A team is tasked with updating an application's cloud infrastructure where a serverless function and its serving endpoints are being modified. The team intends to roll out the changes with minimal interruption to users and needs a reliable way to revert to the previous version if the new release encounters unforeseen issues. What deployment methodology should they implement to properly roll out the modifications, monitor performance, and provide a quick rollback mechanism?
Construct a duplicate setup with modifications, validate its performance, and then redirect all user traffic manually to the new setup.
Incorporate a traffic distribution strategy to evenly split requests between the new and old versions of the serverless endpoints without phased monitoring.
Release the modifications by overwriting the live production infrastructure in a single deployment action
Adopt a progressive deployment technique that begins by directing a small amount of traffic, say 10%, to the updated elements and increases this percentage gradually while monitoring for issues.