Question

A developer configures an AWS Lambda function to process messages from a standard Amazon SQS queue using an event source mapping. The Lambda function normally completes in about 45 seconds, but during peak traffic some invocations take up to 90 seconds. Customers report that the same order is occasionally processed more than once. What is the MOST likely cause and the correct fix?

Accepted Answer

The SQS queue visibility timeout is shorter than the Lambda function's maximum execution time, so messages become visible again and are redelivered; increase the visibility timeout to at least the function timeout.

Answer

The Lambda reserved concurrency is too low, causing throttling; increase reserved concurrency so all messages are processed once.

Answer

The SQS queue is FIFO but the message group ID is not set; switch to a standard queue to guarantee exactly-once delivery.

Answer

The Lambda function memory is too low, causing timeouts; increase memory so the function always completes before the batch window closes.