Ultimate access to all questions.
A company's application, which runs on AWS Lambda, utilizes an Amazon Aurora PostgreSQL DB cluster consisting of one small primary instance and three larger replica instances for read-only operations. During high traffic periods, the application's reliability diminishes due to an excessive number of database connections. Given the unpredictable nature of these traffic spikes, which solution will enhance the application's reliability?