Productivity Tips
April 6, 2024
You are probably paying AWS too much — Part 3
Introduction
Significant efforts in your AWS cost optimization should be directed towards optimizing AWS’s Relational Database Service (RDS). RDS is considered a cornerstone of any standard infrastructure; however, its scalability and ease of management come with challenges, particularly its cost structure.
This part of our guide aims to help DevOps, CTOs, and Backend developers in navigating these challenges, showcasing our journey to cut RDS costs by over 50%.
The Pivot to Graviton2: A Leap in Cost-Performance
Our cost-saving journey starts with transitioning to Graviton2 processors for RDS instances. Following AWS’s recommendation for up to 52% better price/performance, this shift requires upgrading the MySQL engine from version 5.7 to 8.0.23 to provide the necessary suitability for Graviton2 instances.
Fill out our cloud cost optimization questionnaire and get a free consultation session with a solution architect — click here
Important! This procedural complexity involving potential downtime and the need for snapshots, the performance gains and cost savings will be immediate and notable.
The switch to M6g instances from M5 resulted in an 11.2% cost reduction per hour and a 9.14% improvement in CPU utilization efficiency.
M6g instances
Critical Strategies for Maximizing Savings
Database version upgrade: Upgrading the database engine was a prerequisite for Graviton2 adoption. This process involves meticulous planning due to the associated downtime and compatibility checks, especially for significant version upgrades, which may also require OS updates for the DB instances.
Optimizing instance types and storage: The realignment of our instance sizes, informed by performance gains and job distribution analysis, furthered our cost reduction goals. Notably, adopting T4g instances for non-production environments and adjusting storage configurations from Provisioned IOPS to General Purpose SSD for our staging databases were significant steps. Such adjustments reduced costs and aligned storage provisioning with actual needs, highlighting the importance of continuous performance and cost monitoring.
For a free infrastructure consultation session with a cloud solution architect — click here
Conclusion
Concrete Outcomes and Takeaways
Put to our tests, these strategies resulted in a staggering 50% reduction in RDS expenses, with notable savings across production and staging environments. The detailed breakdown of savings across different areas is a testament to the multifaceted approach required for effective cost optimization in cloud infrastructure.
Make sure to constantly leverage newer and updated technologies which are usually more cost-effective i.e. Graviton2 processors.
Continuously evaluate and optimize your database and instance configurations.
Implement thorough testing and adaptation to navigate potential upgrade pitfalls.
Based on your actual usage and performance data, engage in strategic downsizing and storage optimization when possible.
Subscribe to newsletter
By subscribing you agree to with our
Other sticky posts
Explore valuable insights and dive into relevant topics.
Project Management
Migrating a DynamoDB table to a Global table with 4 easy steps. No downtime no data loss.
This article is intended for organizations that have a DynamoDB table associated with a CloudFormation stack and want to convert it into a global DynamoDB table without experiencing downtime or data loss. The article also includes CloudFormation templates generated with CDK.
Productivity Tips
You are probably paying AWS too much!
A complete guide on cloud cost optimization
Productivity Tips
You are probably paying AWS too much — Part 2
Part 2: Leverage DynamoDB to your advantage