Skip to content

Latest commit

 

History

History
5 lines (4 loc) · 778 Bytes

linux-compatibility.md

File metadata and controls

5 lines (4 loc) · 778 Bytes

The current implementation of Mobius requires a minor customization to be applied on Apache Spark for use in Linux. To build the customized Apache Spark, follow the steps below:

  1. Download and unpack the binary package of Apache Spark 1.6.0.
  2. Download and unpack the source package of Apache Spark 1.6.0, apply the diff patch PythonWorkerFactory.scala.patch on core/src/main/scala/org/apache/spark/api/python/PythonWorkerFactory.scala, and build Spark following the instructions.
  3. Replace lib/spark-assembly*hadoop*.jar in the binary package with assembly/target/scala-2.10/spark-assembly*hadoop*.jar built in Step 2. Use/deploy this modified binary package for Spark.