netdev - Re: limited network bandwidth with 3.2.x kernels

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <1329421492.2645.23.camel@edumazet-laptop>
Date:	Thu, 16 Feb 2012 20:44:52 +0100
From:	Eric Dumazet <eric.dumazet@...il.com>
To:	Neal Cardwell <ncardwell@...gle.com>
Cc:	David Miller <davem@...emloft.net>, bhutchings@...arflare.com,
	netdev@...r.kernel.org, alekcejk@...glemail.com
Subject: Re: limited network bandwidth with 3.2.x kernels

Le jeudi 16 février 2012 à 19:22 +0100, Eric Dumazet a écrit :

> On sender side, there are also issues, because if I have TSO on or off
> on my 'netem delay 50ms' machine, performance is completely different.


Well, my sender is a 32bit kernel, so tcp_wmem[2] was less than
1Mbytes..

$ cat /proc/sys/net/ipv4/tcp_wmem
4096	16384	903780

Socket cwnd is then capped to ~400 with TSO=off, and ~1500 with TSO=on


TSO off, a 1500 bytes frame now has truesize=2048+192, while in previous
kernel it had truesize=1500+192 before linux 3.2

(903780/2240 = 403)

(TSO permits a better truesize/len ratio, close to 1)

With 50ms RTT, 400 packets in flight means no more than 90Mbit/s


--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html