[c-nsp] Sup 720 question

[Rodney Dunn]

Those packets coming in would be switched by hardware through the box
and would not affect the RP processing of a telnet session.

He needs to figure out what the traffic looked like and why
the RP was affected.

Unless (and I don't claim to be a 65xx architecuture expert) the
bottleneck is the port asic the session is coming in on.
That didn't sound like it was the case based on what he said.

He needs a trace of the traffic stream.

Rodney



On Thu, Sep 28, 2006 at 03:54:30PM -0500, Dale W. Carder wrote:
> Thus spake Drew Weaver (drew.weaver at thenap.com) on Thu, Sep 28, 2006 at 12:13:35PM -0400:
> >     Does anyone know if the CPU overhead is greater when using the
> > On-board sup 720 gig-e as opposed to using a line card? We had an
> > incident where 190kpps going over Gig 5/1 (attached to the sup720)
> > choked an entire switch.
> 
> The port asic that the gig-e on the front of the 720 sits on
> also feeds the RP and the SP CPU's.  If you anger that port 
> asic, you could anger the router.
> 
> Don't use the the ports on the sup720 unless for OOB mgmt where
> you have control over the situation.
> 
> Dale
> 
> ----------------------------------
> Dale W. Carder - Network Engineer
> University of Wisconsin at Madison
> http://net.doit.wisc.edu/~dwcarder
> 
> _______________________________________________
> 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/