Hi Max,
On Mon, 08 Dec 2025 01:11:56 -0700
Max Chernoff
Hi Stefano,
On Mon, 2025-12-08 at 08:20 +0100, Stefano Brivio wrote:
With this series, I'm getting the expected (wirespeed) throughput for transfers between peers with varying non-local RTTs: I checked different guests bridged on the same machine (~500 us) and hosts with increasing distance using iperf3, as well as HTTP transfers only for some hosts I have control over (500 us and 5 ms case).
With increasing RTTs, I can finally see the throughput converging to the available bandwidth reasonably fast:
Thanks for the patch, but this unfortunately seems to make things worse in my testing (with curl/https). Using my benchmarking script from the earlier thread, with a 10MB file size and a 30s timeout:
Thanks for re-testing. I actually wanted to get back to you about your sysctl values, but, in general, I don't think things can work reliably with the values you shared for tcp_notsent_lowat. Does this (upload now taking longer/timing out with 50 ms RTT) also happen without "custom" values for tcp_notsent_lowat? I tested things quite extensively in that RTT region (without custom sysctl values) and the improvement looks rather consistent to me. -- Stefano