Thanks for the tip! In my case though there were no single processes loading
the cpu:

#sh proc cpu | e 0.00
CPU utilization for five seconds: 48%/35%; one minute: 32%; five minutes:
32%
 PID Runtime(ms)   Invoked      uSecs   5Sec   1Min   5Min TTY Process 
 196    21322228 427846338         49  0.31%  0.16%  0.17%   0 IP Input

 460     5474156    285374      19182  0.55%  0.07%  0.03%   0 BGP Scanner

 466        2868      6596        434 11.59%  0.96%  0.23%   1 Virtual Exec


A nice command to check how much traffic goes to the cpu:

#show ibc
Interface information:
        Interface IBC0/0(idb 0x13764774)
        5 minute rx rate 285758000 bits/sec, 23046 packets/sec
        5 minute tx rate 594270000 bits/sec, 91869 packets/sec
        110219721075 packets input, 170099314990812 bytes
        109544766443 broadcasts received
        218978746093 packets output, 173382310588137 bytes
        283552367 broadcasts sent
        0 Bridge Packet loopback drops
        109343818646 Packets CEF Switched, 8 Packets Fast Switched
        0 Packets SLB Switched, 0 Packets CWAN Switched
        Label switched pkts dropped: 0
        0 total martian ip v4 packets dropped
        Xconnect pkts processed: 0, dropped: 0
        Total paks copied for process level 0
        Total throttle drops 612719    Input queue drops 2662
        IBC resets   = 1; last at 09:35:51.771 UTC Mon Aug 11 2008
 Dev_Instance=0x1377F774


In our case the load was caused by packet fragmentation performed by the CPU
in interrupt mode I suppose.

Cheers,
Mihai



-----Original Message-----
From: Ge Moua [mailto:moua0...@umn.edu] 
Sent: Thursday, February 19, 2009 7:05 PM
To: Mihai Todor
Cc: cisco-nsp@puck.nether.net
Subject: Re: [c-nsp] high cpu load not by processes

sh proc cpu | ex 0.00

sh proc cpu detailed [pid] | ex 0.00


Regards,
Ge Moua | Email: moua0...@umn.edu

Network Design Engineer
University of Minnesota | Networking & Telecommunications Services




Mihai Todor wrote:
> Hello,
>
> Do you have any ideea on how to see what interrups are loading a router's
> CPU?
>
> We're experiencing a high cpu load (60%) on a router, yet resources are
not
> consumed by processes. Some background info - the machine is a 7609-S
router
> with redundant SUP720-3BXL supervizors having a PE role.
>
> Thanks!
>
> Mihai
>
>
> _______________________________________________
> cisco-nsp mailing list  cisco-nsp@puck.nether.net
> https://puck.nether.net/mailman/listinfo/cisco-nsp
> archive at http://puck.nether.net/pipermail/cisco-nsp/
>   

_______________________________________________
cisco-nsp mailing list  cisco-nsp@puck.nether.net
https://puck.nether.net/mailman/listinfo/cisco-nsp
archive at http://puck.nether.net/pipermail/cisco-nsp/

Reply via email to