[
https://issues.apache.org/jira/browse/SPARK-15393?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15309952#comment-15309952
]
Jie Huang commented on SPARK-15393:
-----------------------------------
Is it possible not to write any thing (including _SUCCESS) for those empty
Dataframe? We are using dataframe to write some data to parquet file(s). Even
there is no record(row) in dataframe, it still outputs several small files
there. Actually, we should not write the entire folder if the DF is empty,
right?
> Writing empty Dataframes doesn't save any _metadata files
> ---------------------------------------------------------
>
> Key: SPARK-15393
> URL: https://issues.apache.org/jira/browse/SPARK-15393
> Project: Spark
> Issue Type: Bug
> Components: SQL
> Affects Versions: 2.0.0
> Reporter: Jurriaan Pruis
> Priority: Critical
>
> Writing empty dataframes is broken on latest master.
> It omits the metadata and sometimes throws the following exception (when
> saving as parquet):
> {code}
> 8-May-2016 22:37:14 WARNING:
> org.apache.parquet.hadoop.ParquetOutputCommitter: could not write summary
> file for file:/some/test/file
> java.lang.NullPointerException
> at
> org.apache.parquet.hadoop.ParquetFileWriter.mergeFooters(ParquetFileWriter.java:456)
> at
> org.apache.parquet.hadoop.ParquetFileWriter.writeMetadataFile(ParquetFileWriter.java:420)
> at
> org.apache.parquet.hadoop.ParquetOutputCommitter.writeMetaDataFile(ParquetOutputCommitter.java:58)
> at
> org.apache.parquet.hadoop.ParquetOutputCommitter.commitJob(ParquetOutputCommitter.java:48)
> at
> org.apache.spark.sql.execution.datasources.BaseWriterContainer.commitJob(WriterContainer.scala:220)
> at
> org.apache.spark.sql.execution.datasources.InsertIntoHadoopFsRelation$$anonfun$run$1.apply$mcV$sp(InsertIntoHadoopFsRelation.scala:144)
> at
> org.apache.spark.sql.execution.datasources.InsertIntoHadoopFsRelation$$anonfun$run$1.apply(InsertIntoHadoopFsRelation.scala:115)
> at
> org.apache.spark.sql.execution.datasources.InsertIntoHadoopFsRelation$$anonfun$run$1.apply(InsertIntoHadoopFsRelation.scala:115)
> at
> org.apache.spark.sql.execution.SQLExecution$.withNewExecutionId(SQLExecution.scala:57)
> at
> org.apache.spark.sql.execution.datasources.InsertIntoHadoopFsRelation.run(InsertIntoHadoopFsRelation.scala:115)
> at
> org.apache.spark.sql.execution.command.ExecutedCommandExec.sideEffectResult$lzycompute(commands.scala:57)
> at
> org.apache.spark.sql.execution.command.ExecutedCommandExec.sideEffectResult(commands.scala:55)
> at
> org.apache.spark.sql.execution.command.ExecutedCommandExec.doExecute(commands.scala:69)
> at
> org.apache.spark.sql.execution.SparkPlan$$anonfun$execute$1.apply(SparkPlan.scala:115)
> at
> org.apache.spark.sql.execution.SparkPlan$$anonfun$execute$1.apply(SparkPlan.scala:115)
> at
> org.apache.spark.sql.execution.SparkPlan$$anonfun$executeQuery$1.apply(SparkPlan.scala:136)
> at
> org.apache.spark.rdd.RDDOperationScope$.withScope(RDDOperationScope.scala:151)
> at
> org.apache.spark.sql.execution.SparkPlan.executeQuery(SparkPlan.scala:133)
> at org.apache.spark.sql.execution.SparkPlan.execute(SparkPlan.scala:114)
> at
> org.apache.spark.sql.execution.QueryExecution.toRdd$lzycompute(QueryExecution.scala:85)
> at
> org.apache.spark.sql.execution.QueryExecution.toRdd(QueryExecution.scala:85)
> at
> org.apache.spark.sql.execution.datasources.DataSource.write(DataSource.scala:417)
> at org.apache.spark.sql.DataFrameWriter.save(DataFrameWriter.scala:252)
> at org.apache.spark.sql.DataFrameWriter.save(DataFrameWriter.scala:234)
> at org.apache.spark.sql.DataFrameWriter.parquet(DataFrameWriter.scala:626)
> 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:237)
> at py4j.reflection.ReflectionEngine.invoke(ReflectionEngine.java:357)
> at py4j.Gateway.invoke(Gateway.java:280)
> at py4j.commands.AbstractCommand.invokeMethod(AbstractCommand.java:128)
> at py4j.commands.CallCommand.execute(CallCommand.java:79)
> at py4j.GatewayConnection.run(GatewayConnection.java:211)
> at java.lang.Thread.run(Thread.java:745)
> {code}
> It only saves an _SUCCESS file (which is also incorrect behaviour, because it
> raised an exception).
> This means that loading it again will result in the following error:
> {code}
> Unable to infer schema for ParquetFormat at /some/test/file. It must be
> specified manually;'
> {code}
> It looks like this problem was introduced in
> https://github.com/apache/spark/pull/12855 (SPARK-10216).
> After reverting those changes I could save the empty dataframe as parquet and
> load it again without Spark complaining or throwing any exceptions.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
