AWS Lambda supports Amazon S3 as a failed-event destination for asynchronous and stream event sources
AWS Lambda now supports Amazon S3 as a failed-event destination for asynchronous invocations, and for Amazon Kinesis and Amazon DynamoDB event source mappings (ESMs). This enables customers to route the failed batch of records and function execution results to S3 using a simple configuration, without the overhead of writing and managing additional code.
Customers building event-driven applications with asynchronous event sources or stream event sources for Lambda can configure services like Amazon Simple Queue Service (SQS) and Amazon Simple Notification Service (SNS) as failed-event destinations to store the results of failed invocations. However, in scenarios where existing failed-event destinations do not support the payload size requirements for the failed events, customers need to write custom logic to retrieve and redrive event payload data. With today’s announcement, customers can configure S3 as a failed-event destination for Lambda functions invoked via asynchronous invocations, Kinesis ESMs, and DynamoDB ESMs. This enables customers to deliver complete event payload data to the failed-event destination, and helps reduce the overhead of managing custom logic to reliably retrieve and redrive failed event data.
This feature is generally available in all AWS Commercial Regions where AWS Lambda and the configured event source or event destination are available.
To enable S3 as a failed-event destination, refer to our documentation for configuring destinations with asynchronous invocations, Kinesis ESMs, and DynamoDB ESMs. This feature incurs no additional charge to use. You pay for charges associated with Amazon S3 usage.