spark-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Hyukjin Kwon (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (SPARK-26381) Pickle Serialization Error Causing Crash
Date Wed, 19 Dec 2018 07:32:00 GMT

    [ https://issues.apache.org/jira/browse/SPARK-26381?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16724750#comment-16724750
] 

Hyukjin Kwon commented on SPARK-26381:
--------------------------------------

[~ryan.clancy], please provide the codes to reproduce.

> Pickle Serialization Error Causing Crash
> ----------------------------------------
>
>                 Key: SPARK-26381
>                 URL: https://issues.apache.org/jira/browse/SPARK-26381
>             Project: Spark
>          Issue Type: Bug
>          Components: PySpark
>    Affects Versions: 2.3.1, 2.4.0
>         Environment: Tested on two environments:
>  * Spark 2.4.0 - single machine only
>  * Spark 2.3.1 - YARN installation with 5 nodes and files on HDFS
> The error occurs in both environments.
>            Reporter: Ryan
>            Priority: Major
>
> There is a pickle serialization error when I try and use AllenNLP for doing NER within
a Spark worker - it is causing a crash. When running on just the Spark driver or in a standalone
program, everything works as expected.
>  
> {code:java}
> Caused by: org.apache.spark.api.python.PythonException: Traceback (most recent call last):

>  File "/data/disk12/yarn/local/usercache/raclancy/appcache/application_1543437939000_1040/container_1543437939000_1040_01_000002/pyspark.zip/pyspark/worker.py",
line 217, in main 
>    func, profiler, deserializer, serializer = read_command(pickleSer, infile) 
>  File "/data/disk12/yarn/local/usercache/raclancy/appcache/application_1543437939000_1040/container_1543437939000_1040_01_000002/pyspark.zip/pyspark/worker.py",
line 61, in read_command 
>    command = serializer.loads(command.value) 
>  File "/data/disk12/yarn/local/usercache/raclancy/appcache/application_1543437939000_1040/container_1543437939000_1040_01_000002/pyspark.zip/pyspark/serializers.py",
line 559, in loads 
>    return pickle.loads(obj, encoding=encoding) 
> TypeError: __init__() missing 3 required positional arguments: 'non_padded_namespaces',
'padding_token', and 'oov_token' 
>        at org.apache.spark.api.python.BasePythonRunner$ReaderIterator.handlePythonException(PythonRunner.scala:298)

>        at org.apache.spark.api.python.PythonRunner$$anon$1.read(PythonRunner.scala:438)

>        at org.apache.spark.api.python.PythonRunner$$anon$1.read(PythonRunner.scala:421)

>        at org.apache.spark.api.python.BasePythonRunner$ReaderIterator.hasNext(PythonRunner.scala:252)

>        at org.apache.spark.InterruptibleIterator.hasNext(InterruptibleIterator.scala:37)

>        at scala.collection.Iterator$class.foreach(Iterator.scala:893) 
>        at org.apache.spark.InterruptibleIterator.foreach(InterruptibleIterator.scala:28)

>        at scala.collection.generic.Growable$class.$plus$plus$eq(Growable.scala:59)

>        at scala.collection.mutable.ArrayBuffer.$plus$plus$eq(ArrayBuffer.scala:104)

>        at scala.collection.mutable.ArrayBuffer.$plus$plus$eq(ArrayBuffer.scala:48)

>        at scala.collection.TraversableOnce$class.to(TraversableOnce.scala:310)

>        at org.apache.spark.InterruptibleIterator.to(InterruptibleIterator.scala:28)

>        at scala.collection.TraversableOnce$class.toBuffer(TraversableOnce.scala:302)

>        at org.apache.spark.InterruptibleIterator.toBuffer(InterruptibleIterator.scala:28)

>        at scala.collection.TraversableOnce$class.toArray(TraversableOnce.scala:289)

>        at org.apache.spark.InterruptibleIterator.toArray(InterruptibleIterator.scala:28)

>        at org.apache.spark.rdd.RDD$$anonfun$collect$1$$anonfun$12.apply(RDD.scala:939)

>        at org.apache.spark.rdd.RDD$$anonfun$collect$1$$anonfun$12.apply(RDD.scala:939)

>        at org.apache.spark.SparkContext$$anonfun$runJob$5.apply(SparkContext.scala:2074)

>        at org.apache.spark.SparkContext$$anonfun$runJob$5.apply(SparkContext.scala:2074)

>        at org.apache.spark.scheduler.ResultTask.runTask(ResultTask.scala:87) 
>        at org.apache.spark.scheduler.Task.run(Task.scala:109) 
>        at org.apache.spark.executor.Executor$TaskRunner.run(Executor.scala:345)

>        at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1149)

>        at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:624)

>        ... 1 more
> {code}



--
This message was sent by Atlassian JIRA
(v7.6.3#76005)

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


Mime
View raw message