Live instance health check (hardware and software)

[bugzmanov]

(still have doubts if this is needed)
Background:

• genesis is used to deploy to production
• Resources created and managed by genesis (vms) can be influenced outside from genesis

In order to being able to react changes in instance health status (problems with servers, fatal failures of the apps), genesis should provide information (notifications?) of current instance health.

Example: if some of virtual machine became unavailable genesis should report that.

Open questions: how to determine state of running software or provide generic means of specifying check algorithm