[c-nsp] x6148 vs. x6548

[Tim Stevenson]

At 09:37 AM 6/22/2009, Kevin Graham stated:

> > You are correct. That only applies to the 6148. Originally it also
>
> > applied to the 6548 as well, but that limitation was removed later by
> > s/w optimizations in the LTL programming scheme. So you *can* get
> > more than 1G thru an etherchannel with 6548s, but of course, you
> > still can only get 1G max thru a given port group on the card.
>
>Does it hold though that etherchannel traffic gets replicated to all
>modules with bundle members?

No.

>  (ie. does a bundle spanning two 6724's
>consume bandwidth on both module's fabric connections, irrespective
>of the egress port?)

No, it is filtered on ingress to the fabric by the LTL/FPOE logic.

>  Would this also be true for ports behind
>different fabrics of a dual fabric card (ie. opposite ends of a
>6748)?

The filtering is effective on a per fabric channel basis, so only the 
half of the card w/the member port selected to actually tx the packet 
will get the packet over the fabric.

HTH,
Tim




Tim Stevenson, tstevens at cisco.com
Routing & Switching CCIE #5561
Technical Marketing Engineer, Cisco Nexus 7000
Cisco - http://www.cisco.com
IP Phone: 408-526-6759
********************************************************
The contents of this message may be *Cisco Confidential*
and are intended for the specified recipients only.