Re: Kernel Routing and the IntraNet

From: Al Boldi
Date: Tue Sep 13 2005 - 09:41:21 EST


Henrik Nordstrom wrote:
> On Mon, 12 Sep 2005, Al Boldi wrote:
> > Is there a reason why bridging in linux is not faster than routing?
>
> The main overhead seem to be elsewhere, mainly hardware related (interrupt
> handling, NIC driver, FSB bandwidth, PCI bandwidth and latency).

How could this be improved?
Or, does this imply an underlying (design) problem?

> For best bridge performance make sure your kernel is NOT built with
> support for bridge firewalling. The bridge firewalling support adds a fair
> bit of overhead to bridgeing even if not used.

Firewalling on a bridge?
Does this mean that the bridge does not operate on the physical level of the
dev, or is it ARP-FW?

Also, ifconfig eth0 up does not add the dev in /proc/../ipv4/conf, but
ifconfig eth0 0 does?!

Thanks!

--
Al

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