Answer: Change the storage type to Provisioned IOPS SSD.

## Explanation The correct answer is **A. Change the storage type to Provisioned IOPS SSD.** ### Why Option A is Correct: 1. **Storage Performance Issue**: The problem statement explicitly states that "the company has determined that the database storage performance is the problem." This indicates the issue is with I/O performance, not compute or memory. 2. **General Purpose SSD Limitations**: General Purpose SSD (gp2) storage provides baseline performance of 3 IOPS per GB, with the ability to burst to 3,000 IOPS. For a 2 TB volume, this provides: - Baseline: 2,000 GB × 3 IOPS/GB = 6,000 IOPS - Burst: Up to 3,000 IOPS (but limited by burst balance) 3. **Provisioned IOPS SSD (io1/io2)**: This storage type allows you to provision specific IOPS (up to 256,000 IOPS for io1 and 256,000 IOPS for io2) independent of storage size, providing consistent low-latency performance. 4. **High Update Volume**: With "millions of updates every day," the database requires consistent high I/O performance that Provisioned IOPS SSD can provide. ### Why Other Options are Incorrect: **B. Change the DB instance to a memory optimized instance class.** - Memory optimized instances (R5, X1, etc.) are designed for workloads that require large amounts of memory, not specifically for storage I/O performance. - While they may help with caching, they don't directly address the identified storage performance issue. **C. Change the DB instance to a burstable performance instance class.** - Burstable instances (T3, T4g) provide baseline CPU performance with the ability to burst, but they're not designed for consistent high-performance database workloads. - This would likely make performance worse for a database with millions of daily updates. **D. Enable Multi-AZ RDS read replicas with MySQL native asynchronous replication.** - Read replicas help with read scalability and disaster recovery, but they don't improve write performance for the primary database. - The issue is with insert operations (writes), not reads. - Multi-AZ provides high availability through synchronous replication to a standby instance, but doesn't improve storage I/O performance. ### Key AWS Concepts: - **General Purpose SSD (gp2/gp3)**: Best for small to medium-sized databases with moderate I/O requirements - **Provisioned IOPS SSD (io1/io2)**: Designed for I/O-intensive workloads requiring consistent performance - **Storage vs. Compute**: It's important to distinguish between storage performance issues (I/O) and compute/memory issues ### Additional Considerations: 1. Monitor RDS metrics like `ReadIOPS`, `WriteIOPS`, `ReadLatency`, `WriteLatency`, and `QueueDepth` to confirm storage bottlenecks. 2. Consider using RDS Performance Insights to identify specific queries causing I/O issues. 3. For extremely high-performance requirements, consider Amazon Aurora which provides better performance than standard RDS MySQL.