如何使并发与写入配置单元表的数据帧一起工作？

Question

我在 Spark 1.6 上有多个线程写入同一个 hive 表（使用 parquet 文件），当它们尝试同时写入时，在将写入文件的部分重命名为 HDFS 期间会提示错误。我正在寻找绕过这个已知 Spark 问题的解决方案。

class MyThread extends Runnable {
          def run {
          //some code
          myTable.write.format("parquet").mode("append")
                 .saveAsTable("hdfstable")
          //some code
          }
}
Executors.defaultThreadFactory().newThread(new MyThread).start()

我收到此错误：

org.apache.spark.SparkException: Job aborted.
    at org.apache.spark.sql.execution.datasources.InsertIntoHadoopFsRelation$$anonfun$run$1.apply$mcV$sp(InsertIntoHadoopFsRelation.scala:156)
    at org.apache.spark.sql.execution.datasources.InsertIntoHadoopFsRelation$$anonfun$run$1.apply(InsertIntoHadoopFsRelation.scala:108)
    at org.apache.spark.sql.execution.datasources.InsertIntoHadoopFsRelation$$anonfun$run$1.apply(InsertIntoHadoopFsRelation.scala:108)
    at org.apache.spark.sql.execution.SQLExecution$.withNewExecutionId(SQLExecution.scala:53)
    at org.apache.spark.sql.execution.datasources.InsertIntoHadoopFsRelation.run(InsertIntoHadoopFsRelation.scala:108)
    at org.apache.spark.sql.execution.ExecutedCommand.sideEffectResult$lzycompute(commands.scala:58)
    at org.apache.spark.sql.execution.ExecutedCommand.sideEffectResult(commands.scala:56)
    at org.apache.spark.sql.execution.ExecutedCommand.doExecute(commands.scala:70)
    at org.apache.spark.sql.execution.SparkPlan$$anonfun$execute$5.apply(SparkPlan.scala:132)
    at org.apache.spark.sql.execution.SparkPlan$$anonfun$execute$5.apply(SparkPlan.scala:130)
    at org.apache.spark.rdd.RDDOperationScope$.withScope(RDDOperationScope.scala:150)
    at org.apache.spark.sql.execution.SparkPlan.execute(SparkPlan.scala:130)
    at org.apache.spark.sql.execution.QueryExecution.toRdd$lzycompute(QueryExecution.scala:55)
    at org.apache.spark.sql.execution.QueryExecution.toRdd(QueryExecution.scala:55)
    at org.apache.spark.sql.DataFrameWriter.insertInto(DataFrameWriter.scala:189)
    at org.apache.spark.sql.DataFrameWriter.saveAsTable(DataFrameWriter.scala:239)
    at org.apache.spark.sql.DataFrameWriter.saveAsTable(DataFrameWriter.scala:221)
    at fr.neolink.spark.streaming.StreamingNeo$.algo(StreamingNeo.scala:837)
    at fr.neolink.spark.streaming.StreamingNeo$$anonfun$main$3$$anonfun$apply$18$MyThread$1.run(StreamingNeo.scala:374)
    at java.lang.Thread.run(Thread.java:748)

引起：

java.io.IOException: Failed to rename 
FileStatus{path=hdfs://my_hdfs_master/user/hive/warehouse/MYDB.db/hdfstable/_temporary/0/task_201812281010_1770_m_000000/part-r-00000-9a70cbea-d105-4f50-ba1b-372f555906ce.gz.parquet; 
isDirectory=false; length=4608; replication=3; blocksize=134217728; modification_time=1545988247575; 
access_time=1545988247494; owner=owner; group=hive; permission=rw-r--r--; isSymlink=false} 
to hdfs://my_hdfs_master/user/hive/warehouse/MYDB.db/hdfstable/part-r-00000-9a70cbea-d105-4f50-ba1b-372f555906ce.gz.parquet

我在 jira 上发现了这个问题：https://issues.apache.org/jira/browse/SPARK-18626

有没有办法让写作部分线程安全？一个接一个地执行？

谢谢。

Answer 1

解决方案

使用this.synchronized{}，如下所示

class MyThread extends Runnable{
      def run{
      //some code
         this.synchronized{
            myTable.write.format("parquet").mode("append")
                   .saveAsTable("hdfstable")
         }
      //some code
      }
}
Executors.defaultThreadFactory().newThread(new MyThread).start()