Spark 2.4.1 无法从 HDFS 读取 Avro 文件

Question

我有一个简单的代码块来编写然后读取数据帧作为 Avro 格式。由于 Avro 库已经内置在 Spark 2 中。4.x、

Avro 文件写入成功，文件在HDFS 中生成。但是，当我读取文件时抛出 AbstractMethodError 异常。任何人都可以分享我一些光吗？

我通过在我的 Zeppelin nodebook Spark 解释器中添加包 org.apache.spark:spark-avro_2.11:2.4.1 来使用 Spark 内部库。

我的简单代码块：

%pyspark

test_rows = [ Row(file_name = "test-guangzhou1", topic='camera1', timestamp=1, msg="Test1"),  Row(file_name = "test-guangzhou1", topic='camera1', timestamp=2, msg="Test2"), Row(file_name = "test-guangzhou3", topic='camera3', timestamp=3, msg="Test3"), Row(file_name = "test-guangzhou1", topic='camera1', timestamp=4, msg="Test4") ]

test_df = spark.createDataFrame(test_rows)

test_df.write.format("avro")
    .mode('overwrite').save("hdfs:///tmp/bag_parser279181359_3")

loaded_df =  spark.read.format("avro").load('hdfs:///tmp/bag_parser279181359_3')

loaded_df.show()

我看到的错误信息：

Py4JJavaError: An error occurred while calling o701.collectToPython.
: java.lang.AbstractMethodError
    at org.apache.spark.sql.execution.FileSourceScanExec.inputRDD$lzycompute(DataSourceScanExec.scala:337)
    at org.apache.spark.sql.execution.FileSourceScanExec.inputRDD(DataSourceScanExec.scala:331)
    at org.apache.spark.sql.execution.FileSourceScanExec.inputRDDs(DataSourceScanExec.scala:357)
    at org.apache.spark.sql.execution.WholeStageCodegenExec.doExecute(WholeStageCodegenExec.scala:627)
    at org.apache.spark.sql.execution.SparkPlan$$anonfun$execute.apply(SparkPlan.scala:137)
    at org.apache.spark.sql.execution.SparkPlan$$anonfun$execute.apply(SparkPlan.scala:133)
    at org.apache.spark.sql.execution.SparkPlan$$anonfun$executeQuery.apply(SparkPlan.scala:161)
    at org.apache.spark.rdd.RDDOperationScope$.withScope(RDDOperationScope.scala:151)
    at org.apache.spark.sql.execution.SparkPlan.executeQuery(SparkPlan.scala:158)
    at org.apache.spark.sql.execution.SparkPlan.execute(SparkPlan.scala:133)
    at org.apache.spark.sql.execution.SparkPlan.getByteArrayRdd(SparkPlan.scala:289)
    at org.apache.spark.sql.execution.SparkPlan.executeTake(SparkPlan.scala:381)
    at org.apache.spark.sql.execution.CollectLimitExec.executeCollect(limit.scala:38)
    at org.apache.spark.sql.Dataset$$anonfun$collectToPython.apply(Dataset.scala:3259)
    at org.apache.spark.sql.Dataset$$anonfun$collectToPython.apply(Dataset.scala:3256)
    at org.apache.spark.sql.Dataset$$anonfun.apply(Dataset.scala:3373)
    at org.apache.spark.sql.execution.SQLExecution$$anonfun$withNewExecutionId.apply(SQLExecution.scala:79)
    at org.apache.spark.sql.execution.SQLExecution$.withSQLConfPropagated(SQLExecution.scala:144)
    at org.apache.spark.sql.execution.SQLExecution$.withNewExecutionId(SQLExecution.scala:74)
    at org.apache.spark.sql.Dataset.withAction(Dataset.scala:3367)
    at org.apache.spark.sql.Dataset.collectToPython(Dataset.scala:3256)
    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:357)
    at py4j.Gateway.invoke(Gateway.java:282)
    at py4j.commands.AbstractCommand.invokeMethod(AbstractCommand.java:132)
    at py4j.commands.CallCommand.execute(CallCommand.java:79)
    at py4j.GatewayConnection.run(GatewayConnection.java:238)
    at java.lang.Thread.run(Thread.java:748)

(<class 'py4j.protocol.Py4JJavaError'>, Py4JJavaError(u'An error occurred while calling o701.collectToPython.\n', JavaObject id=o702), <traceback object at 0x7fc031b5c878>)

Answer 1

AbstractMethodError :

当应用程序试图调用抽象方法时抛出。通常，此错误会被 compiler;如果某些 class 的定义自上次编译当前执行的方法以来发生了不兼容的更改，则此错误只会在运行时间发生。

AFAIK 你必须调查你使用的编译版本和运行.

Answer 2

有一个类似但不同的问题被问到 here that pertains to using spark-avro on emr-5.28.0. It's not the same cause as what's being discussed here in this question (since this question was asked long before emr-5.28.0 was available), but it's similar enough that I figured I'd link to my answer 以防万一有人因为看起来相似的堆栈跟踪和相似的问题而偶然发现这个问题。

Spark 2.4.1 无法从 HDFS 读取 Avro 文件

Spark 2.4.1 can not read Avro file from HDFS

bigdata

hdfs

avro

apache-spark

avro-tools