> We just discovered a very weird problem, we're not sure what to
> attribute it to. We run a port-channel between a cisco (6509E,
> WS-X6548-GE-TX) and a Huawei NE40E. Port channel consists of 2 copper
> links and runs at about 1.2G. We've noticed huge number of input queue
> drops and overruns:

Entirely expected (we've seen similar). The 6548 is overbooked 8:1, each
ASIC handling 8 ports with 1 Gig towards the backplane bus. 2/5 and 2/8
are on the same ASIC and drops are expected if you run above 1 Gbps. You
may have better luck if you spread the channels on two different ASICs.
For high bandwidth applications a card with small or no overbooking is
better (e.g. the 6724 / 6748 family).

Steinar Haug, Nethelp consulting, [EMAIL PROTECTED]
_______________________________________________
cisco-nsp mailing list  [email protected]
https://puck.nether.net/mailman/listinfo/cisco-nsp
archive at http://puck.nether.net/pipermail/cisco-nsp/

Reply via email to