Henery309

@ahmetb what if we mark the container as keep warm. If we try to send logs after request has finished will they be dropped?

@vdjuric don't think this approach will because: - HttpProxy is already reading the request stream, you won't be able to read the request again in Retry middleware. - _httpProxy.ProxyAsync sets...

I Agree with @vdjuric, should be off by default. We want to retry request that result in intermittent failure. Our downstream services return custom http code when a intermittent failure...

> How will that help avoid buffering the request body? You are right it's not possible. I just read through `CreateRequestMessageAsync` and it just appears to be just forwarding the...

Had to roll up my own middleware for my specific use case . Not sure if this is optimal solution, any feedback is appreciated :) Still unfinished have to complete...

@Tratcher thanks for you feedback really appreciated it. Based on the feedback I am trying to simplify my solution. But when retrying on second destination I am getting. `Reading the...

Thanks @Tratcher it worked. I actually did not know that you can re-try middleware by just calling _next again. I thought pipelines only run sequence in 1 direction. Not sure...

Hi @Tratcher, `PathPattern` transformation does not work correctly when using retry middleware. For the first request the path is appended to the request url as expected. However, for second retry...

@Tratcher done. https://github.com/microsoft/reverse-proxy/issues/1163