[c-nsp] Would millions of TxPause mean my 6500 is too slow?

[Adrian Chadd]

On Mon, Mar 17, 2008, Deny IP Any Any wrote:
> I have a 6506 with a Sup2 running in Hybrid (7.6/12.1) mode. It has a
> X6548-GE-TX, with many high-bandwidth devices on it. I am not seeing
> any interface errors, and nothing but zero's in a 'show asicreg port
> pinnacle err', however, I am getting millions (per day) of flowcontrol
> TxPause frames on some of my ports.  Does this mean I am overflowing
> the (limited) buffers on the 6548 card, or perhaps overworking the sup
> itself?

>From the Cisco docs - http://www.cisco.com/en/US/products/hw/switches/ps700/products_tech_note09186a00801751d7.shtml#ASIC

(quote)
On these modules there is a single 1-Gigabit Ethernet uplink from the port
ASIC that supports eight ports. These cards share a 1 Mb buffer between a
group of ports (1-8, 9-16, 17-24, 25-32, 33-40, and 41-48) since each block
of eight ports is 8:1 oversubscribed. The aggregate throughput of each block
of eight ports cannot exceed 1 Gbps.
(/quote)

I've had these issues with oversubscribed gige cards (in the Cisco-4500
series switches) which was (partially) solved by distributing the "busy"  
devices correctly to minimise oversubscription. You may have some success
doing this.

HTH,



Adrian