CPU states on Cisco Nexus 4000

[k.ramalingam]

                   Hi Expert,

We have Nexus 4000 that is constantly having CPU spikes and shows 100% on CPU States - Users

SYD-ES-L06-SA01# sh system resources

Load average:   1 minute: 0.11   5 minutes: 0.11   15 minutes: 0.09

Processes   :   211 total, 1 running

CPU states  :   100.0% user,   0.0% kernel,   0.0% idle

Memory usage:   2076512K total,    871832K used,   1204680K free

                  10088K buffers,  521584K cache

SYD-ES-L06-SA01# sh ver
Cisco Nexus Operating System (NX-OS) Software
TAC support: http://www.cisco.com/tac
Copyright (c) 2002-2010, Cisco Systems, Inc. All rights reserved.
The copyrights to certain works contained herein are owned by
other third parties and are used and distributed under license.
Some parts of this software are covered under the GNU Public
License. A copy of the license is available at
http://www.gnu.org/licenses/gpl.html.

Software
  BIOS:      version 1.0.37
  loader:    version N/A
  kickstart: version 4.1(2)E1(1f)
  system:    version 4.1(2)E1(1f)
  BIOS compile time:       09/29/09
  kickstart image file is: bootflash:/n4000-bk9-kickstart.4.1.2.E1.1f.bin
  kickstart compile time:  11/3/2010 19:00:00 [11/04/2010 15:04:36]
  system image file is:    bootflash:/n4000-bk9.4.1.2.E1.1f.bin
  system compile time:     11/3/2010 19:00:00 [11/04/2010 14:49:00]

Hardware
  cisco Nexus 4000 Chassis ("20x10GE/supervisor")
  Motorola, e500v2  with 2076512 kB of memory.
  Processor Board ID JAF1502CBBT

  Device name: SYD-ES-L06-SA01
  bootflash:     609168 kB

Kernel uptime is 669 day(s), 2 hour(s), 14 minute(s), 15 second(s)

Last reset at 319430 usecs after  Sat Jul 30 07:33:49 2011

  Reason: Reset Requested by CLI command reload
  System version: 4.1(2)E1(1f)
  Service:

plugin
  Core Plugin, Ethernet Plugin

sh users

SYD-ES-L06-SA01# sh users
NAME     LINE         TIME         IDLE          PID COMMENT
admin    pts/0        May 30 13:48   .         13508 (203.8.5.74) session = ssh *

Do you guys think its bug related, couldn't find any info on this occurance. Release notes on this version did not indicate any related issues.

Thanks in advance

Rgds

Kumar

[sbhadrav@cisco.com]

Intermittent High CPU spikes observed under the show process CPU history is expected normal activity on Cisco Nexus 4000 switch. This brief High CPU utilization could be caused by platform processes and does not affect the normal functioning of the switch.
 

Using the show process cpu history command on a switch with little traffic would show the following:
switch# show processes cpu history
 

CPU% per hour (last 72 hours)
  * = maximum CPU%   # = average CPU%

 Using the show processes cpu sort command, you might see CPU utilization spikes caused by platform processes.

switch# show processes cpu sort | exc 0.0

PID    Runtime(ms)  Invoked   uSecs  1Sec    Process
-----  -----------  --------  -----  ------  -----------
4425         2488   4460919      0   23.5%  statsclient
4296         2343   7178248      0   17.6%  pfma
4404           40  71615766      0    2.0%  gatosusd
CPU util  :   13.9% user,   23.8% kernel,   62.4% idle
switch#
This activity is normal for the switch and does not affect traffic because all traffic is switched by hardware ASICs and not by the CPU. Therefore intermittent high CPU observed in Nexus 4000 is expected and does not affect the normal functioning of the switch. However, the high CPU utilization could also be related to any of the following bugs.