system.cpu.util

[julien56]

Hi all,

On my zabbix server 1.6.6, i want to monitor and make an alert when the windows server processor exceed 80 %.

My items CI is : system.cpu.util[avg1]

My Triggers is : {hostname:system.cpu.util[avg1].last(0)}>80

There are no problems for zabbix when i validate but i haven't a warning when the processor of my windows server exceed 80 %(I do this with cpu burner).

Of course the agent on the windows server is ok. Agent 1.6.6.

My triggers is wrong ?

Thank you.

[stage]

Try this

{hostname:system.cpu.util[]last(0)}>80

It works for me.

[julien56]

humm okay but...

for me the command is

my triggers : {hostname:system.cpu.util[].last(0)}>80

have you forgot the "." before last(0)}>80 in your command ?

thank you for your help.

[stage]

.........

Yes i forgot the "."

i have :

item: system.cpu.util[]
trigger : {hostname:system.cpu.util[].last(0)}>80

above gives you a alert every time it go's over the value (what gives a lot of alerts if the use is around the 80%). I promis you, you will go nuts.

trigger: {hostname:system.cpu.util[].avg(600)}>80

this will give u more comfort, it gives you an alert every time the averige over 600 sec go's over 80%

[julien56]

ok. I repeat because i haven't a good english and i want understand correctly.

trigger: {hostname:system.cpu.util[].avg(600)}>80

This line give an alert if the charge of processor exceed 80% during 600 seconds ?

thank you

[julien56]

For me it works but i have a strange mode.
I explain :

With this line {hostname:system.cpu.util[].last(0)}>80 i have an alert but after 7 minutes (Items update interval for system.cpu.util is 10 second !)

So with your Improvement trigger: {hostname:system.cpu.util[].avg(600)}>80 I have an alert but after 600 seconds + 7 minutes !

Why i have 7 minutes i dont know but i look.

[zabbix_zen]

{hostname:system.cpu.util[].last(0)}>80 i have an alert but after 7 minutes (Items update interval for system.cpu.util is 10 second

Something's going wrong between your actions queue and your action dispatcher(sms, email server,..). It should take about a minute.

Last of(most recent) value. where in the Parameter,
num – ignored seconds
#num – Nth value

so .avg(600) computes the average between 600sec of your last collected data, whereas .avg(#10) would compute the average between your last 10 gathered values.

I think you should decrease your Item's update interval, maybe you're overloading your server with a uselessly high Item's refresh rate.
Why would you want to gather data in 10sec interval only to be alerted if a threshold of a 600sec has been crossed anyways?

[Thib]

about this .. !

this works on every of my servers, excepted one

383:20100311:124611 Requested [system.cpu.util[,,avg1]]
383:20100311:124611 Sending back [0.000000]

it always stays at 0% (min, max, && average )

[subba5678]

Hi,
Can you please post how to Create an Item for the same ,as because i created an item with Zabbix Active Agent and i am not getting any alert . I am using the below parameter , is this correct , if not please help me out.

{BLR-NOC-06:system.cpu.util[].last(0)}>10 as well as

{BLR-NOC-06:system.cpu.util[].last(0)}>80



Thanks,
Subbu