A developer wants the option to roll back to a previous Lambda version if a new deployment causes errors, while minimizing user impact. Which deployment pattern accomplishes this with the least user disruption?

The correct answer describes a canary deployment strategy using Lambda aliases. By routing a small percentage (e.g., 10%) of traffic to the new version first, the developer can monitor for errors with minimal impact on the overall user base. If errors are detected, the alias can be quickly updated to route 100% of traffic back to the stable, previous version, effectively rolling back the deployment with minimal disruption. The first incorrect option describes a blue/green deployment, which involves an immediate cutover. While it allows for quick rollback, the initial cutover to a potentially faulty version could impact all users. The third option requires modifying the application to change the Lambda version ARN, which is not a seamless or automated rollback strategy and would cause significant disruption. The fourth option is overly complex and achieves a similar outcome to the correct answer but with unnecessary alias management.