spark-shell 中的 Apache Hudi 示例为 Spark 2.3.0 引发错误

问题描述

我正在尝试使用 spark-shell 运行此示例 (https://hudi.apache.org/docs/quick-start-guide.html)。 Apache Hudi 文档说“Hudi 使用 Spark-2.x 版本” 环境细节是：

平台：HDP 2.6.5.0-292 Spark 版本：2.3.0.2.6.5.279-2 Scala 版本：2.11.8

我正在使用下面的 spark-shell 命令（注意 - spark-avro 版本不完全匹配，因为我找不到 Spark 2.3.2 的相应 spark-avro 依赖项）

spark-shell \
--packages org.apache.hudi:hudi-spark-bundle_2.11:0.6.0,org.apache.spark:spark-avro_2.11:2.4.4,org.apache.avro:avro:1.8.2 \
--conf 'spark.serializer=org.apache.spark.serializer.KryoSerializer'

当我尝试写入数据时，出现以下错误：

scala> df.write.format("hudi").
     |   options(getQuickstartWriteConfigs).
     |   option(PRECOMBINE_FIELD_OPT_KEY,"ts").
     |   option(RECORDKEY_FIELD_OPT_KEY,"uuid").
     |   option(PARTITIONPATH_FIELD_OPT_KEY,"partitionpath").
     |   option(TABLE_NAME,tableName).
     |   mode(Overwrite).
     |   save(basePath)
20/12/27 06:21:15 WARN HoodieSparksqlWriter$: hoodie table at file:/u/users/j0s0j7j/tmp/hudi_trips_cow already exists. Deleting existing data & overwriting with new data.
java.lang.NoSuchMethodError: org.apache.avro.Schema.createUnion([Lorg/apache/avro/Schema;)Lorg/apache/avro/Schema;
  at org.apache.hudi.spark.org.apache.spark.sql.avro.SchemaConverters$.toAvroType(SchemaConverters.scala:185)
  at org.apache.hudi.spark.org.apache.spark.sql.avro.SchemaConverters$$anonfun$5.apply(SchemaConverters.scala:176)
  at org.apache.hudi.spark.org.apache.spark.sql.avro.SchemaConverters$$anonfun$5.apply(SchemaConverters.scala:174)
  at scala.collection.Iterator$class.foreach(Iterator.scala:893)
  at scala.collection.AbstractIterator.foreach(Iterator.scala:1336)
  at scala.collection.IterableLike$class.foreach(IterableLike.scala:72)
  at org.apache.spark.sql.types.StructType.foreach(StructType.scala:99)
  at org.apache.hudi.spark.org.apache.spark.sql.avro.SchemaConverters$.toAvroType(SchemaConverters.scala:174)
  at org.apache.hudi.AvroConversionUtils$.convertStructTypetoAvroSchema(AvroConversionUtils.scala:77)
  at org.apache.hudi.HoodieSparksqlWriter$.write(HoodieSparksqlWriter.scala:132)
  at org.apache.hudi.DefaultSource.createRelation(DefaultSource.scala:125)
  at org.apache.spark.sql.execution.datasources.SaveIntoDataSourceCommand.run(SaveIntoDataSourceCommand.scala:46)
  at org.apache.spark.sql.execution.command.ExecutedCommandExec.sideEffectResult$lzycompute(commands.scala:70)
  at org.apache.spark.sql.execution.command.ExecutedCommandExec.sideEffectResult(commands.scala:68)
  at org.apache.spark.sql.execution.command.ExecutedCommandExec.doExecute(commands.scala:86)
  at org.apache.spark.sql.execution.SparkPlan$$anonfun$execute$1.apply(SparkPlan.scala:131)
  at org.apache.spark.sql.execution.SparkPlan$$anonfun$execute$1.apply(SparkPlan.scala:127)
  at org.apache.spark.sql.execution.SparkPlan$$anonfun$executeQuery$1.apply(SparkPlan.scala:155)
  at org.apache.spark.rdd.RDDOperationScope$.withScope(RDDOperationScope.scala:151)
  at org.apache.spark.sql.execution.SparkPlan.executeQuery(SparkPlan.scala:152)
  at org.apache.spark.sql.execution.SparkPlan.execute(SparkPlan.scala:127)
  at org.apache.spark.sql.execution.QueryExecution.toRdd$lzycompute(QueryExecution.scala:80)
  at org.apache.spark.sql.execution.QueryExecution.toRdd(QueryExecution.scala:80)
  at org.apache.spark.sql.DataFrameWriter$$anonfun$runcommand$1.apply(DataFrameWriter.scala:654)
  at org.apache.spark.sql.DataFrameWriter$$anonfun$runcommand$1.apply(DataFrameWriter.scala:654)
  at org.apache.spark.sql.execution.sqlExecution$.withNewExecutionId(sqlExecution.scala:77)
  at org.apache.spark.sql.DataFrameWriter.runcommand(DataFrameWriter.scala:654)
  at org.apache.spark.sql.DataFrameWriter.savetoV1Source(DataFrameWriter.scala:273)
  at org.apache.spark.sql.DataFrameWriter.save(DataFrameWriter.scala:267)
  at org.apache.spark.sql.DataFrameWriter.save(DataFrameWriter.scala:225)
  ... 68 elided

在我看来，正确的 avro 版本没有添加到类路径中，也没有被提取。任何人都可以请建议解决方法吗？我已经坚持了很长一段时间了。

解决方法

这个问题是因为 avro jar 是从 spark2/jars/avro-1.7.7.jar 引用的。这导致了错误。

我必须使用 --jars、spark.driver.extraClassPath 和 spark.executor.extraClassPath 选项来指定 .ivy2/jars 位置以覆盖默认的 avro jar。

Spark Shell 命令：

spark-shell \
 --packages org.apache.hudi:hudi-spark-bundle_2.11:0.6.0,org.apache.spark:spark-avro_2.11:2.4.4,org.apache.avro:avro:1.8.2 \\
 --jars $HOME/.ivy2/jars/org.apache.avro_avro-1.8.2.jar \
 --conf spark.driver.extraClassPath=$HOME/.ivy2/jars/org.apache.avro_avro-1.8.2.jar \
 --conf spark.executor.extraClassPath=$HOME/.ivy2/jars/org.apache.avro_avro-1.8.2.jar \
 --conf 'spark.serializer=org.apache.spark.serializer.KryoSerializer'

使用下面的代码片段打印 spark-shell 的类路径

import java.lang.ClassLoader
val cl = ClassLoader.getSystemClassLoader
cl.asInstanceOf[java.net.URLClassLoader].getURLs.foreach(println)

检查是否确实从 extraClassPath 选项中提取了 avro 文件

sc.getClass().getResource("/org/apache/avro/generic/GenericData.class")
res3: java.net.URL = jar:file:/users/joyan/.ivy2/jars/org.apache.avro_avro-1.8.2.jar!/org/apache/avro/generic/GenericData.class

apache-hudi apache-spark avro spark-avro spark-shell