[c-nsp] High ESMP value for 4006 CPU
Amol Sapkal
amolsapkal at gmail.com
Tue Feb 13 01:23:17 EST 2007
Hi all,
Here's an interesting problem that I came across on one of my 4006's.
------------------------------
SWITCH#sh proc cpu sorted
CPU utilization for five seconds: 58%/1%; one minute: 20%; five minutes: 18%
PID Runtime(ms) Invoked uSecs 5Sec 1Min 5Min TTY Process
28 1137757060 936939968 1214* 43.03% 6.73% 5.48% 0 Cat4k Mgmt
LoPri*
27 2148680476 419997595 5115 11.11% 10.84% 10.71% 0 Cat4k Mgmt
HiPri
Cisco says, a high 'Cat4k LoPri' indicates that packets are being
process-switched (58%/1% indicates the same)
Now, on running a cpu packet statistics, I get this:
SWITCH#sh platform cpu packet statistics
--cut--
Packets Received by Packet Queue
Queue Total 5 sec avg 1 min avg 5 min avg 1 hour
avg
---------------------- --------------- --------- --------- ---------
----------
*Esmp 2783535625 118 119 118
118*
Control 122376210 7 7
7 7
Host Learning 13976079 0 0
0 0
L2 Fwd Low 96492617 7 7
7 8
L3 Rx High 100616 0 0
0 0
L3 Rx Low 2167446 0 0
0 0
ACL fwd(snooping) 314400304 23 22 22
24
--cut--
ESMP stands for Even Simple Management Protocol. Documentation says that
these are control packets sent to the CPU. Anyone here has more information
on this? My switch goes upto 95% utilization (cat4k mgmt loPri eats most of
it).
About the switch:
IOS: 12.2(18)EW4
Ports: Gig ports for connectivity to servers (ports are bundled and connect
to servers)
Spanning tree: enabled
Port-fast: not enabled
Switch is doing L2 (only configured with a management IP)
Can anyone let me know what does ESMP exactly do? I would like to reduce the
CPU utilization.
--
Warm regards,
Amol Sapkal
-------------------------------------------------------------------
"When I'm not in my right mind, my left mind
gets pretty crowded"
-------------------------------------------------------------------
More information about the cisco-nsp
mailing list