Hi, is there a way to configure timeouts for requests?

for outgoing you can just configure the client. for incoming it's something I've actually never seen done before TBH..

For incoming, we wrote our own throttling logic, and we had a pentester testing the app, and we discovered the throttling breaks because of hanging requests.

can you reuse something from resilience4k maybe?

will check

You might consider a proxy in front of the jvm to deal with stuff like that. Varnish and others can help mitigate slow loris type attacks.

I think for undertow the idle timeout is already exposed.

<http://UndertowOptions.NO|UndertowOptions.NO>_REQUEST_TIMEOUT

It can also be don for netty server if we add the ReadTimeoutHandler in the server pipeline

does not seem to work for undertow