AWS RDS

Required AWS RDS Attributes

The following cost calculation tags are required for traces associated with queries running against AWS RDS instances.

Always Required Fields

These fields are always required regardless of the database engine:

Attribute Name to Send to Beakpoint Insights	Example Value	Allowed Values
aws.rds.instance.id	mydbinstance	Any valid RDS instance identifier
aws.rds.instance.class	db.t3.micro	Any valid RDS instance class
aws.region	us-east-1	Any valid AWS region
aws.rds.deployment.option	Single-AZ	Single-AZ, Multi-Az
aws.rds.engine	aurora-postgresql	aurora-postgresql, postgres, mysql, mariadb, oracle-se2, oracle-ee, sqlserver-ex, sqlserver-web, sqlserver-se, sqlserver-ee
aws.rds.engine_version	15.2	Any valid version for selected engine
aws.rds.storage.type	aurora	aurora, gp2, gp3, io1, io2, standard
aws.rds.license.model	postgresql-license	Any valid license model

Required for CUR Cost Finalization

Costs calculated from the tags above are provisional — estimated from list prices. When you connect your AWS Cost and Usage Report (CUR), Beakpoint Insights replaces provisional costs with your actual billed costs by matching spans to CUR line items. That matching requires one additional tag:

Attribute Name	Example Value	Allowed Values
cloud.resource_id	arn:aws:rds:us-east-1:123456789012:db:mydbinstance	The full RDS instance ARN

Must be a resource attribute

cloud.resource_id should be set as an OpenTelemetry resource attribute. Spans sent directly via OTLP will not match your CUR data if the tag only appears at the span level, and their costs will remain provisional indefinitely. (Traces imported from AWS X-Ray have span-level values promoted automatically, but do not rely on this.)

Visualization with Mermaid

The diagram below shows which attributes are required depending on the instance type.

Retrieving RDS Instance Details

You can retrieve instance details using AWS CLI:

AWS CLI

aws rds describe-db-instances --db-instance-identifier mydbinstance --region us-east-1

which will return:

{
    "DBInstances": [
        {
            "DBInstanceIdentifier": "mydbinstance",
            "DBInstanceClass": "db.t3.micro",
            "Engine": "postgres",
            "EngineVersion": "15.2",
            "MultiAZ": true,
            "StorageType": "gp2",
            "ReadReplicaSourceDBInstanceIdentifier": "source-db-id",
            "DBClusterIdentifier": "my-cluster",
            "AvailabilityZone": "us-east-1a"
        }
    ]
}