Proxy health check

[benbjohnson]

It would be helpful for the LiteFS proxy to return a health check based on a configurable lag amount. For example, if the LiteFS replica is lagging more than 10s behind, return a non-2xx status code. The URL path can live at GET /litefs/healthz.

The endpoint could optionally return more detailed information about the health in plain text, however, this URL is public facing so we'll need to be careful about what details we expose.