This is a sample adhoc solution for Problem Defined in blog "http://sleepythread.blogspot.in/2014/10/hadoop-problem-simple-predection-system.html"

1. Data Setup:

First you need to setup the data in the Hadoop and schemas in Hive.


cd prediction-system/seed/
sudo chmod u+x prediction_system_setup.sh
./prediction_system_setup.sh

Note:
  1. Change the "hive" alias location to the actual location of hive on you env.


2. Next you need to update the job_sample.properties and add namenode & jobtracker url.

3. Build the workflow package.
 1. Building the package
    = mvn clean package
 2. Copy the tar from target/prediction-system-0.1-workflow-package.tar.gz to host having Oozie client install
 3. Untar the package
    = tar -xvf prediction-system-0.1-workflow-package.tar.gz

4. Running the workflow:

  1. ./prediction-system-0.1/bin/run_job.sh