Ask Your Question

Revision history [back]

click to hide/show revision 1
initial version

How would a ceilometer alarm trigger when monitoring an AutoScalingGroup?

I'm trying to understand when a Ceilometer alarm that monitors a Heat AutoScalingGroup triggers.

Here's an example:

This alarm is configured to indicate when CPU utilization goes above a threshold:

  type: OS::Ceilometer::Alarm
    description: Scale-up if the average CPU > 75% for 1 minute
    meter_name: cpu_util
    statistic: avg
    period: 60
    evaluation_periods: 1
    threshold: 75
      - {get_attr: [web_server_scaleup_policy, alarm_url]}
    matching_metadata: {'metadata.user_metadata.stack': {get_param: "OS::stack_id"}}
    comparison_operator: gt

If the AutoScalingGroup includes two or more VMs at some point in time, does the "avg" statistics field refer to average CPU utilization across all VMs? Or average per VM? For example, if one VM is running on 100% CPU and the other on 10% (for over a minute) should that trigger this alarm or not?

Is there any documentation out there explaining details like this?