spark-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Mala Chikka Kempanna (Jira)" <j...@apache.org>
Subject [jira] [Updated] (SPARK-30225) "Stream is corrupt" exception on reading disk-spilled data of a shuffle operation
Date Wed, 11 Dec 2019 18:58:00 GMT

     [ https://issues.apache.org/jira/browse/SPARK-30225?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]

Mala Chikka Kempanna updated SPARK-30225:
-----------------------------------------
    Summary: "Stream is corrupt" exception on reading disk-spilled data of a shuffle operation
 (was: "stream is corrupt" exception on reading disk-spilled data of a shuffle operation)

> "Stream is corrupt" exception on reading disk-spilled data of a shuffle operation
> ---------------------------------------------------------------------------------
>
>                 Key: SPARK-30225
>                 URL: https://issues.apache.org/jira/browse/SPARK-30225
>             Project: Spark
>          Issue Type: Bug
>          Components: Input/Output
>    Affects Versions: 2.4.0
>            Reporter: Mala Chikka Kempanna
>            Priority: Major
>
> There is issues with spark.unsafe.sorter.spill.read.ahead.enabled in spark 2.4.0, which
is introduced by https://issues.apache.org/jira/browse/SPARK-23366
> Workaround for this problem is to disable readahead of unsafe spill with following.
> --conf spark.unsafe.sorter.spill.read.ahead.enabled=false
> <code>
> 19/12/10 01:51:31 INFO sort.ShuffleExternalSorter: Thread 142 spilling sort data of 5.1
GB to disk (1  time so far)19/12/10 01:51:31 INFO sort.ShuffleExternalSorter: Thread 142
spilling sort data of 5.1 GB to disk (1  time so far)19/12/10 01:52:48 INFO sort.ShuffleExternalSorter:
Thread 142 spilling sort data of 5.1 GB to disk (2  times so far)19/12/10 01:53:53 ERROR
executor.Executor: Exception in task 6.0 in stage 0.0 (TID 6)java.io.IOException: Stream is
corrupted at net.jpountz.lz4.LZ4BlockInputStream.refill(LZ4BlockInputStream.java:202) at net.jpountz.lz4.LZ4BlockInputStream.refill(LZ4BlockInputStream.java:228)
at net.jpountz.lz4.LZ4BlockInputStream.read(LZ4BlockInputStream.java:157) at org.apache.spark.io.ReadAheadInputStream$1.run(ReadAheadInputStream.java:168)
at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1149) at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:624)
at java.lang.Thread.run(Thread.java:748)19/12/10 01:53:53 INFO executor.CoarseGrainedExecutorBackend:
Got assigned task 3319/12/10 01:53:53 INFO executor.Executor: Running task 8.1 in stage 0.0
(TID 33)19/12/10 01:54:00 INFO sort.UnsafeExternalSorter: Thread 142 spilling sort data of
3.3 GB to disk (0  time so far)19/12/10 01:54:30 INFO executor.Executor: Executor is trying
to kill task 8.1 in stage 0.0 (TID 33), reason: Stage cancelled19/12/10 01:54:30 INFO executor.Executor:
Executor killed task 8.1 in stage 0.0 (TID 33), reason: Stage cancelled19/12/10 01:54:52 INFO
executor.CoarseGrainedExecutorBackend: Driver commanded a shutdown
> </code>
>  
>  



--
This message was sent by Atlassian Jira
(v8.3.4#803005)

---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@spark.apache.org
For additional commands, e-mail: issues-help@spark.apache.org


Mime
View raw message