AWS Cost and Usage Report Specialized2015年〜
A reporting service that delivers detailed AWS cost and usage data to S3
What It Does
AWS Cost and Usage Report (CUR) delivers the most detailed data about your AWS spending and usage to an S3 bucket in CSV or Parquet format. It includes resource-level billing details, usage hours, data transfer volumes, discount application status, and every other piece of billing information. It provides finer-grained data than Cost Explorer, enabling you to build your own cost analysis platform.
Use Cases
It is used in large-scale AWS environments for accurate resource-level cost allocation, and for building custom cost dashboards with Athena or QuickSight. FinOps teams also use it to auto-generate monthly cost reports and to analyze Savings Plans and Reserved Instance coverage in detail.
Everyday Analogy
Think of it like a credit card transaction statement. If Cost Explorer is the monthly bill (total amount and category breakdown), CUR is the individual transaction detail (date, merchant name, amount, points applied). With the raw transaction data, you can run your own custom aggregations and analyses.
What Is Cost and Usage Report?
AWS Cost and Usage Report (CUR) provides the most detailed raw billing data available from AWS. While Cost Explorer is a graphical analysis tool, CUR delivers raw line-item data to S3. By analyzing this data with Athena, Redshift, QuickSight, and other tools, you can perform advanced cost analysis that goes beyond what Cost Explorer offers.
Report Content and Granularity
CUR contains hundreds of columns including resource ID, usage type, operation, charges, discount amounts, and tag information. You can choose hourly, daily, or monthly granularity. With hourly granularity, you can track exactly when a specific EC2 instance was running and how much it cost. Including resource IDs lets you pinpoint costs for individual resources.
Integration with Athena
Analyzing CUR data with Athena is the most common pattern. When you enable 'Athena integration' in the CUR settings, a table is automatically created in the Glue Data Catalog, allowing you to analyze cost data with SQL queries. For example, you can run queries like 'Aggregate last month's EC2 costs by instance type' or 'Total cost of resources without tags.' Choosing Parquet format reduces query scan volume and lowers Athena costs. For a deeper understanding of Athena integration, technical books on Amazon are a handy resource.
Getting Started
To get started with CUR, create a new report from the 'Cost and Usage Reports' section of the Billing console. Specify the report name, time granularity, data format (CSV or Parquet), and the destination S3 bucket. Reports are updated several times a day and delivered to your S3 bucket. If you enable Athena integration, you can immediately start analyzing the delivered data with SQL.
Things to Watch Out For
- CUR itself is free, but you will incur S3 storage costs and Athena query execution costs. Choosing Parquet format reduces data size and helps keep costs down
- Including resource IDs significantly increases the report data volume. If detailed analysis is not needed, exclude resource IDs to keep data volume manageable
- CUR data has up to a 24-hour delay. For real-time cost monitoring, use Cost Explorer or Budgets alongside it
