Add Uptime to telemetry
We'd like to understand the uptime of our self-managed instances, to determine if we are making improvements to:
- Recommended deployment topologies (uptake of HA, for example)
- Application reliability
We could potentially gather this data from the Prometheus instance.