A company recently deployed its web application on AWS. The company is preparing for a large- scale sales event and must ensure that the web application can scale to meet the demand. The application's

Configuring higher reserved concurrency only sets a maximum limit for concurrent executions and does not address or mitigate cold start latency or the time spent establishing database connections.

Provisioned concurrency ensures a specified number of Lambda instances are pre-initialized, eliminating cold starts and reducing the latency associated with environment setup, including initial database connection establishment during bursts.

Aurora global databases and additional replicas primarily enhance read scalability and disaster recovery across regions, which does not directly solve the problem of high latency due to database connection establishment within individual Lambda invocations.

Refactoring Lambda functions to initialize database connections outside the handler function (in the global scope) allows connection reuse across multiple invocations on the same warm container, significantly reducing connection establishment time for subsequent requests.

Amazon RDS Proxy maintains a pool of database connections and reuses them across Lambda invocations, efficiently managing connections and drastically reducing the overhead and latency of establishing new database connections for each function call.