Support alerting for all services deployed on dedicated/local Bluemix platforms

As an administrator of a client platform, I experienced several situations where applications crashed because of exhausted resources on various services of the platform. It happened with network bandwidth saturation (vyatta) and with Compose RAM (over consumption upto 122%) that hung services and even control over RabbitMQ service instances. All services deployed on Bluemix dedicated SHOULD implement a minimal alerting compliant with the Bluemix notification mechanism. Today only a few situations are notified by Bluemix. Examples of enhancements for services that should generate alerts:

- Compose: Thresholds on storage, RAM, CPU, I/O usage, outage

- Cloudant: Thresholds on storage, RAM, CPU, I/O usage, outage

- Network: Internal and External bandwidth (Vyatta, DirectLink, Fortigate), outage

- Bluemix CF: RAM, CPU over consumption by specific application (cell # should be included in alert), outage

- Docker: Thresholds on storage, RAM, CPU, I/O usage, outage

- GitHub Enterprise: Thresholds on storage, I/O usage, outage

- Continuous Delivery: Thresholds on storage, CPU usage, outage