[c-nsp] CRS-1 MSC utilization

[Oliver Boehmer (oboehmer)]

Bas,


> > If you want to look at the forwarding asic utilization, "show
> > controllers pse utilization location <loc>" is the command you're
> > looking for..
> 
> How should I interpret the output of that command?
> When I issue it on a single PLIM:
> 
> -----
> #show controllers pse utilization location 0/0/CPU0
>                 PPE Utilization
>         Node            Ingress   Egress
> ========================================
>         0/0/CPU0:          9.2       0.6
> ----
> 
> From this output I would think there is 10% of capacity ingress and
0.6%
> egress.
>
>[...]

Right, but that is forwarding capacity (i.e. pps), rather than bandwidth
(bps).. this is why it doesn't match the interface utilization wich you
also posted. Generally, you would run into bandwidth rather than PPS/PSE
shortage, my bad..

> Or what else should/could we monitor to prevent loss due to too much
> traffic on a PLIM.
 
I did not find a similar simple command to show the other relevant
ASIC's util. based on bandwidth, so I guess it is plain interface
bandwidth monitoring, possibly using some tools like (M)RTG where you
can graph a sum of multiple OIDs so you have a single figure which you
can monitor..

	oli