A production Multi-AZ Amazon RDS for MySQL (db.m6g.xlarge) is hitting 'too many connections' errors frequently. The administrator wants to minimize code changes and fix the problem cost-effectively. Which approach best meets these constraints?

Configuring RDS Proxy is the most effective solution because it pools and shares database connections, reducing the overhead of establishing new connections and preventing 'too many connections' errors without requiring significant application code changes. It also improves application resilience by automatically routing connections to a healthy replica during failovers. Scaling up the instance might temporarily alleviate the issue but doesn't address the underlying connection management problem and is more expensive. Migrating to DynamoDB is a major architectural change, not a cost-effective or minimal-code-change solution for a relational database issue. Switching to a memory-optimized class is unlikely to resolve connection limits, as these are typically governed by database configuration parameters and connection handling, not primarily memory.