InfoQ

AWS has introduced a new capability for triggering Lambda functions directly from Amazon RDS for SQL Server database events, paving the way for event-driven architectures on SQL Server databases hosted on RDS. This feature enables database events to automatically invoke Lambda functions, allowing for a seamless integration of cloud-based serverless computing with traditional SQL Server databases.
The implementation relies on an event-driven pattern where stored procedures write formatted messages to SQL Server’s ERRORLOG using the RAISERROR command. These messages are then detected by CloudWatch Logs subscription filters, which invoke a callback Lambda function. The callback Lambda publishes messages to Amazon SNS, which routes them to SQS queues to trigger designated Lambda functions. By leveraging this pattern, business logic is decoupled from the database, avoiding the pitfalls of tight coupling and offering scalability without the bottlenecks associated with traditional in-database triggers.
Security is a key consideration, with the guidance emphasizing the need to encrypt sensitive data before it reaches the ERRORLOG and CloudWatch. In addition, AWS provides a Cloud Development Kit (CDK) template available on GitHub for easy deployment of this architecture.
While this AWS solution contrasts with Azure’s native SQL trigger bindings—which utilize SQL Server change tracking instead of log parsing—practitioners have observed up to 40% cost reductions due to selective triggering and batch processing, showcasing the potential efficiency gains from this new capability.

Collections

AWS now supports triggering Lambda functions from Amazon RDS for SQL Server database events, enabling event-driven architectures on SQL Server. The pattern works by having stored procedures write formatted messages to SQL Server's ERRORLOG via RAISERROR, which CloudWatch Logs subscription filters detect to invoke a callback Lambda. That Lambda publishes to SNS, which routes messages through SQS to trigger target Lambda functions. This decouples business logic from the database, improving scalability. Security guidance recommends encrypting sensitive data before it reaches ERRORLOG and CloudWatch. A CDK deployment template is available on GitHub. Compared to Azure's native SQL trigger bindings using change tracking, this log-parsing approach has shown up to 40% cost reductions through selective triggering and batch processing.

AWS Introduces Lambda Function Triggers for RDS SQL Server Database Events