It turns out both features (separating computation of the status and request handling as well as timelimits for computation of the status) have been requested by other users and are under consideration by spring. You can check out these issues to track the progress on these issues: https://github.com/spring-projects/spring-boot/issues/25459 https://github.com/spring-projects/spring-boot/issues/43449

In addition the tests currently run sequentially and running them parallel is also under consideration here: https://github.com/spring-projects/spring-boot/issues/2652

Since all of these issues are already open for a while, the threads mention a few ways to implement the respective behavior on the application side overwriting the existing status checks. In particular, there is also this library that combines these features and allows to use them for any health check via annotation: https://github.com/antoinemeyer/spring-boot-async-health-indicator Note that it still would need you to overwrite any existing status checks to apply the different behavior to them (and disable the default version).