Documentation request: Failover considerations, operating experience

[reachbach]

Firstly, kudos for a very useful contribution to the community. For a prospective user evaluating rearview, including the following in the documentation would be of great help:

1. Documentation around failover / restoration considerations for rearview itself.

2. Duration of use in production and documentation of any observed gotchas, numbers related to rearview's uptime, performance (with respect to number of metrics, poll interval) and scale (w.r.t number of metrics and frequency of sampling) in a production environment.

3. Is rearview still in active use / development internally ? Any other adopters outside of LivingSocial?

Thanks.