Extract health check in its own middleware
This way we short-circuit health check requests and avoid doing domain lookup for them. We also do not report them in exported Prometheus metrics and this way avoid trigger alerts during deploys.
Related to:
Edited by Krasimir Angelov