Aws Rds Reservations

Lists every running RDS DB instance (Aurora cluster members AND standalone single-instance RDS, via DescribeDBInstances) and every reserved DB instance (via DescribeReservedDBInstances), per profile × region — the read-only data behind reserved-instance coverage planning. Read-only (Describe* + sts:GetCallerIdentity).

A single sweep method fans out over every configured profile and region in one execution (one model lock, all output in one pass). It emits one instance resource per provisioned DB instance, one reserved resource per reservation, and one scan_error per (profile, region) phase that fails — an expired SSO token or an SCP-denied region is surfaced as a row, never aborting the sweep. regions is required: with an empty list the sweep makes no AWS calls and writes a single no_regions scan_error (zero data rows), which the companion report treats as degraded rather than a healthy empty fleet.

Pairs with the companion report @jentz/aws-rds-reservation-coverage, which normalizes these rows into size-flexible large-equivalent units and reports the running-minus-reserved coverage gap.

Instance-level + reservations scope. For cluster-level inventory (one resource per Aurora / Multi-AZ DB cluster and per cluster member) see the sibling @jentz/aws-rds-inventory.