Tom Deseyn

cc @davidfowl

``` at System.Net.Sockets.Socket.SendPacketsAsync(SocketAsyncEventArgs e, CancellationToken cancellationToken) ``` I guess this is at: https://github.com/dotnet/runtime/blob/99297533df71e0d9e38ccf4a6bb6afd236a5f665/src/libraries/System.Net.Sockets/src/System/Net/Sockets/Socket.cs#L2981-L2984 So on these platforms the new cases added in https://github.com/dotnet/runtime/pull/72000 throw because we can't determine the `Socket`...

You could add an overload to `Run` that allows to set `askUserForInput`, and set it to `false` in the template. Otherwise you'll break anyone that relies on getting this prompt.

> For what it's worth, our new hardware environment has a switch with 6x QSFP-100G ports if you were looking to donate three network cards and a few cables lol....

> I can see 9M RPS with the latest runs on Citrine ... how is that possible then? Indeed. Citrine must have more than 10G then.

I'm looking a bit at these daily results (https://tfb-status.techempower.com/). The variation across runs is huge. plaintext: Date | Netty | aspnetcore ------|-------------------|----------------- 16/04 | 5.5 Mrps (35.7ms) | 3.0 Mrps...

@msmith-techempower the math is good for the response. For the request there are about 400 bytes: ``` GET /json HTTP/1.1 Host: server User-Agent: Mozilla/5.0 (X11; Linux x86_64) Gecko/20130501 Firefox/30.0 AppleWebKit/600.00...

> Is that an actual request sent via wrk? You can send that via wrk. I copied the request from https://www.techempower.com/benchmarks/#section=code

> The requests have about 400B (TCP content only). So 2.7M rps is about 8.6Gbps. My math assumed 400B requests when creating this issue. So round15 didn't hit 10G. And...

Any plans to work on this? I'm looking forward to see that plaintext 10xtop-1 become a top-10.