Skip to main content

Create an AWS Billing Export (CUR 2.0) for Spotto

Every AWS connection needs a billing export: Spotto reads AWS costs from a CUR 2.0 Data Export that AWS delivers to S3. Create a New Export prepares regional private storage; after access connects, Create Billing Export in AWS creates the separate global export definition. Use this page to create one by hand, connect one you already have, or request billing history.

Supported Exports​

SettingSupportedWhy
TableCUR 2.0AWS's current schema. Legacy CUR works only through Manual Setup.
File formatParquet (required)CSV Data Exports are rejected.
Time granularityDaily (preferred) or HourlySpotto analyses by day. Hourly also works, but its files are about 24 times larger and slower to import. Monthly isn't supported: it dates the whole month on the 1st.
Resource IDsOn (required)Needed for per-resource cost and savings on recommendations.
Capacity reservation dataOff, or omittedThe capacity-reservation table configuration is not supported.
ColumnsThe 8 required columnsidentity_line_item_id, line_item_usage_start_date, line_item_usage_end_date, line_item_usage_account_id, line_item_line_item_type, line_item_product_code, line_item_currency_code, line_item_unblended_cost. Keeping every column, the default, includes them.
Reservation and Savings Plans columnsRecommendedNeeded for complete amortized cost. The portal warns when they're missing.

Other settings:

SettingChooseWhy
Split cost allocation dataOnly if you run ECS or EKSAdds a container-level cost split.
Manual discount compatibilityNoOnly for private-pricing edge cases.
File versioningOverwrite existing data export fileKeeps the bucket small. Both options work.
S3 bucketPrivate, Block Public Access onThe only bucket policy needed lets the Data Exports service write. Nothing is public.

Create the export in the management account, so it covers every account in the organization.

Create The Export In The Console​

  1. Open Billing and Cost Management > Data Exports and select Create.
  2. Under Export type, choose Standard data export and enter an Export name, such as spotto-cur2.
  3. Under Data table content settings, choose CUR 2.0, tick Include resource IDs, leave capacity reservation data off and set Time granularity to Daily (or Hourly).
  4. Under Column selection, keep every column.
  5. Under Data export delivery options, choose Parquet and Overwrite existing data export file.
  6. Under Data export storage settings, choose an S3 bucket in the company geography (AU Sydney, US N. Virginia, EU Stockholm) and enter an S3 path prefix, such as cur2. For a new bucket, let AWS apply the bucket policy it offers.
  7. Select Create.

Then connect it, as below.

Connect An Existing Export​

  1. On the connect page, choose Use an Existing Export, then select Launch Stack in AWS and create the stack.
  2. When the portal connects, choose the export from the list and select Finish to check access. Spotto reads the definition and tests storage access before saving; no grant stack is launched before this check.
  3. If the current source needs more access and its encryption has been identified, the error offers Grant Access to Export with the exact bucket, prefix, export name and any resolved KMS key. For a new SpottoBillingAccess-<deployment> stack, tick the IAM box and select Create stack.
  4. If that billing-access stack already exists, use the offered update instructions: open it in CloudFormation, select Update, choose Replace current template, and use the current versioned template URL and exact parameters shown. Creating another stack with the same name fails.
  5. After the grant stack finishes, return to Spotto and retry saving. A failed check leaves the connection unsaved. See what the preflight checks.

Use every displayed parameter on an update, including an empty KmsKeyArn when the checked destination uses S3-managed encryption. Keeping the previous key would retain the wrong grant for the new source.

Grant Access to Export adds SpottoExistingBillingExportRead to the role. It reads only that export's folder, and sits beside any SpottoBillingExportRead the access stack added, so neither replaces the other. It needs:

RequirementIf not
A supported exportCreate a new export.
The bucket is in the account you're connectingUse Manual Setup and choose the account that stores the bucket.
A customer-managed KMS key, if any, allows the roleEnter the key in the stack's KmsKeyArn parameter, and allow the role kms:Decrypt and kms:DescribeKey in the key policy.

The CloudShell command handles the stack steps, including detecting the KMS key. When the destination is empty, the portal's preflight independently reads the bucket's encryption settings and resolves any SSE-KMS key through read-only lookups. If a readable file has already arrived, its actual read and key metadata establish access instead. An unreadable or unusable key blocks saving. A key you entered must match a delivered file's key, or the bucket's default key if no file exists yet. See billing preflight.

When Data Arrives​

  • AWS takes up to 24 hours to deliver the first files.
  • The first delivery has the current month only. For earlier months, request billing history.
  • After that, AWS refreshes the export at least daily, and Spotto imports new files automatically.
  • To follow it, open the account from Cloud Accounts and check Billing Data in AWS Setup Status: it shows when AWS is preparing the files, when they've arrived, and if a run failed. See billing is waiting for delivery.

You can finish connecting before the first delivery once the source passes policy readiness. The empty-folder check uses IAM simulation; it doesn't prove bucket or key policies, resource control policies or VPC endpoint policies allow reading the future files. Actual file access is tested when AWS delivers them. If nothing arrives after 24 hours, check the export's status in Billing and Cost Management > Data Exports.

Request Billing History​

Spotto uses 13 months of cost data: the current month plus the previous 12. AWS fills in past months only when you ask, so open an AWS Support case from the account that owns the export:

  1. Open AWS Support Center and select Create case.

  2. Choose Account and billing, service Cost and Usage Report, category Backfill a Report.

  3. Paste the case text. Request Billing History in the portal generates it with your account, export and dates filled in:

    Please backfill our Data Export with cost and usage data for the previous 12 months.

    Payer account ID: <account ID>
    Export name: <export name>
    Export ARN: <export ARN>
    S3 bucket and prefix: s3://<bucket>/<prefix>
    Months: <first day, such as 2025-10-01> to <last day, such as 2026-09-30>

    Please keep the export's existing settings and S3 destination.
  4. Submit the case.

Account and billing cases are available on every AWS Support plan, including Basic. AWS delivers the backfilled files to the same export, and Spotto imports them automatically. There's no need to ask for more than 12 months.

Definition Region And File Region​

AWS Data Exports creates its definition through us-east-1 even for AU/EU sources. This does not require a US bucket: the export's S3Region is the regional delivery destination. Spotto's automated flow uses a separate definition-only CloudFormation stack with no registration payload or company External ID. The access/storage stack and SNS registration remain regional. Existing sources keep their real location; review that location against your requirements before selecting one. AWS permits five CUR 2.0 exports per account; reuse a supported source when capacity is exhausted.