What is good in modular routers these days?

Mikael Abrahamsson swmike at swm.pp.se
Mon Jul 20 00:31:13 CDT 2009


On Mon, 20 Jul 2009, Edward B. DREGER wrote:

> With a little creativity, it can _almost_ be done for IPv4.

That's most likely a big _almost_.

> With an efficient FIB algorithm, a single core on a Xeon 5400 will
> exceed 30 million lookups per second for IPv4 -- full table and lots
> of peers.

When someone asks for "2600 class router" they probably also want 
WFQ/fairqueue/LLQ, L2TPv3, PPPoE and a heap of other things that impede 
pps quite a lot on a CPU based platform.

> Of course, that fails to accomodate RIB maintenance and FIB updates.  It
> also doesn't take into account modern SMP CPUs; the RIB-handling code is
> still under development.

If you can bring all (or most) of the IOS functionality into a modern 
Intel Xeon/i7 platform with all that memory access speed etc and you use 
all the cores efficiently, then you might be able to do a lot. I've heard 
a lot of claims before (Luleå Algorithm from Effnet for instance) but it 
never came to much because functionality/stability is everything, if I 
want a stupid pps forwarding device I might as well get myself an L3 
switch, it'll use less power and have less parts that can break.

-- 
Mikael Abrahamsson    email: swmike at swm.pp.se


More information about the NANOG mailing list