RE: Low capped TCP network speeds over the Internet

[Marcin Gondek <drixter%e-utp.net@localhost>]

Hi All,

I've collected some sysctl for my purpose: https://fido.e-utp.net/2025/01/22/NETBSD-SYSCTL.html

Not sure that this help in that case, but helps me.

Thanks,


--
Marcin Gondek / Drixter
http://fido.e-utp.net/
AS56662

-----Original Message-----
From: netbsd-users-owner%NetBSD.org@localhost <netbsd-users-owner%NetBSD.org@localhost> On Behalf Of Michael van Elst
Sent: Thursday, March 5, 2026 6:55 AM
To: netbsd-users%netbsd.org@localhost
Subject: Re: Low capped TCP network speeds over the Internet

z411%omaera.org@localhost (z411) writes:

>UDP speeds are great, but NetBSD specifically has an issue with 
>TCP/HTTP. The speed seems to be capped to exactly 986 KB/s and it's 
>very unstable (congestion handling issues I assume).

>I tried tuning the following:

>net.inet.tcp.recvbuf_auto=1
>net.inet.tcp.sendbuf_auto=1
>net.bpf.jit=1
>net.bpf.maxbufsize=4194304
>net.inet.tcp.sendbuf_max=16777216
>net.inet.tcp.recvbuf_max=16777216
>net.inet.ip.ifq.maxlen=4096
>net.inet.tcp.mssdflt=1072


You need to increase kern.sbmax too, this is a global limit on the socket buffer size. The default is 256kB, so at 200ms RTT you are limited to about 256kB/200ms ~ 10Mbit/s.

If you have many connections you may also need to increase kern.somaxkva that limits how much memory can be allocated to the network in total.


>Since post-tuning it's very stable I don't think it's about congestion 
>anymore - but I'm not sure why I'm hitting this hard cap, even after 
>increasing the buffer sizes. Could it be that NetBSD is not increasing 
>the window size?

NetBSD will grow the window and has window scaling enabled by default.
However, it is very slow growing the send window after a congestion event.