AWS AppConfig Feature Flags are designed to release features safely with minimal code changes. The lowest operational overhead approach is to use a single feature flag and configure it to deliver the feature to a percentage of users through built-in targeting rules and variants.
With option C, the developer creates one AppConfig feature flag and defines a variant that represents the new feature behavior (for example, enabled=true or a “newExperience” variant). Then the developer creates a rule in AppConfig to target 15% of users. AppConfig feature flags support rules that can evaluate attributes (such as user IDs or other contextual attributes supplied by the application) and apply percentage-based rollout. This avoids building and maintaining custom rollout logic in the application.
Option A and D both require implementing and maintaining custom traffic splitting in code, which increases complexity and operational burden, and risks inconsistent behavior across clients/services.
Option B adds extra configuration overhead and complexity by managing multiple feature flags for “groups” instead of using a single flag with a variant and rule. A single flag with variants is the clean, scalable pattern.
Therefore, create one AppConfig feature flag, define a variant, and configure a rule to target 15% of users.
Contribute your Thoughts:
Chosen Answer:
This is a voting comment (?). You can switch to a simple comment. It is better to Upvote an existing comment if you don't have anything to add.
Submit