High CPU Utilization CPU1

CPU: 0 (clock ticks) Last 5 sec Last 1 min Last 5 min Total - (avg/sec) (avg/sec) (avg/sec) - User 8 3 2 26.7M Niced 0 0 0 3.9M System 58 13 3 17.3M Idle 33 83 94 2.7G Irq 0 0 0 8.5K Softirq 0 0 0 2.0M Iowait 0 0 0 2.0M Stolen 0 0 0 0 Util% (last 10 sec) - - - 0 CPU: 1 (clock ticks) Last 5 sec Last 1 min Last 5 min Total - (avg/sec) (avg/sec) (avg/sec) - User 5 2 1 28.3M Niced 0 0 0 6.5M System 30 4 1 45.6M Idle 63 94 97 2.7G Irq 0 0 0 582 Softirq 0 0 0 501.0K Iowait 0 0 0 2.5M Stolen 0 0 0 0 Util% (last 10 sec) - - - 100

11 Replies

Hamish
Cirrocumulus
Jun 06, 2019
Have a look and see which process is using all that CPU...

'top' may help there...

H
Rodrigo_Albuque
MVP
Jun 06, 2019
Hi Cathy

There is no high CPU utilisation from the output you posted.

The Idle output is just how much of its time the CPU was just sitting there doing nothing.

The System output average for the past 5 minutes for CPU 1 was just 1%.

The rest is all zeroes.
cathy_123
Cirrus
Jun 06, 2019
Thanks Everyone!

Hi Hamish I did try top command see below

Swap: 1023996k total,  782240k used,  241756k free,  262580k cached

PID USER   PR NI VIRT RES SHR S %CPU %MEM  TIME+ COMMAND
  1 root   20  0 598m 188m 1140 R 96.6 4.8  8122:08 init . -> init has 96% utilization
10580 root   RT  0 2790m 71m 52m S 3.9 1.8  6555:32 tmm.0
  9 root   20  0   0  0  0 S 2.0 0.0 74:45.46 ksoftirqd/1
5796 root   20  0 44756 17m 15m S 2.0 0.5  1427:46 bigd

Hi Rodrigo,

Would like to ask isnt the CPU0 has 0% Util and CPU1 has 100% Util?

Thank you!
- Rodrigo_Albuque
  MVP
  Jun 06, 2019
  ah, yeah. I missed that.. sorry.
  
  Are you able to take a snapshot of your system and upload to ihealth? https://ihealth.f5.com
  
  You can log in to BIG-IP and click on System --> Support.
  
  You can then upload your snapshot (called qkview) to ihealth and lastly copy the ihealth link and paste it to us.
  
  I can have a quick look at your CPU usage.
  - cathy_123
    Cirrus
    Jun 06, 2019
    Thanks Rodrigo I did upload it. will you be able to see this https://ihealth.f5.com/qkview-analyzer/qv/11773442/status/overview?offset=0 thanks
Hamish
Cirrocumulus
Jun 06, 2019
96% utilisation for init isn't normal.

It's a centos (i.e. RedHat) base. (Not sure which one for 12.1.2 - But my v24 uses Centos 6.8).

However it shouldn't normally use a lot of CPU. Anything logged? (/var/adm/messages) or in the 'dmesg' output?

Have you edited inittab at all? Normally there's just a single entry in there, but some people (Myself also) like to put important daemons in there with restart enabled. Breaking those can cause init to spin.

I take it from the discrepenecy between your tmmsh cpu output and top that it's bursty?

H
- cathy_123
  Cirrus
  Jun 06, 2019
  Hi Hamish,
  
  I surely didnt change it.. I am seeing messages on dmesg but I am not really sure about it.
  - Hamish
    Cirrocumulus
    Jun 06, 2019
    So what are the messages you're seeing? Nobody can help much without the info...