To maximize our confidence in our cloud-native services, we need to be alerted about issues ahead of the end users so that we can respond quickly and minimize the mean time to recovery. This also means that we need to eliminate alert fatigue and only alert on what really matters, otherwise important alerts will be lost in the noise. This recipe demonstrates how to create alerts on key metrics.
- Log in to your Datadog account.
- Create an IAM alert with the following settings:
- Select Monitors | New Monitor | Event
- Match events from Amazon Cloudtrail over aws-service:iam
- Multi Alert—aws_account
- Alert threshold—above 0
- Include trigger tags—true