# CPUUtilization metrics for an auto-scaling group shows only an aggregation

**URL:** <https://community.grafana.com/t/cpuutilization-metrics-for-an-auto-scaling-group-shows-only-an-aggregation/60248>\
**Category:** Cloudwatch\
**Tags:** telegraf, cloudwatch, aws\
**Created:** [February 4, 2022, 4:43pm UTC](https://community.grafana.com/t/cpuutilization-metrics-for-an-auto-scaling-group-shows-only-an-aggregation/60248 "2022-02-04T16:43:07Z")\
**Posts on this page:** 3\
**Page:** 1

<div class="post-metadata">

**Author:** ![ft7](https://avatars.discourse-cdn.com/v4/letter/f/5f9b8f/32.png) [@ft7](https://community.grafana.com/u/ft7)\
**Post date:** [February 4, 2022, 4:43pm UTC](https://community.grafana.com/t/cpuutilization-metrics-for-an-auto-scaling-group-shows-only-an-aggregation/60248/1 "2022-02-04T16:43:07Z")

</div>

- What Grafana version and what operating system are you using?

Latest Grafana on RHEL

- **What** are you trying to achieve?

Add a panel that shows CPUUtilization for each instance in an auto-scaling group

- **How** are you trying to achieve it?

Please see attached screenshot

- What happened?

Only one metric is shown on the graph

- What did you **expect** to happen?

Multiple metrics shown on the graph, one for each instance

- Can you copy/paste the configuration(s) that you are having problems with?

Please see attached screenshot

- Did you receive any errors in the Grafana UI or in related logs? If so, please tell us **exactly** what they were.

No

- Did you follow any online instructions? If so, what is the URL?

Tried many different things on the internet…

Please note the CloudWatch source is working fine. Also, when using metrics exported by Telegraf, we immediately see one metric pre instance, for example for disk usage.

Thank you very much in advance for any help!

Fabrice

 ![CPU_utilization](https://us1.discourse-cdn.com/grafana/original/3X/9/a/9a16d5fe1c9cd0134bc8c8d5c698a93b3a87ef9e.jpeg)

---

<div class="post-metadata">

**Author:** ![ft7](https://avatars.discourse-cdn.com/v4/letter/f/5f9b8f/32.png) [@ft7](https://community.grafana.com/u/ft7)\
**Post date:** [February 8, 2022, 3:44pm UTC](https://community.grafana.com/t/cpuutilization-metrics-for-an-auto-scaling-group-shows-only-an-aggregation/60248/2 "2022-02-08T15:44:41Z")

</div>

Hi everybody,

So after some research, I am quite sure the problem is that the `AWS/EC2` metric `CPUUtilization` has no dimension named `AutoScalingGroupName` . Consequently, we can’t filter on the `AutoScalingGroupName` for instances’ `CPUUtilization` .

Confusingly, autoscaling groups publish a metric also named `CPUUtilization` , which is an aggregate of the CPU utilization for all the instances in the group. Now this metric does have a dimension named `AutoScalingGroupName` , explained [here](https://docs.aws.amazon.com/autoscaling/ec2/userguide/viewing-monitoring-graphs.html), so when we filter based on the autoscaling group name, that’s the metric we are seeing, not the per-instance CPU utilization.

As far as I can see, the `CPUUtilization` published by AWS for each EC2 instance has only one metric: the instance id. That’s not enough for Grafana to filter on the autoscaling group name. In our organisation, we had to use a metric exported by an agent running on the instances.

---

<div class="post-metadata">

**Author:** ![martinetcheverry](https://avatars.discourse-cdn.com/v4/letter/m/f6c823/32.png) [@martinetcheverry](https://community.grafana.com/u/martinetcheverry)\
**Post date:** [November 30, 2022, 10:04pm UTC](https://community.grafana.com/t/cpuutilization-metrics-for-an-auto-scaling-group-shows-only-an-aggregation/60248/3 "2022-11-30T22:04:47Z")

</div>

> [@ft7](#):
>
> t’s not enough for Grafana to filter on the autoscaling group name. In our organisation, we had to use a metric exported by an agent running on the instances.

Hey Fabrice , any new insight in this issue? there other option than export the metric using the agent?
