[c-nsp] C3560 as CPE, possible TCAM contention

[Peter Rathlev]

Thank you Tassos and Brian. Of course "it was decided" to just change to
SDM and reboot, so now we can't test any more. :-( The problem seems to
have disappeared though. I think I'll try and set up a test box with the
same FIB-entries and SDM template, just to see if I can reproduce the
behaviour.

On Tue, 2008-04-29 at 18:09 +0200, Brian Turnbow wrote:
> Note that the tcam utilization is based on the assumtion of up to 8
> routed interfaces
> If you have more you will not be able to reach the max values.
> 
> We have some with similar values on routing templates that work fine,
> this particular unit has 13 routed interfaces.
> 
> 
>  Unicast mac addresses:                        400/3200         29/163
>  IPv4 IGMP groups + multicast routes:          144/1152          6/26
>  IPv4 unicast directly-connected routes:       400/3200         29/163
>  IPv4 unicast indirectly-connected routes:    1040/8320        246/1873
>  IPv4 policy based routing aces:               512/512           2/2
>  IPv4 qos aces:                                528/528          82/82
>  IPv4 security aces:                          1024/1024        103/103

Ok, so I guess the box should do fine for now.

> As tassos mentioned checking  the sh controllers cpu  can tell you what
> kind of traffic is making to the cpu

The "show controllers cpu-interface" and "show platform ip unicast
counts" are probably nice to have handy, thanks for the tips both of
you. :-)

Regards,
Peter