[nsp] ip verify unicast reverse-path confirmation?

[Nicolas Sayer]

Yep, CEF was activated on the interface... It's a GigE interface, the 
troubling part of debugging that issue was that under 500Mb/s traffic 
was going through the interface flawlessly, over 500Mb/s we started 
losing packets (yep losing). CPU was fairly well loaded under 500Mb/s, 
99% over that limit (to note that no actual process was using the CPU 
that much, the high load-average was probably due to a high interupt 
rate). I actually thought someone was rate-limiting at 500Mb/s on the 
line...
Of course, when i take uRPF, all suddenly goes well (traffic flows over 
500 and CPU's back at normal).

Couldn't find any information on CCO about similar problems... Am i the 
only one to go over 500Mb/s on a 6500 ?

thanks for your help, from the information i got in this thread, it 
seems that it's an MSFC1 issue... crap...


On jeudi, déc 4, 2003, at 19:59 Europe/Paris, Streiner, Justin wrote:
> Did you have CEF enabled and specifically enabled on the interface?  
> Also,
> was the CPU OK before enabling unicast RPF, e.g. are you sure that just
> enabling unicast RPF on the interface is what killed your CPU?

-- 
INRIA/DirDRI/VTHD - Tél:+33 13963 5795 - Fax:+33 66344 7365 - Mob:+33 
663440227
Domaine de Voluceau - B.P. 105 Rocquencourt - 78153 - Le Chesnay CEDEX 
- FRANCE
fingerprint: DC9C 4D03 E26F A303 2443 A185 B1B4 32E2 FD8E 1178    UIN: 
85201551