[c-nsp] policy routing on c4948

[Alley Hasan]

Mark-
The 75% is due to a process.
Do a sh proc cpu | ex 0.00 and send the complete output. 

Thanks
Alley

-----Original Message-----
From: cisco-nsp-bounces at puck.nether.net
[mailto:cisco-nsp-bounces at puck.nether.net] On Behalf Of Mark Kent
Sent: Thursday, June 22, 2006 12:28 PM
To: cisco-nsp at puck.nether.net
Subject: [c-nsp] policy routing on c4948

On a c4948-10GE(12.2(25)EWA), 
I have a simple one-line access-list:

access-list 150 permit ip x.y.0.0 0.0.255.255 any

and a simple route-map:

route-map foo permit 10
 match ip address 150
 set ip default next-hop a.b.c.d

where IP address a.b.c.d is "directly connected, via Vlan55"

And I do this:

int vlan6
 ip policy route-map foo

where vlan6 is pulling in about 44Mbs (8500pps).

This bumps the CPU from:

CPU utilization for five seconds: 10%/0%; one minute: 12%; five minutes:
11%

up to

CPU utilization for five seconds: 75%/2%; one minute: 60%; five minutes:
31%

Two questions:

a) I don't like the 75%, but since the second number is only 2%
   should I worry about this?

b) Is there some other magic I can do to get this kind of policy
   routing without bumping up the cpu so much?

I figured I should be able to get away with this simple example
without affecting the cpu much, and "sh int stats" does reveal that
the traffic is still hardware switched (as does, I'ld guess, the "2%").

Thanks,
-mark
_______________________________________________
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/