Refactor service status checkers to make it pluggable

During the discussion on #7267, @kishoreg suggests making service status checkers pluggable so that different customers can plugin their own implementation of status checkers. Apparently there have been some needs of customizing health status check in Kubernetes as well.

Start by reading the discussion in #7267 and locating the existing service status checker implementation in the Java codebase. Define the extension points needed for customer-provided and Kubernetes-specific health checks; done means different checker implementations can be plugged in without changing the core status-checking flow.

Written by the indexing model from the issue text.