cancel
Showing results for 
Search instead for 
Did you mean: 
cancel
738
Views
0
Helpful
1
Replies

High CPU load on Cat6500 Sup 720

admin_2
Level 3
Level 3

Hi all,

I notice very often a high cpu load on both of our 6500 switches, they all have a sup 720.

The output from sh proc cpu det 16408 (ios base)

CPU utilization for five seconds: 59%; one minute: 56%; five minutes: 47%

PID/TID 5Sec 1Min 5Min Process Prio STATE CPU

16408 53.7% 45.7% 36.1% ios-base 100d

1 0.3% 0.4% 0.3% 10 Receive 22h57m

2 0.8% 1.0% 0.4% 5 Condvar 1d17h

4 1.3% 1.4% 2.1% 10 Receive 80.956

6 3.9% 3.1% 2.8% 10 Reply 50m46s

7 29.6% 23.9% 16.7% 21 Intr 39d08h

8 1.0% 1.0% 1.0% 22 Intr 2d06h

11 5.0% 3.6% 3.1% 10 Reply 28m26s

13 1.0% 1.9% 1.7% 10 Receive 35m06s

15 3.0% 3.3% 2.9% 10 Receive 13m08s

17 4.2% 3.4% 2.9% 10 Receive 82.244

18 3.3% 2.5% 2.2% 10 Receive 42m58s

Process sbin/ios-base, type IOS, PID = 16408

CPU utilization for five seconds: 17%/36%; one minute: 16%; five minutes: 15%

Task Runtime(ms) Invoked uSecs 5Sec 1Min 5Min TTY Task Name

2 372992544 947435113 393 1.83% 1.59% 1.43% 0 Service Task

4 75682716 833732606 90 0.47% 0.39% 0.34% 0 Service Task

16 19224691551794399521 1071 6.14% 6.69% 6.59% 0 ARP Input

49 20079392 37178586 540 0.15% 0.15% 0.12% 0 Per-Second Jobs

90 36906803 11312351 3262 0.31% 0.19% 0.16% 0 Compute load avg

98 24336567 252594777 96 0.23% 0.15% 0.13% 0 Service Task

131 5302205993018582975 175 2.07% 1.84% 1.82% 0 IP Input

174 27704490 10355870 2675 0.23% 0.11% 0.11% 0 IPC LC Message H

177 71464178 50188603 1423 0.29% 0.29% 0.24% 0 CEF process

179 76593510 7524416 10179 0.37% 0.27% 0.25% 0 Adj Manager

185 42703197 246284017 173 0.31% 0.29% 0.22% 0 DHCPD Receive

241 166972491 563571741 296 0.87% 0.41% 0.41% 0 Port manager per

317 48625672 168431557 288 0.23% 0.23% 0.23% 0 IGMP Input

318 34415659 174301349 197 0.63% 0.20% 0.15% 0 PIM Process

320 27446549 77322191 354 0.23% 0.18% 0.14% 0 MLSM Process

324 1751262601760848851 99 0.87% 0.84% 0.80% 0 Standby (HSRP)

I notice that the process INTR with prio 21 is always showing 30 % cpu and more.

Do you know why it is showing very often 70 % cpu (total)

best regards

Jorg

1 Reply 1

corey
Level 1
Level 1

Hard to say as I do not know what all the processes do. you could try and hook up a packet analyzer that will capture data from say VLAN1 or some VLANS that the MSFC might see traffic from. Some packet storms/floods with specific types of addresses can do things like this. Broadcasts, Multicasts, etc... In any case it might give you some clues.

Review Cisco Networking for a $25 gift card