Asset Library / FinOps Use Cases / Analyze cost per compute service for a subaccount
Compute spend tends to be the largest component of spend in most cloud bills. Drilling into the resource-level view of cost for a given day can provide a detailed view of the heaviest spending compute resources, including a view of which time periods the consumption was at its highest level.
SELECT Min(ChargePeriodStart), Max(ChargePeriodEnd), ServiceName, ResourceId, ResourceName, SUM(PricingQuantity), SUM(EffectiveCost) AS MonthlyEffectiveCost FROM focus_data_table WHERE SubAccountId = ? AND ServiceCategory = 'Compute' AND ChargePeriodStart >= ? and ChargePeriodEnd < ? GROUP BY ServiceName, ResourceId, ResourceName ORDER BY MonthlyEffectiveCost DESC
Average cost, after amortization of unused commitment discounts, of each Core per calendar month.
(Effective Cost + Unused Commitment Discount Cost + Compute Cost) / Total number of Cores
Data Sources:
Measures the percentage of compute effective cost (excluding Spot) covered by commitment discount for the previous month.
Compute Cost after Commitment Discount / On-Demand Compute Cost
Data Sources:
Related Capabilities
Measure of the percentage of resources running on ‘modern’ types (e.g. instance types, volume type).
In general, cloud service providers come out with modern resource types that often come with improved price/performance.
NOTE: Always compare the actual cost impact of a migration vs. level of effort required. Not all modernizations will result in cost savings. Always consider your active reservations before migrating workloads.
Number of instances on ‘legacy’ resource types / total number of instances (can be done by app, account, total)
Data Sources:
Measure of cloud compute resource utilization, e.g. Compute (EC2), Block Storage (EBS), or Object Storage (S3).
For each resource type, calculate the amount of resources utilized as a percentage of the the total capacity allocated. Resource utilization rates should be considered over time, and as compared to the expected utilization based on the workloads running. Practitioners should also consider average vs. peak utilization rates.
Compute: CPU Utilization Rate (total consumed quantity in core-hrs or cpu-hrs / Total CPUs Allocated), Memory Utilization Rate (Total Usage in GB / Total Memory Allocated)
Volumes: Throughput Utilized / Throughput Provisioned, IOPS Utilized / IOPS provisioned, Storage Utilized / Storage Provisioned (per volume)
Data Sources:
Related Capabilities
Related Assets
Workload Optimization Library
A collection of resources to help identify types of cloud cost waste by service provider, including links to additional tools.
How to Optimize Cloud Usage
Use this paper to better understand what usage optimization is and what to consider if your organization wants to maximize the value of the cloud.
Engineering
Workload Optimization
Rate Optimization 
Cloud Sustainability