[c-nsp] Increased CPU utilization on 7206VXR

Stephen Kratzer kratzers at pa.net
Mon Jan 23 11:38:43 EST 2006


See anything unusual in the output of 'show proc cpu sort' or 'show logging'? 
Refer to http://www.cisco.com/warp/public/63/highcpu.html for high router CPU 
troubleshooting tips.

On Friday 20 January 2006 12:29, Roger Weeks wrote:
> We terminate our DSL customer connections on a 7206VXR:
> IOS 12.3(11)T2
> NPE-300 with 256MB RAM
> PA-A3-T3 DS3 interface
>
> We're using RBE so there is no PPPoE overhead for these connections.
>
> Since I have been keeping CPU statistics with Cacti, from April 2005,
> our average CPU utilization grew slowly from 5% up to something just
> over 10%.  Keeping watch on the 5 minute CPU graph, our usage was
> fairly flat all through the day.
>
> During this time we also grew from 150 DSL circuits to 500.
>
> In early December we added two things: 60 circuits all at once, and
> in addition to our existing BGP-published /20 and /22, a third BGP-
> published Class C.
>
> Our average CPU usage jumped from 10% to 45% immediately.  On a 5-
> minute graph, the CPU usage now peaks at 60% during the day when our
> traffic is highest, and bottoms out at 20% during the night when
> there is low traffic.
>
> Can anyone tell me why this changed so dramatically?  I had not
> counted on 60 additional circuits causing this kind of jump in CPU
> utilization.  We were planning on adding another 200+ customers to
> this router in another larger batch, but I am now thinking that the
> NPE-300 will be overutilized if we do that.
>
> --
> Roger J. Weeks
> Systems & Network Administrator
> Mendocino Community Network
> _______________________________________________
> 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/


More information about the cisco-nsp mailing list