To prevent loss of SNS notification messages when the downstream RDS database is unavailable (for example, shut down), which changes will prevent message loss? (Choose two.)

To prevent message loss when the downstream RDS database is unavailable, an Amazon SQS queue can act as a buffer. Subscribing an SQS queue directly to the SNS topic ensures that messages are durably stored in the queue even if the Lambda function (which would typically write to RDS) is unable to process them immediately. The Lambda function can then poll messages from this SQS queue. Configuring an Amazon SQS dead-letter queue (DLQ) for the SNS topic provides a fallback mechanism. If SNS attempts to deliver a message to a subscriber (like a Lambda function or HTTP endpoint) and the delivery fails after a configured number of retries, the message is moved to the DLQ. This prevents messages from being lost due to transient or prolonged subscriber unavailability. Replacing RDS with DynamoDB or configuring an SQS queue as a destination for a Lambda function doesn't directly address message loss from SNS when the database is unavailable. Replacing SNS with EventBridge doesn't inherently solve the message loss problem; EventBridge would still need a robust target for durable storage.