Skip to main content

Backfill historical CUR


We support backfill of historical data with the following constraints:

  • You must be on a dedicated payer account (i.e., non-consolidated).

  • The CUR files are no older than one year.

Add CUR files to S3 bucket

The CMP automatically discovers CUR files in the Amazon S3 bucket under your dedicated payer account.

To backfill data from historical CUR files which do not exist in the bucket, you have to add them manually into the bucket:

  1. Sign in to your AWS Billing & Cost Management console.

  2. Select Cost & Usage Reports on the navigation pane, verify that cost and usage reports are enabled.

  3. Find the S3 bucket for your dedicated payer account.

  4. Use the following path structure and naming convention to add your historical CUR files.

    • base-CUR-path: The base path of the CUR file (each payer has its own base folder path).

    • yyyymmdd-yyyymmdd: The billing month notation. For example: 20210101-20210201 for the billing month of January 2021.

    • <manifest-file>-Manifest.json: This manifest file is the map of the CUR files related to that billing month. In each folder there must be such a file with its name ending in -Manifest.json.

  5. Make sure that you have granted the CMP access to the S3 bucket.


Once you uploaded the files, they will be discovered by the CMP in about six hours.

CUR files

The CUR files must follow the AWS standard:

  • The format of the billing files is CSV.

  • The billing files are compressed using zip or gz.

All the CUR files have to be in the folder specified in the %-Manifest.json file under reportKeys. Below is an example from a manifest file:

0: "ch/cloudhealth-hourly-cur/20190301-20190401/3d4115f5-cbe4-4e76-b480-c116dc632458/cloudhealth-hourly-cur-1.csv.gz"
1: "ch/cloudhealth-hourly-cur/20190301-20190401/3d4115f5-cbe4-4e76-b480-c116dc632458/cloudhealth-hourly-cur-2.csv.gz"