public inbox for linux-kernel@vger.kernel.org
 help / color / mirror / Atom feed
* linux/iptables + smp question
@ 2006-04-28 11:46 Juan Pablo Abuyeres
  2006-04-28 22:50 ` Harald Welte
  0 siblings, 1 reply; 2+ messages in thread
From: Juan Pablo Abuyeres @ 2006-04-28 11:46 UTC (permalink / raw)
  To: linux-kernel

Hi guys,

I've been using an old single processor / linux 2.4 iptables based 
firewall for a few years.

Now it's time to upgrade that machine, so, I am wondering, would it be 
of real benefit if I put a two-processor system for a firewall? This 
machine is going to have 4 NICs, it's going to make routing (lots of 
routes), and firewall (iptables). I don't know if these kind of tasks 
take advantage from a multiple-processor architecture. Please enlighten 
me :)

Thank you!


^ permalink raw reply	[flat|nested] 2+ messages in thread

* Re: linux/iptables + smp question
  2006-04-28 11:46 linux/iptables + smp question Juan Pablo Abuyeres
@ 2006-04-28 22:50 ` Harald Welte
  0 siblings, 0 replies; 2+ messages in thread
From: Harald Welte @ 2006-04-28 22:50 UTC (permalink / raw)
  To: Juan Pablo Abuyeres; +Cc: linux-kernel, netfilter

[-- Attachment #1: Type: text/plain, Size: 1308 bytes --]

On Fri, Apr 28, 2006 at 07:46:13AM -0400, Juan Pablo Abuyeres wrote:
> Hi guys,

Hi, please follow up to the netfilter mailinglist, since this is not a
kernel [development] question.

> I've been using an old single processor / linux 2.4 iptables based firewall for a few years.
>
> Now it's time to upgrade that machine, so, I am wondering, would it be of real benefit if I put a 
> two-processor system for a firewall? This machine is going to have 4 NICs, it's going to make 
> routing (lots of routes), and firewall (iptables). I don't know if these kind of tasks take 
> advantage from a multiple-processor architecture. Please enlighten me :)

some notes:

1) 2.6. network stack scales better on smp
2) iptables and routing both scale very good on smp systems, if you use
   multiple interfaces and distribute the interrupts over multiple cpus
3) connection tracking inherently scales less good on SMP systems

-- 
- Harald Welte <laforge@netfilter.org>                 http://netfilter.org/
============================================================================
  "Fragmentation is like classful addressing -- an interesting early
   architectural error that shows how much experimentation was going
   on while IP was being designed."                    -- Paul Vixie

[-- Attachment #2: Type: application/pgp-signature, Size: 189 bytes --]

^ permalink raw reply	[flat|nested] 2+ messages in thread

end of thread, other threads:[~2006-04-28 22:49 UTC | newest]

Thread overview: 2+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2006-04-28 11:46 linux/iptables + smp question Juan Pablo Abuyeres
2006-04-28 22:50 ` Harald Welte

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox