A developer’s application uses AWS Lambda and CloudFormation. As usage increased, Lambda functions began hitting rate limits when they fetch a Systems Manager Parameter Store advanced parameter on every invocation. The parameter only changes during deployments. Because usage is unpredictable, the developer wants a cost-effective way to avoid the Parameter Store rate limiting. Which solution best meets this requirement?

Using an SSM dynamic reference in the CloudFormation template to set the parameter as an environment variable is the most cost-effective and efficient solution. CloudFormation resolves the dynamic reference during deployment, injecting the parameter value directly into the Lambda function's environment variables. This means the Lambda function reads the parameter once when it initializes, not on every invocation, eliminating repeated calls to Parameter Store and preventing rate limiting. Setting reserved concurrency doesn't address the Parameter Store call frequency. Adding retries with exponential backoff would still involve multiple calls to Parameter Store, potentially exacerbating the rate limit issue during high concurrency and adding latency. Requesting a service quota increase is a reactive measure, may not be granted immediately, and doesn't solve the underlying architectural inefficiency of fetching a static parameter on every invocation.