2

I am trying to create an alarm in CloudWatch which triggers when any node in a given ASG has greater than 90% average CPU utilisation.

I don't want to just do MAX, because we often see short spikes to above 90%, where the AVG would still be ~80% for that time.

I'm only interested if a node in the ASG is consistently over 90% utilisation.

You can do this in Explorer web interface with an "Aggregate" but it looks like it's actually adding a metric for each instance - so I can't make an alarm for it because it references the instances by their ID (which will change over time):

CloudWatch query

0

You must log in to answer this question.

Browse other questions tagged .