[c-nsp] ip tcp adjust-mss on LNS 7206VXR

[Rodney Dunn]

Yep to move the same amount of data you are correct.

I'd expect that removing the fragmentation part would result
in a net gain of overall performance through the box.

Rodney

On Thu, Jan 11, 2007 at 02:22:43PM +0000, Kristofer Sigurdsson wrote:
> Hi,
> 
> 
> 2007/1/11, Piotr Nowacki <piotr.nowacki at interia.pl>:
> > On Thu, Jan 11, 2007 at 08:19:35AM -0500, Rodney Dunn wrote:
> > > It's done in the CEF path so the extra cycles to burn by the CPU
> > > would be pretty small.
> > >
> > > I've never seen it be a "CPU killer".
> 
> On the aggregation server layer, it's not that heavy, but if you lower
> this significantly, you are potentially creating smaller packages,
> hence more pps, hence more CPU in forwarding, all over the network.
> 
> Interestingly enough, I've seen a case where this first and foremost
> affected the CPE of residential users, resulting in their forwarding
> rate going down (cheap SOHO routers).
> 
> -Kristo
> _______________________________________________
> cisco-nsp mailing list  cisco-nsp at puck.nether.net
> https://puck.nether.net/mailman/listinfo/cisco-nsp
> archive at http://puck.nether.net/pipermail/cisco-nsp/