Consider temporary instances

[gvasquez95]

Current setup considers instances that are in running state when the daily process executes. This leaves out of the aggregation temporary instances like spot instances and instances from autoscaling groups.

[gvasquez95]

Possible solution could be changing de behavior of the Get Instances Lambda to get data directly from CloudWatch CPU Utilization metrics, rather than from EC2 running instances.