Failure scenario: Test 'etcd' crash

Once VOLTHA is operational (working) if etcd crashes (becomes unavailable) the r/w and r/o cores should return error messages to all in-flight and pending requests. The error message should be focused on the operator (not the troubleshooter), but should container an error number that can be used by the troubleshooter to identify the cause is a data store unavailable. i.e. not helpful to tell the operator "data store unavailable" because they just want to add a subscriber.