Go to All Forums

AWS Custom Cloudwatch metric and AWS Cloudwatch Alarm ingestion

I notice that you poll the Cloudwatch API to get metric ingestion for a number of AWS products, but are missing the ability to ingest Custom Cloudwatch metrics.

We produce custom metrics, and would really like to be able to use them in Site 24x7.

They tend to have a pretty standard 5 minute polling interval. Currently there are about 20 of these metrics we alert on using Cloudwatch Alarms.

Even having the ability to ingest Cloudwatch alarms would be a great start.

Like (3) Reply
Replies (2)

Re: AWS Custom Cloudwatch metric and AWS Cloudwatch Alarm ingestion

Hi Quentin,

Thanks for raising your requirement.

We do have plans to support custom metrics. We are currently in the development phase . In the meantime , we would like to understand about your usage , in case of custom metrics. It would be great , if you can share the following details

1. Dimensions used.
2. Statistics preferred for these metrics(for alerting).  
3. How you classify those metrics?(For example , a custom metric shall be used for a particular instance of type cx.large. How do you prefer to see that classification , in case of metric data being shown?)

With respect to CloudWatch Alarms, we will add this to our product roadmap. We will update you on this pretty soon.


Ananthkumar K S

Like (0) Reply

Re: Re: AWS Custom Cloudwatch metric and AWS Cloudwatch Alarm ingestion

Thanks for look at this Ananthkumar K S.

Our primary usage is the output from a Cloudwatch agent installed on a EC2 instance. This gives a lot of additional detail that is not present with the Native Cloudwatch metrics for EC2.
Here's AWS's documentation on what can be collected: docs.aws.amazon.com/AmazonCloudWatch/latest/monitoring/metrics-collected-by-CloudWatch-agent.html

The metric and units are all detailed in the AWS documentation.

Outside of the EC2 CW agent use case, our developers are outputting their logs to CW logs, and event output to their own custom Cloudwatch metrics. The metrics and units will be unique for each product, and are used to alert if a threshold is met. e.g. Some output a file count every 5 minutes, and alert actions are taken on too few / too many files.

Like (2) Reply

Was this post helpful?