[c-nsp] 12008 GSR performance

[Kevin Scheunemann]

I have a basic ACL (access-list 102) applied to the inbound of the slot1
interface (ip access-group 102 in).
Here's the output of those commands:
========= Line Card (Slot 1) =========

CPU utilization for five seconds: 1%/1%; one minute: 1%; five minutes:
1%
 PID Runtime(ms)   Invoked      uSecs   5Sec   1Min   5Min TTY Process

========= Line Card (Slot 2) =========

CPU utilization for five seconds: 4%/4%; one minute: 3%; five minutes:
1%
 PID Runtime(ms)   Invoked      uSecs   5Sec   1Min   5Min TTY Process

     0 input errors, 0 CRC, 0 frame, 0 overrun, 0 ignored
     0 input errors, 0 CRC, 0 frame, 0 overrun, 0 ignored
     0 input errors, 0 CRC, 0 frame, 0 overrun, 0 ignored, 0 abort
     0 input errors, 0 CRC, 0 frame, 0 overrun, 0 ignored

Kevin,

-----Original Message-----
From: Pete Templin [mailto:petelists at templin.org] 
Sent: Wednesday, December 07, 2005 4:08 PM
To: Kevin Scheunemann; Cisco Nsp
Subject: Re: [c-nsp] 12008 GSR performance

Kevin Scheunemann wrote:
> I just setup a 12008 with 3 SFCs 1 CSC, and GRP-B, with two 
> GE-GBIC-SC-B cards I am taking on full routes with BGP.
> I also have dCEF turned on.
>>From what it looks like I can't push more than 60mbit of traffic 
>>across
> the two line cards.
> Is there any switch fabric configuration that I need to do to enable 
> more performance?

SFC/CSC must be working, or you'd be passing 0mbit.  E1 cards shut down
if you drop below 4 (SFC+CSC).

What sort of ACLs are you using?  What do you see from "exe a sh proc c
s | e 0.0.%" or "sh int | i ign"?

pt