Unable to generate alarms/events for virtual machine CPU thresholds

bellch

I am unable to get Orion to register when a virtual machines CPU has exceeded a specified limit.

The virtualization tab reflects the VM has exceeded 80% CPU, but no event or alarm is generated.

I have generated an alert policy in the Alert Manager with the following information:

Trigger Condition: CPU Utilization is greater than or equal to 70

Reset Condition left as default

Trigger Actions: NetPerfMon Event Log: Node ${NodeName} CPU usage is ${CpuUtilization}

When I run a test alarm it gets displayed in Orion

Can anyone tell me what I am missing.

Thank you

Find more posts tagged with

Accepted answers

All comments

mcbridea

Looks like you are doing the right things. Can you open a ticket and we'll help you get through this?

Andy

bshopp

In the advanced alert manager UI there is a drop down to select if this alert is against a node, interface, volume etc. under trigger conditions. Under there, you should find an option specific to virtual machines.

bellch

I was using the Virtual Machine option. With further testing if I change the type to Node I get the alert for the virtual machine.

So what is the differentce between the Node and Virtual Machine type. What CPU/Memory values is it using. Is it a percentage of the total Node CPU/Memory or only the virtual machine CPU/Memory that is found in VirtualCenter under performance for that VM?