将大型 Spark 数据帧从数据块写入 csv 失败

Posted 2023-03-23

【中文标题】将大型 Spark 数据帧从数据块写入 csv 失败

【英文标题】：Failure in writing large spark dataframes to csv from databricks

【发布时间】：2020-03-24 17:06:08

【问题描述】：

我正在处理数据块中的大型 spark 数据帧，当我尝试将最终数据帧写入 csv 格式时，它给了我以下错误： org.apache.spark.SparkException：作业中止。

#Creating a data frame with entire date seuence for each user
df=pd.DataFrame('transaction_date':dt_range2,'msno':msno1)

from pyspark.sql.types import *

mySchema1 = StructType([StructField("transaction_date", DateType(), True),
                         StructField("msno", StringType(), True)
                         ])

#user wise date sequence conversion to pyspark data frame
df = spark.createDataFrame(df,schema=mySchema1)

#writing to a csv
df.write.csv("dbfs:/mnt/entracermount1/kkbox/Data_test.csv")

#full traceback
Py4JJavaError                             Traceback (most recent call last)
<command-3754966373119602> in <module>
----> 1 df.write.csv("dbfs:/mnt/entracermount1/kkbox/Data_test.csv")

/databricks/spark/python/pyspark/sql/readwriter.py in csv(self, path, mode, compression, sep, quote, escape, header, nullValue, escapeQuotes, quoteAll, dateFormat, timestampFormat, ignoreLeadingWhiteSpace, ignoreTrailingWhiteSpace, charToEscapeQuoteEscaping, encoding, emptyValue)
    930                        charToEscapeQuoteEscaping=charToEscapeQuoteEscaping,
    931                        encoding=encoding, emptyValue=emptyValue)
--> 932         self._jwrite.csv(path)
    933 
    934     @since(1.5)

/databricks/spark/python/lib/py4j-0.10.7-src.zip/py4j/java_gateway.py in __call__(self, *args)
   1255         answer = self.gateway_client.send_command(command)
   1256         return_value = get_return_value(
-> 1257             answer, self.gateway_client, self.target_id, self.name)
   1258 
   1259         for temp_arg in temp_args:

/databricks/spark/python/pyspark/sql/utils.py in deco(*a, **kw)
     61     def deco(*a, **kw):
     62         try:
---> 63             return f(*a, **kw)
     64         except py4j.protocol.Py4JJavaError as e:
     65             s = e.java_exception.toString()

/databricks/spark/python/lib/py4j-0.10.7-src.zip/py4j/protocol.py in get_return_value(answer, gateway_client, target_id, name)
    326                 raise Py4JJavaError(
    327                     "An error occurred while calling 012.\n".
--> 328                     format(target_id, ".", name), value)
    329             else:
    330                 raise Py4JError(

Py4JJavaError: An error occurred while calling o1459.csv.
: org.apache.spark.SparkException: Job aborted.
    at org.apache.spark.sql.execution.datasources.FileFormatWriter$.write(FileFormatWriter.scala:201)
    at org.apache.spark.sql.execution.datasources.InsertIntoHadoopFsRelationCommand.run(InsertIntoHadoopFsRelationCommand.scala:192)
    at org.apache.spark.sql.execution.command.DataWritingCommandExec.sideEffectResult$lzycompute(commands.scala:108)
    at org.apache.spark.sql.execution.command.DataWritingCommandExec.sideEffectResult(commands.scala:106)
    at org.apache.spark.sql.execution.command.DataWritingCommandExec.doExecute(commands.scala:126)
    at org.apache.spark.sql.execution.SparkPlan$$anonfun$execute$1.apply(SparkPlan.scala:146)
    at org.apache.spark.sql.execution.SparkPlan$$anonfun$execute$1.apply(SparkPlan.scala:134)
    at org.apache.spark.sql.execution.SparkPlan$$anonfun$executeQuery$5.apply(SparkPlan.scala:187)
    at org.apache.spark.rdd.RDDOperationScope$.withScope(RDDOperationScope.scala:151)
    at org.apache.spark.sql.execution.SparkPlan.executeQuery(SparkPlan.scala:183)
    at org.apache.spark.sql.execution.SparkPlan.execute(SparkPlan.scala:134)
    at org.apache.spark.sql.execution.QueryExecution.toRdd$lzycompute(QueryExecution.scala:113)
    at org.apache.spark.sql.execution.QueryExecution.toRdd(QueryExecution.scala:113)
    at org.apache.spark.sql.DataFrameWriter$$anonfun$runCommand$1.apply(DataFrameWriter.scala:710)
    at org.apache.spark.sql.DataFrameWriter$$anonfun$runCommand$1.apply(DataFrameWriter.scala:710)
    at org.apache.spark.sql.execution.SQLExecution$$anonfun$withCustomExecutionEnv$1.apply(SQLExecution.scala:111)
    at org.apache.spark.sql.execution.SQLExecution$.withSQLConfPropagated(SQLExecution.scala:240)
    at org.apache.spark.sql.execution.SQLExecution$.withCustomExecutionEnv(SQLExecution.scala:97)
    at org.apache.spark.sql.execution.SQLExecution$.withNewExecutionId(SQLExecution.scala:170)
    at org.apache.spark.sql.DataFrameWriter.runCommand(DataFrameWriter.scala:710)
    at org.apache.spark.sql.DataFrameWriter.saveToV1Source(DataFrameWriter.scala:306)
    at org.apache.spark.sql.DataFrameWriter.save(DataFrameWriter.scala:292)
    at org.apache.spark.sql.DataFrameWriter.save(DataFrameWriter.scala:235)
    at org.apache.spark.sql.DataFrameWriter.csv(DataFrameWriter.scala:698)
    at sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method)
    at sun.reflect.NativeMethodAccessorImpl.invoke(NativeMethodAccessorImpl.java:62)
    at sun.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:43)
    at java.lang.reflect.Method.invoke(Method.java:498)
    at py4j.reflection.MethodInvoker.invoke(MethodInvoker.java:244)
    at py4j.reflection.ReflectionEngine.invoke(ReflectionEngine.java:380)
    at py4j.Gateway.invoke(Gateway.java:295)
    at py4j.commands.AbstractCommand.invokeMethod(AbstractCommand.java:132)
    at py4j.commands.CallCommand.execute(CallCommand.java:79)
    at py4j.GatewayConnection.run(GatewayConnection.java:251)
    at java.lang.Thread.run(Thread.java:748)
Caused by: org.apache.spark.SparkException: Job aborted due to stage failure: Serialized task 64:0 was 614833921 bytes, which exceeds max allowed: spark.rpc.message.maxSize (268435456 bytes). Consider increasing spark.rpc.message.maxSize or using broadcast variables for large values.
    at org.apache.spark.scheduler.DAGScheduler.org$apache$spark$scheduler$DAGScheduler$$failJobAndIndependentStages(DAGScheduler.scala:2360)
    at org.apache.spark.scheduler.DAGScheduler$$anonfun$abortStage$1.apply(DAGScheduler.scala:2348)
    at org.apache.spark.scheduler.DAGScheduler$$anonfun$abortStage$1.apply(DAGScheduler.scala:2347)
    at scala.collection.mutable.ResizableArray$class.foreach(ResizableArray.scala:59)
    at scala.collection.mutable.ArrayBuffer.foreach(ArrayBuffer.scala:48)
    at org.apache.spark.scheduler.DAGScheduler.abortStage(DAGScheduler.scala:2347)
    at org.apache.spark.scheduler.DAGScheduler$$anonfun$handleTaskSetFailed$1.apply(DAGScheduler.scala:1101)
    at org.apache.spark.scheduler.DAGScheduler$$anonfun$handleTaskSetFailed$1.apply(DAGScheduler.scala:1101)
    at scala.Option.foreach(Option.scala:257)
    at org.apache.spark.scheduler.DAGScheduler.handleTaskSetFailed(DAGScheduler.scala:1101)
    at org.apache.spark.scheduler.DAGSchedulerEventProcessLoop.doOnReceive(DAGScheduler.scala:2579)
    at org.apache.spark.scheduler.DAGSchedulerEventProcessLoop.onReceive(DAGScheduler.scala:2527)
    at org.apache.spark.scheduler.DAGSchedulerEventProcessLoop.onReceive(DAGScheduler.scala:2515)
    at org.apache.spark.util.EventLoop$$anon$1.run(EventLoop.scala:49)
    at org.apache.spark.scheduler.DAGScheduler.runJob(DAGScheduler.scala:896)
    at org.apache.spark.SparkContext.runJob(SparkContext.scala:2280)
    at org.apache.spark.sql.execution.datasources.FileFormatWriter$.write(FileFormatWriter.scala:170)
    ... 34 more

即使我尝试另存为镶木地板文件，它也给出了同样的错误。如果您可以为此提出任何解决方案，那就太好了。谢谢。

【问题讨论】：

请将完整的回溯显示为文本，而不是图像

添加了完整的回溯

Serialized task 64:0 was 614833921 bytes, which exceeds max allowed: spark.rpc.message.maxSize (268435456 bytes). Consider increasing spark.rpc.message.maxSize or using broadcast variables for large values

任何解决方案

你能不能试试 df.cache().count() 然后保存文件。还有你用的是哪个java版本？

【参考方案1】：

初始化集群时需要更改 Spark 配置。详情请参考https://kb.databricks.com/execution/spark-serialized-task-is-too-large.html。

问题已解决。 谢谢大家的帮助。