: Spark
> Issue Type: Improvement
> Components: Shuffle, Spark Core
> Affects Versions: 3.2.0
> Reporter: wuyi
> Priority: Major
>
> When the disk is full, Spark throws FileNotFoundException instead of
> IOException with the hint. It's quite a confusing error to users：
> {code:java}
> 9/03/26 09:03:45 ERROR ShuffleBlockFetcherIterator: Failed to create input
> stream from local block
> java.io.IOException: Error in reading
> FileSegmentManagedBuffer{file=/local_disk0/spark-c2f26f02-2572-4764-815a-cbba65ddb315/executor-b4b76a4c-788c-4cb6-b904-664a883be1aa/blockmgr-36804371-24fe-4131-a3dc-00b7f98f3a3e/11/shuffle_113_1029_0.data,
> offset=110254956, length=1875458}
> at
> org.apache.spark.network.buffer.FileSegmentManagedBuffer.createInputStream(FileSegmentManagedBuffer.java:111)
> at
> org.apache.spark.storage.ShuffleBlockFetcherIterator.next(ShuffleBlockFetcherIterator.scala:442)
> at
> org.apache.spark.storage.ShuffleBlockFetcherIterator.next(ShuffleBlockFetcherIterator.scala:64)
> at scala.collection.Iterator$$anon$12.nextCur(Iterator.scala:434)
> at scala.collection.Iterator$$anon$12.hasNext(Iterator.scala:440)
> at scala.collection.Iterator$$anon$11.hasNext(Iterator.scala:408)
> at
> org.apache.spark.util.CompletionIterator.hasNext(CompletionIterator.scala:31)
> at
> org.apache.spark.InterruptibleIterator.hasNext(InterruptibleIterator.scala:37)
> at scala.collection.Iterator$$anon$11.hasNext(Iterator.scala:408)
> at
> org.apache.spark.sql.catalyst.expressions.GeneratedClass$GeneratedIteratorForCodegenStage2.sort_addToSorter_0$(Unknown
> Source)
> at
> org.apache.spark.sql.catalyst.expressions.GeneratedClass$GeneratedIteratorForCodegenStage2.processNext(Unknown
> Source)
> at
> org.apache.spark.sql.execution.BufferedRowIterator.hasNext(BufferedRowIterator.java:43)
> at
> org.apache.spark.sql.execution.WholeStageCodegenExec$$anonfun$11$$anon$1.hasNext(WholeStageCodegenExec.scala:622)
> at
> org.apache.spark.sql.execution.aggregate.SortAggregateExec$$anonfun$doExecute$1$$anonfun$3.apply(SortAggregateExec.scala:98)
> at
> org.apache.spark.sql.execution.aggregate.SortAggregateExec$$anonfun$doExecute$1$$anonfun$3.apply(SortAggregateExec.scala:95)
> at
> org.apache.spark.rdd.RDD$$anonfun$mapPartitionsWithIndexInternal$1$$anonfun$12.apply(RDD.scala:839)
> at
> org.apache.spark.rdd.RDD$$anonfun$mapPartitionsWithIndexInternal$1$$anonfun$12.apply(RDD.scala:839)
> at
> org.apache.spark.rdd.MapPartitionsRDD.compute(MapPartitionsRDD.scala:60)
> at org.apache.spark.rdd.RDD.computeOrReadCheckpoint(RDD.scala:340)
> at org.apache.spark.rdd.RDD.iterator(RDD.scala:304)
> at
> org.apache.spark.rdd.MapPartitionsRDD.compute(MapPartitionsRDD.scala:60)
> at org.apache.spark.rdd.RDD.computeOrReadCheckpoint(RDD.scala:340)
> at org.apache.spark.rdd.RDD.iterator(RDD.scala:304)
> at
> org.apache.spark.rdd.MapPartitionsRDD.compute(MapPartitionsRDD.scala:60)
> at org.apache.spark.rdd.RDD.computeOrReadCheckpoint(RDD.scala:340)
> at org.apache.spark.rdd.RDD.iterator(RDD.scala:304)
> at
> org.apache.spark.scheduler.ShuffleMapTask.runTask(ShuffleMapTask.scala:99)
> at
> org.apache.spark.scheduler.ShuffleMapTask.runTask(ShuffleMapTask.scala:55)
> at org.apache.spark.scheduler.Task.doRunTask(Task.scala:139)
> at org.apache.spark.scheduler.Task.run(Task.scala:112)
> at
> org.apache.spark.executor.Executor$TaskRunner$$anonfun$13.apply(Executor.scala:497)
> at org.apache.spark.util.Utils$.tryWithSafeFinally(Utils.scala:1432)
> at org.apache.spark.executor.Executor$TaskRunner.run(Executor.scala:503)
> at
> java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1149)
> at
> java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:624)
> at java.lang.Thread.run(Thread.java:748)
> Caused by: java.io.FileNotFoundException:
> /local_disk0/spark-c2f26f02-2572-4764-815a-cbba65ddb315/executor-b4b76a4c-788c-4cb6-b904-664a883be1aa/blockmgr-36804371-24fe-4131-a3dc-00b7f98f3a3e/11/shuffle_113_1029_0.data
> (No such file or directory)
> at java.io.FileInputStream.open0(Native Method)
> at java.io.FileInputStream.open(FileInputStream.java:195)
> at java.io.FileInputStream.<init>(FileInputStream.java:138)
> at
> org.apache.spark.network.buffer.FileSegmentManagedBuffer.createInputStream(FileSegmentManagedBuffer.java:100)
> ... 35 more{code}
> And there's probably a way to detect the disk full: when we get
> `FileNotFoundException`, we try
> [http://weblog.janek.org/Archive/2004/12/20/ExceptionWhenWritingToAFu.html]
> to see if SyncFailedException throws. If SyncFailedException throws, then we
> throw IOException with the disk full hint.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscr...@spark.apache.org
For additional commands, e-mail: issues-h...@spark.apache.org
