health check and status interface

[peterfroehlich]

We are using resec on Nomad and ran into race conditions on deployments with our bigger clusters (>5gb data).

We implemented an health check interface to monitor if all data synced successfully. Nomad can use this to control the deployment process and only progresses if the instance is in sync and ready. We are using this on our production systems and deployments work well now.

Additionally the http interface exposes the resec states and redis info output as well. The interface is disabled by default and the listen address can be configured.

I've added an example with a more complex Nomadconfig, with health checking, blue/green deployment and metrics sidecar as well.

[0xlee]

Also added a feature announcing elected master to zookeeper.

[bdossantos]

Hi, thank you for this contribution. It would be cool if it merged, especially the health check route ... :/