Pierre De Rop

Rebased the PR on top of latest netty5 branch. To summarize, this PR has merged the following files from latest 1.0.x branch (SHA a001ed3e6c21e666a7d54ed3fb4f6ecd42c59a8f): reactor-netty-http/src/main/java/reactor/netty/http/client/HttpClient.java reactor-netty-http/src/main/java/reactor/netty/http/client/HttpClientFinalizer.java reactor-netty-http/src/main/java/reactor/netty/http/client/HttpClientOperations.java reactor-netty-http/src/main/java/reactor/netty/http/server/HttpServerFormDecoderProvider.java reactor-netty-http/src/main/java/reactor/netty/http/server/HttpServerOperations.java reactor-netty-http/src/main/java/reactor/netty/http/server/HttpServerRequest.java...

@violetagg , in order to fix the additional compilation warnings, I had to do a fix in the netty-contrib, so I have created this PR (but I have not merged...

The netty-contrib PR is now used from the [httpcontent-used-as-raw-type](https://github.com/pderop/codec-multipart/tree/httpcontent-used-as-raw-type), because the PR is not yet merged. the CI still reports some errors, I'm now looking into them.

Rebase in order to pick up the last changes in netty5 branch (expecially PR #2449)

fixed up previous commit, where a missing import was remaining.

rebased on netty5, in order to fix a conflict in .github/workflows/check_transport.yml, which is now using all netty-contribs from their main branches.

Hi @crankydillo , I just have read the discussion from Gitter you have mentioned, I'm sorry but I'm still not sure why you can't just leverage the Reactor _Retry_ operator...

Hi @yanwenliang , Without having a reproducer scenario for the issue, it is impossible to guess just based on the provided logs. From the logs, we see that the application/framework...

@bruto1 Hi, Out of curiosity, can you confirm if performance is better when using the timeout operator with a "single" scheduler (**Schedulers.single()**) instead of using default parallel scheduler ? for...

Hi @bruto1 , sure, the sample project is [here](https://github.com/pderop/test-reactor-flux-timer-benchmark), I hope it will help to track this issue.