Hi,
We are trying to write some tests to prove following statements about reactive servers
1. Thread context switching has cost associated with it. So creating lots of threads is problematic.
2. With tens of thousands connections per server, reactive servers utilise the server resources more effectively.
We have seen these situations in production on one or more projects, but there is no code with tests which we can point people to and say 'run and see yourself'.
I was curious to know if there are already some samples which demonstrate the effects of thread context switching in blocking servers and how akka http's reactive architecture scales better. (in terms of resource utilisation and stability)
Thanks,
Unmesh