How CloudFix Consulting Slashed RDS Costs for a Major Low-Code PaaS Vendor

This case study highlights the innovative approach taken to overcome technical constraints and achieve substantial
savings.

The Challenge

The client was operating thousands of m5.large SQL Server Standard RDS instances, each with 1-year partial upfront Reserved Instances (RIs) applied. While these RIs provided a 25% discount, the effective rate for each instance was still $0.730/hour or $6,395/year. With RDS costs running into tens of millions of dollars annually, drastic cost reduction was imperative.

Two main challenges presented themselves:

1. The use of RDS Reserved Instances locked the client into an older generation of instances (m5 family), preventing an upgrade to the more efficient m6i/m7i families.
2. The client’s product was deeply integrated with SQL Server, making an immediate switch to a less expensive database engine unfeasible.

The Solution

After careful analysis, CloudFix Consulting proposed a bold yet effective solution: migrating from Amazon RDS to Amazon EC2 instances running SQL Server Standard on Linux.

Why EC2?

While Amazon RDS is a managed service that handles many administrative tasks, the cost savings potential of EC2 was too significant to ignore. According to AWS documentation, EC2 offers more control over the database instance and operating system, which aligned well with the client’s needs.

Performance Benefits of Newer Instance Types

In addition to cost savings, the migration to EC2 m7i instances provided significant performance improvements.

1. Improved Transactions Per Minute: M7i instances showed up to 1.41x higher transactions per minute (TPM) compared to M5 instances, and up to 1.15x higher TPM compared to M6i instances.
2. Increased Query Performance: For specific database operations, M7i instances showed notable improvements:

• Up to 1.46x faster performance for point selects
• Up to 1.44x faster performance for simple range selects
• Up to 1.39x faster performance for sum range selects

These performance enhancements not only improved the client’s application responsiveness but also allowed for potential consolidation of workloads, further contributing to cost optimization.

The Results

The migration from RDS to EC2 yielded impressive results:

1. Per-Instance Savings: Each database instance now saves $1,787 per year.
2. Overall Impact: Given the thousands of instances in use, this translated to millions of dollars in annual savings.
3. Additional Benefits: The increase in EC2 spend allowed for the deployment of RightSpend cRIs, leading to an extra $1 million per year in savings on non-database instances.
4. Performance Gains: The move to newer instance types provided significant performance improvements, enhancing the overall user experience of the client’s platform.

Looking Ahead

While this solution provided immediate and substantial cost savings, CloudFix Consulting also advised on long-term strategies. The client’s development team is working on a new version of their product that will support PostgreSQL, which could lead to even greater savings (up to 85%) once fully implemented and migrated.

Conclusion

This case study demonstrates the value of innovative thinking in cloud cost optimization. By challenging the status quo and leveraging deep AWS knowledge, CloudFix Consulting was able to deliver significant value to their client, proving that sometimes the best solutions require thinking outside the “managed service” box. The combination of cost savings and performance improvements made this migration a resounding success.