* Bump api spec version
* Do not panic on cpu monitoring error
* Centralize the socket disconnect logic and fire it also when socket errors occur. Hopefully closes#421
* Add support for ending the inbound stream. Closes#191
* Add a simple success response to API requests
* Add viewers over time API
* Move controllers to admin directory
* Add CPU and RAM usage alerting
* Create basic troubleshooting document to point alerts at
* Limit max number of hardware values collected
* Save metric value with the point in time it was taken