Support HA for Agent failure events

[thedude42]

This issue is a request for feature behavior based on feedback from the field concerning the LBaaSv1 F5 Agent.

• LBaaSv1 Agent behavior designates an LBaaS pool resource to be owned by a single LBaaSv1 agent.
• In a multi-Agent environment, this behavior results in an environment where although we don't have a single point of failure for the entire LBaaS environment, each LBaaS pool configured is under a single point of failure where the pool and its attached resources are tied directly to the owning-Agent's fate.

The desired behavior and thus the enhancement being requested is that when a user deploys a multi-Agent environment, Agent failure events result in a transfer of ownership of LBaaS (et al, e.g. vpnaas, fwaas) resources be moved to healthy Agents, allowing those resources to continue to be managed by OpenStack.

[dflanigan]

This is a high priority feature for next quarter (Early in 2017).