In most use cases the predefined metrics are used to define and create a predictive scaling policy. However, you can use Custom metrics when the predefined metrics (CPU, network I/O, and Application Load Balancer request count) do not sufficiently describe your application load.
As mentioned in previous chapter you don’t have 24 hours data for predictive scaling to start forecasting. Therefore, as part of the CloudFormation stack you created, a bash script has been executed to update two CloudWatch custom metrics which can be used in creating the predictive scaling policy.
Verify scaling and load metrics data in CloudWatch.
If you’re running the workshop in your own account, there might be up to 30 minutes delay for CloudWatch to make the metric data available.