Using job object to execute agents

[bronhaim]

Execution will run in a job that runs shell script with the extracted parameter. The idea behind a job here is to allow executing the script from any proxy node in the cluster and to monitor the job status:

1. All possible agent scripts are installed in the job image
2. Creating agent-image with all available fence script
3. Creating the job by the controller which uses same image with different commands
4. Job objects will be posted in fencing::executeFence
5. Controller will follow jobs status before moving to next steps (in controller::handleExistingNodeFences)

the controller needs to update the nodefence object status to "running" when job is executed.