Best practices for implementing Lambda event filtering

Familiarize yourself with the Lambda service’s filter rule syntax—specifically the filter rule syntax Lambda uses for Amazon SQS. Then, design an event filter according to your business logic. The effectiveness of any event filter depends on its ability to invoke a Lambda function only when it should. To accomplish this, you must apply the correct syntax to your filter rules so that the rules reflect the needs of your application’s business logic.

In situations where the application’s business logic demands a more complex series of filter conditions, make sure that you use multi-level filtering. For more information and example data objects, see Multi-level filtering in the AWS Lambda Developer Guide.

Limit how long you want Amazon SQS to retain unprocessed messages for by defining the MessageRetentionPeriod value in your Amazon SQS request parameters. Make sure that the value you define fits your application’s use case by identifying how long it typically takes a valid message to process. For more information, see SetQueueAttributes in the Amazon SQS API Reference. Also, see the Avoiding non-matching message accumulation section of this guide.

Configure an Amazon SQS dead-letter queue in situations where Lambda function unavailability could make it impossible to process valid messages. This extra Amazon SQS queue helps avoid creating a snowball anti-pattern in your serverless application’s architecture. For more information, see the Avoiding snowball anti-patterns section of this guide.

Make sure that you activate the event filter feature on the Lambda service. The feature must be activated when you initially create your Lambda function. For instructions, see Attaching filter criteria to an event source mapping (console) or Attaching filter criteria to an event source mapping (AWS CLI) in the AWS Lambda Developer Guide.