Add support for Healthchecking Dynamic Backends

[camilee03]

Meant to support the changes made in this PR in the Rust SDK to support healthchecking hostcalls for dynamic backends

There is additional healthchecking-related terminology like here, which sets backend health to BackendHealth::Unknown (which the static backend sets as a config). This should be fine to leave as is.

[dgryski]

Does this prevent sending http requests to unhealthy backends too?

[camilee03]

@dgryski No, it doesn't. This is focused on adding healthchecks as an option to dynamic backends on creation re the work in ExecuteD along similar lines, not on the other side of when a user tries to query a created backend and healthcheck it.

Unless I've misinterpreted the code for Viceroy, and it's different from ExecuteD, I believe that functionality shouldn't ~need to be changed because once the dynamic backend is built, it is processed the same way as a static backend (lmk if that assumption is incorrect)?

[dgryski]

When backend health was added to viceroy, the status was not checked during an http send. So, no change then.