Not logged in. Login

Solving spark-submit issue for Windows machines

Step 1: Download Spark (You can download the package pre-built for Hadoop 2.6 and later)

Step 2: Download winutils

Step 3: Copy the winutils.exe file inside the "bin" folder of the unzipped Spark directory

Step 4: In your code add the following lines

from pyspark import SparkContext, SparkConf
SparkContext.setSystemProperty("hadoop.home.dir", "S:\\spark-1.5.2-bin-hadoop2.6\\")

Note: Path in the setSystemProperty should be where you have the "bin" folder containing winutils.exe

Updated Thu Jan. 07 2016, 19:10 by sjishan.