Provide uptime statistics for each service

[uniqueg]

Once #13 is implemented, it would be nice to include some form up uptime statistics in the GET services/{service_id}/health endpoint, e.g., percent of heartbeat calls with a return code suggesting normal operation within the last 24 hours, 1 week, 1 month, since operation. This will require some form of tracking of heartbeat results in the database.