Watchdog feature for resources

[SteveL-MSFT]

Summary of the new feature / enhancement

We don't want a single resource blocking completion of a configuration. Timeouts aren't predictable as the time something takes can differ on different systems and also networks.

Proposed technical implementation details (optional)

DSC engine should have a built in watchdog that will timeout after some predefined time (can be settable via parameter, but maybe default to 5 minutes?). Basically, a resource that does not emit anything to stderr or stdout within the watchdog timeout will cause dsc to kill the resource and report a timeout. Anything sent to stderr (progress, trace, maybe we can have a generate heartbeat message if needed) will reset the watchdog.