AWS Compute Savings Plan

You are correct, compute savings plans will cover the most expensive compute resource. When you say that these 3 instance types are being covered, are the resources running constantly or could it be that as things scale up/down that the savings plan is covering the most expensive resource for that hour?

Just to clarify in any hour your Savings Plans apply to your usage starting with the highest discount percentage. [1]

* Q. Are Saving Plans applied hourly
* A. Yes, they are calculated hourly to the usage that gets the largest discount from the saving plans vs on-demand cost. 

Q. Why am I not seeing my saving plan applied 100% on one resources?
A. Probably because the resources isn't active for that time period or wasn't the resources with the largest discount.

Below is the reference to the AWS documentation

[1] Understanding how Savings Plans apply to your AWS usage - https://docs.aws.amazon.com/savingsplans/latest/userguide/sp-applying.html

Hello, I ran into this same issue a few times! I put together some resources for other AWS users to understand what their Savings Plans are covering (by the hour, if that is what you want to know).

For the full rundown of queries, check out this blog post on Savings Plans. For now, here is a sample query you can run against your Cost and Usage Report to see which instance types your Savings plans are being applied to by region. Instance types in different regions can have different pricing, which is another reason you could be seeing the data you're seeing.

--Savings plan existence cost
--Savings plan existence cost by instance type, region
SELECT 
  [product/instanceType],
  [product/region],
    round(sum([savingsPlan/SavingsPlanEffectiveCost]), 4) as existence_cost
FROM CUR
WHERE
  [lineItem/ProductCode] = 'AmazonEC2'
    and [product/instanceType] <> ""
    and [lineItem/LineItemType] is 'SavingsPlanCoveredUsage'
GROUP BY 
  [product/instanceType],
  [product/region]
ORDER BY
  sum([savingsPlan/SavingsPlanEffectiveCost]);

Best of luck!