[177584] in North American Network Operators' Group
Re: scaling linux-based router hardware recommendations
daemon@ATHENA.MIT.EDU (Robert Bays)
Wed Jan 28 06:02:51 2015
X-Original-To: nanog@nanog.org
From: Robert Bays <robert@gdk.org>
In-Reply-To: <54C7BD4D.1000101@shankland.org>
Date: Wed, 28 Jan 2015 03:02:38 -0800
To: nanog@nanog.org
Errors-To: nanog-bounces@nanog.org
> On Jan 27, 2015, at 8:31 AM, Jim Shankland <nanog@shankland.org> =
wrote:
>=20
> My expertise, such as it ever was, is a bit stale at this point, and =
my=20
> figures might be a little off. But I think the general principle=20
> applies: think about the minimum number of x86 instructions, and the=20=
> minimum number of main memory accesses, to inspect a packet header, do =
a=20
> routing table lookup, and enqueue the packet on an outbound interface. =
I=20
> can't see that ever getting reduced to the point where a generic =
server=20
> can handle 40-byte packets at line rate (for that matter, "line rate" =
is=20
> increasing a lot faster than "speed of generic server" these days).
Using DPDK it=E2=80=99s possible to do everything stated and achieve =
10Gbps line rate at 64byte packets on multiple interfaces =
simultaneously. Add ACLs to the test setup and you can reach =
significant portions of 10Gbps at 64byte packets and full line rate at =
128bytes.
Check out Venky Venkatesan=E2=80=99s presentation at the last DPDK =
Summit for interesting information on pps/CPU cycles and some of the =
things that can be done to optimize forwarding in a generic processor =
environment.
=
http://www.slideshare.net/jstleger/6-dpdk-summit-2014-intel-presentation-v=
enky-venkatesan
