[c-nsp] Redirects / hair-pinning traffic vs. performance

[Peter Rathlev]

On Tue, 2009-06-23 at 11:11 -0400, Rodney Dunn wrote:
> On Sun, Jun 21, 2009 at 11:05:55PM +0200, Peter Rathlev wrote:
> > Interface stats for the two relevant interfaces (policy map attached to
> > Vlan2176, policy routed traffic exits via Vlan507, non policy routed
> > exist via next hop on same interface as it arrived):
> > 
> > Vlan507
> >           Switching path    Pkts In   Chars In   Pkts Out  Chars Out
> >                Processor      73750    4426830     314407   20751714
> >              Route cache          1         90          0          0
> >                    Total      73751    4426920     314407   20751714
> > Vlan2176
> >           Switching path    Pkts In   Chars In   Pkts Out  Chars Out
> >                Processor     210884   13340863     323420   21776624
> >              Route cache         23       5081         24       5267
> >                    Total     210907   13345944     323444   21781891
> 
> I thought the newew sup/rp's would show a distributed cache line that
> would show the hw forwarded counters.

Well, this is a 3560 IP Services which was chosen as a relatively
inexpensive device capable of doing PBR "in hardware". So no distributed
cache for us.

> > The "show ip traffic" seems only to show traffic received. Should it
> > also show policy routed traffic?
> 
> No. I was looking to see if there were a bunch of ICMP generated frames.

Ah, ok.

I've looked more at the load of the unit as we throw more traffic at it.
It might actually be that the peaks we see are only ARP and other
traffic destined for the processor. Since it's placed on directly on teh
intar-tubes it does receive it's fair share.

I'll monitor it closely, but it actually seems to do the job very well.
Compared to a NPE-225 doing basically the same job in another but
comparable place it looks positively relaxed. :-)

Regards,
Peter