RFR(s): Improving performance of Windows socket connect on the loopback adapter

[Nikola Grcevski]

Hi Alan,

I forwarded this to the two network engineers I was working with on this. I had asked before if there are any side-effects of using the SIO_TCP_INITIAL_RTO on OSs that don't support it, but I didn't realize there might be a server vs. workstation OS issue. I'll do some more testing tomorrow on my end.

Thanks again,
Nikola

-----Original Message-----
From: Alan Bateman <Alan.Bateman at oracle.com> 
Sent: July 26, 2020 6:45 AM
To: Nikola Grcevski <Nikola.Grcevski at microsoft.com>; net-dev at openjdk.java.net
Subject: Re: RFR(s): Improving performance of Windows socket connect on the loopback adapter

On 24/07/2020 16:20, Nikola Grcevski wrote:
> Thanks Alan, yes I'll need a sponsor for the patch.
>
>
I tried the patch in our CI and test/jdk/java/net/Socket/Timeouts.java
is consistently failing on Windows Server 2016 systems, specifically
testTimedConnect2 which expects a "connection refused" within 10s of attempting to connect to a port on the loopback that doesn't have any service running. SIO_TCO_INITIAL_RTO seems to be intended for desktop systems so I'm wondering if you can find out if there is any issues with using it on Windows Server editions.

-Alan