[SPARK-24018] Spark-without-hadoop package fails to create or read parquet files with snappy compression

• Type: Bug
• Status: Resolved
• Priority: Minor
• Resolution: Fixed
• Affects Version/s: 2.3.0
• Fix Version/s: 2.3.2
• Component/s: Deploy
• Assignee: Unassigned
• Reporter: Jean-Francis Roy
• Created: 18/Apr/18 18:45
• Updated: 10/Dec/18 16:44
• Resolved: 10/Dec/18 16:44

Description

On a brand-new installation of Spark 2.3.0 with a user-provided hadoop-2.8.3, Spark fails to read or write dataframes in parquet format with snappy compression.

This is due to an incompatibility between the snappy-java version that is required by parquet (parquet is provided in Spark jars but snappy isn't) and the version that is available from hadoop-2.8.3.

 

Steps to reproduce:

• Download and extract hadoop-2.8.3
• Download and extract spark-2.3.0-without-hadoop
• export JAVA_HOME, HADOOP_HOME, SPARK_HOME, PATH
• Following instructions from https://spark.apache.org/docs/latest/hadoop-provided.html, set SPARK_DIST_CLASSPATH=$(hadoop classpath) in spark-env.sh
• Start a spark-shell, enter the following:

 

import spark.implicits._
val df = List(1, 2, 3, 4).toDF
df.write
  .format("parquet")
  .option("compression", "snappy")
  .mode("overwrite")
  .save("test.parquet")

 

 

This fails with the following:

java.lang.UnsatisfiedLinkError: org.xerial.snappy.SnappyNative.maxCompressedLength(I)I
    at org.xerial.snappy.SnappyNative.maxCompressedLength(Native Method)
    at org.xerial.snappy.Snappy.maxCompressedLength(Snappy.java:316)
    at org.apache.parquet.hadoop.codec.SnappyCompressor.compress(SnappyCompressor.java:67)
    at org.apache.hadoop.io.compress.CompressorStream.compress(CompressorStream.java:81)
    at org.apache.hadoop.io.compress.CompressorStream.finish(CompressorStream.java:92)
    at org.apache.parquet.hadoop.CodecFactory$BytesCompressor.compress(CodecFactory.java:112)
    at org.apache.parquet.hadoop.ColumnChunkPageWriteStore$ColumnChunkPageWriter.writePage(ColumnChunkPageWriteStore.java:93)
    at org.apache.parquet.column.impl.ColumnWriterV1.writePage(ColumnWriterV1.java:150)
    at org.apache.parquet.column.impl.ColumnWriterV1.flush(ColumnWriterV1.java:238)
    at org.apache.parquet.column.impl.ColumnWriteStoreV1.flush(ColumnWriteStoreV1.java:121)
    at org.apache.parquet.hadoop.InternalParquetRecordWriter.flushRowGroupToStore(InternalParquetRecordWriter.java:167)
    at org.apache.parquet.hadoop.InternalParquetRecordWriter.close(InternalParquetRecordWriter.java:109)
    at org.apache.parquet.hadoop.ParquetRecordWriter.close(ParquetRecordWriter.java:163)
    at org.apache.spark.sql.execution.datasources.parquet.ParquetOutputWriter.close(ParquetOutputWriter.scala:42)
    at org.apache.spark.sql.execution.datasources.FileFormatWriter$SingleDirectoryWriteTask.releaseResources(FileFormatWriter.scala:405)
    at org.apache.spark.sql.execution.datasources.FileFormatWriter$SingleDirectoryWriteTask.execute(FileFormatWriter.scala:396)
    at org.apache.spark.sql.execution.datasources.FileFormatWriter$$anonfun$org$apache$spark$sql$execution$datasources$FileFormatWriter$$executeTask$3.apply(FileFormatWriter.scala:269)
    at org.apache.spark.sql.execution.datasources.FileFormatWriter$$anonfun$org$apache$spark$sql$execution$datasources$FileFormatWriter$$executeTask$3.apply(FileFormatWriter.scala:267)
    at org.apache.spark.util.Utils$.tryWithSafeFinallyAndFailureCallbacks(Utils.scala:1411)
    at org.apache.spark.sql.execution.datasources.FileFormatWriter$.org$apache$spark$sql$execution$datasources$FileFormatWriter$$executeTask(FileFormatWriter.scala:272)
    at org.apache.spark.sql.execution.datasources.FileFormatWriter$$anonfun$write$1.apply(FileFormatWriter.scala:197)
    at org.apache.spark.sql.execution.datasources.FileFormatWriter$$anonfun$write$1.apply(FileFormatWriter.scala:196)
    at org.apache.spark.scheduler.ResultTask.runTask(ResultTask.scala:87) at org.apache.spark.scheduler.Task.run(Task.scala:109)
    at org.apache.spark.executor.Executor$TaskRunner.run(Executor.scala:345)
    at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1149)
    at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:624)
    at java.lang.Thread.run(Thread.java:748)

 

  Downloading snappy-java-1.1.2.6.jar and placing it in Sparks's jar folder solves the issue.

Issue Links

• is caused by: SPARK-18646 ExecutorClassLoader for spark-shell does not honor spark.executor.userClassPathFirst