implement wrappers for agents

[bronhaim]

implement wrappers to run agents as scripts by executor job. the following are list of desired agents:

• Cluster fence operation - E.g: 1) cordon node 2) cleaning resources - deleting pods from apiserver. https://github.com/ClusterLabs/fence-agents/tree/master/fence/agents/aws - Cloud provider agent for rebooting ec2 machines.
• https://github.com/ClusterLabs/fence-agents - Scripts for executing pm devices.
• https://github.com/ClusterLabs/fence-agents/blob/master/fence/agents/compute/fence_compute.py -
• Fence agent for the automatic resurrection of OpenStack compute instances.
• https://github.com/ClusterLabs/fence-agents/tree/master/fence/agents/vmware_soap - Cloud provider agent for rebooting vmware machines.
• https://github.com/ClusterLabs/fence-agents/tree/master/fence/agents/rhevm - Cloud provider agent for rebooting oVirt hosts.

[bronhaim]

Current implementation uses hard-coded implementations in fence-scripts folder, in agent.go we define map entry for each agent script. This map should be loaded dynamically from specific path /usr/libexec/fence-agents/. all fence scripts should be copied to this dir for executor pod\job. when executor starts it loads all scripts and create agents map.