how to deploy and execute processes

How to Deploy Process in your hadoop environment

After making entries into metadata related to your process, you have to deploy this process in your hadoop environment.

Steps to deploy your process

1. Click deploy button on process page corresponding to the process you want to deploy. ( Deploy button will show status regarding deployment of process, when you hover over the button.)
2. You have to copy process-deploy-env.properties file and process-deploy.sh file from application repositories and put properties file in same directory from where you are running .sh file. process-deploy-env.properties process-deploy.sh .You have to provide executable permissions to every shell script.
3. You have to edit process-deploy-env.properties. Set all properties based on your environment.
4. In jack repository get deploy-env.properties and update it there according to your environment.
5. Process Deploy main class calls deploy script based on type of container process. For Example for process type 1 it will call process-type-1.sh script.
6. You have to specify path of your process-type-n.sh script in md-config.xml file under your environment within in deploy and script-path tags upto file name "process-type-".( If you don't have any environment created according to your local system just create same name as in mybatis-config.xml and copy entries from any other environment and edit it according to your local environment). process-deploy-env.properties should be in same directory from where you are running these scripts. You have to provide executable permissions to every shell script.
7. You have to setup a crontab in your environment. Crontab will run process-deploy.sh script. for example */5 * * * * /home/cloudera/process-deploy.sh
8. After the deployment is complete and in UI the status for the process is deployed, you have to execute.
9. Download Workflow.py and flume.sh. Update Workflow.py and flume.sh according to your environment.
10. Update hostname varible in Workflow.py.
11. Update pathForFlumeng,pathForFlumeconf,pathForFlumeconfFile variables in flume.sh.
12. To store logs for execution, you have to create log directory as mentioned in your md-config.xml. create directory and give permission for every user to write into it.
13. Also update path in md-config.xml under execute and then oozie-script-path tags for Workflow.py. And under execute then standalone-script-path for flume.sh.