In a scenario where applications running on EC2 instances are tasked with processing messages from an Amazon SQS (Simple Queue Service) queue, there are occasional instances where messages fail to be processed, resulting in errors. To identify and isolate these problematic messages for further debugging and analysis, which option would be most effective?