We have currently many async tests which are related to advancing time (like spawning TCP/UDP servers and send data periodically or cancel safety testing, etc...)
The current solution is to call tokio::sleep for actual period of times making tests take more time and not reliable as they should be.
Tokio provides the feature test-util which allows controlling time for test scenarios example link
We need to scan our tests and use this feature for all cases where applicable.
Also it would be great if those changes can avoid forcing running async tests in with multiple threads on all operating systems
We have currently many async tests which are related to advancing time (like spawning TCP/UDP servers and send data periodically or cancel safety testing, etc...)
The current solution is to call tokio::sleep for actual period of times making tests take more time and not reliable as they should be.
Tokio provides the feature
test-utilwhich allows controlling time for test scenarios example linkWe need to scan our tests and use this feature for all cases where applicable.
Also it would be great if those changes can avoid forcing running async tests in with multiple threads on all operating systems