Szymon Habrainski

If we are accepting the breaking changes then we should release this PR together with #6292 to prevent two successive breaking API changes of the same API after each other

@minwoox You are working on this? Otherwise I would be happy to take this issue

It is an [issue](https://github.com/netty/netty/issues/15153) inside Netty and it will be fixed in [4.2.2.Final](https://github.com/netty/netty/milestone/315). EDIT: [4.1.122.Final](https://mvnrepository.com/artifact/io.netty/netty-handler/4.1.122.Final) also resolves this issue.

@laurit I just ran the [reproducer](https://gist.github.com/schiemon/12d878e7c69223f7e335536f179437e1) with [4.1.122.Final](https://mvnrepository.com/artifact/io.netty/netty-handler/4.1.122.Final) and in fact the issue is resolved 👍

@ikhoon Happy to take this issue but I can't reproduce it. I `@RepeatedTest(100)` this test without any failures. Do you have any more information when this might fail? [Here](a9e3e4c97372d7f802cfd958d08a44bab48030e5) is...

There are two other duplicates of this issue by the way: https://github.com/line/armeria/issues?q=closeAfterTimeoutWhenCloseFrameIsNotSent and there is a solution proposal: https://github.com/line/armeria/issues/6156#issuecomment-2772631042

@jrhee17 Thank you for the review 🤙 . I addressed your comments. I am not sure if you like me to close the comments myself or leave them open. Feel...

FYI: This PR is not exclusively for hedging. This PR is more of a cleanup so we all can better argue about `RetryingClient` - when discussing changes for hedging but...

Hi @jrhee17, do you have any updates on this?

> I prefer that hedging is implemented first so that we can be sure that refactoring isn't done just for the sake of refactoring (unless there is a good reason...