Adding autorepair and autoreplace

Autorepair and autoreplace can be added to your assembly to keep it healthy and running in the event of foreseen and unforeseen interruptions. Autorepair automatically tries to heal your instances that are marked as unhealthy by OneOps because they triggered some healthy violation or missed a heartbeat. Both autorepair and autoreplace are configured at platform level and after an assembly is deployed. Autorepair is attempted before autoreplace and the methods differ on a case-by-case basis. For example, for Apache web server autorepair might try to restart the service if it's not responding.

However, for a compute node, it might try to reboot it. Similarly for a disk node, there might not be an autorepair procedure at all. If autorepair does not resolve the issue with a platform, the instance is marked for autoreplace. Autoreplace is triggered when the instance remains unhealthy despite OneOps's best efforts. Once an instance is marked for autoreplace, it...