Hi
Few suggestions:
1. Try storage mode as "memory and disk" both. >> to verify heap memory
error
2. Try to copy and read json source file from local filesystem (i.e.
Without hdfs) >> to verify minimum working code
3. Looks like some library issue which is causing lzo telated error.
On Saturday
Version: Spark 1.5.2
*Spark built with Hive*
git clone git://github.com/apache/spark.git
./make-distribution.sh --tgz -Phadoop-2.4 -Pyarn -Dhadoop.version=2.4.0
-Phive -Phive-thriftserver
*Input:*
-sh-4.1$ hadoop fs -du -h /user/dvasthimal/poc_success_spark/data/input
2.5 G /user/dvasthimal/po