Stop Throwing Hardware at Problems: Optimise Databases First

As a database specialist at DNX, I’ve seen this countless times: When performance issues arise, many businesses default to scaling up hardware. But what if the real solution lies in database optimisation? 

A recent customer engagement reminded me that the best solution often lies not in bigger servers but in smarter code. Let me walk you through how we helped a client improve their app’s performance, save up to 47% on costs, and reduce CPU stress by 60%—without touching their hardware.

The Challenge before the Database Optimisation: Performance, Costs, and User Experience

At DNX, we believe in starting with an investigation before rushing to a solution. Scaling up hardware is a one-click fix — especially with cloud services like AWS —  but adding more instances without resolving underlying issues just masks the problem. 

For this customer, it wasn’t just a technical issue; it was about fixing inefficiencies that were eating into their budget and frustrating users – a very business problem. Here’s what they were dealing with:

• Escalating Costs: Their Amazon RDS instance costs were getting more expensive, driven by resource-heavy database queries.
• Performance Bottlenecks: A critical aggregation job running every 10 minutes across three stored procedures was maxing out CPU resources and slowing everything down.
• Negative User Experience: Users faced over 10-minute waits for updates, making the app frustratingly slow and uncompetitive in the market.

Step-by-Step Approach: Improving Database Performance

Step 1: Understand the Code and Processes

Our first priority was to get a complete picture of the client’s code, database setup, and the processes at play. We collaborated with their team to map how their database was impacting business operations.

• Pinpoint the problem: We identified that a complex aggregation task ran every 10 minutes, consuming excessive CPU resources.
• Assess performance impact: The strain from these tasks affected not just the database, but also other crucial areas of the app.
• Evaluate user experience: The lag caused by the slow aggregation left users waiting over 10 minutes for reports, which was a significant drawback.

Step 2: Identify the Root Cause

After thorough analysis, we determined the main contributors to the performance issues:

• Slow data aggregation: The frequency and complexity of the aggregation tasks were putting unnecessary strain on the system.
• Inefficient data processing: The existing data handling methods were not optimised for performance, leading to excessive CPU load.

Step 3: Implement Strategic Code Refactoring and Optimise Database

Based on our findings, we proposed a solution that many might overlook: strategic code refactoring. Optimising the code would provide a more sustainable solution, improving performance while reducing ongoing costs. Our team worked closely with the client’s developers to implement the necessary changes:

• Optimising queries: We applied a carefully crafted delta to the existing codebase to minimise the frequency and volume of data aggregation.
• Enhancing indexing: Improved indexing strategies for quicker data retrieval and reduced processing time.

The Results: Database Optimisation in Action

The implementation of our code optimisation strategy delivered results that exceeded both the customer’s and our expectations. By addressing the root cause of the database performance issues and implementing the code refactor, we achieved:

• CPU Usage: Reduced by 60%, significantly lowering strain on the system.
• Aggregation Time: Dropped from 12-10 minutes to just 45 seconds—a 93% improvement in efficiency.
• Database Instance Costs: Reduced by 47%, as the optimised system allowed us to resize the database instance to a smaller and more cost-effective type.
  • RDS Cost Before DNX: AUD $1,419/month
  • RDS Cost After DNX: Initially reduced to AUD $884/month and planned to go down further to AUD $489/month with additional tuning.

The Broader Impact of Database Performance Optimisation

The benefits of optimising the client’s database extended beyond solving the immediate technical challenges. The changes created lasting value across multiple facets of their business:

• Enhanced User Experience
  • The app became significantly more responsive, enabling real-time insights.

• • Users could access reports in seconds instead of minutes, improving satisfaction and engagement.

• Increased Capacity

• • Freed-up CPU resources allowed the system to handle higher workloads without additional infrastructure.
  • Scalability improved, positioning the business for future growth without concerns over immediate bottlenecks.

• Cost Optimisation

• • Infrastructure costs dropped by 47%, saving thousands annually.
  • The optimised instance size continues to reduce operational expenses.

• Knowledge Transfer

• • DNX worked closely with the client’s team to ensure they understood the changes, enabling them to maintain and build on the improvements.
  • This collaboration empowered the client’s developers to approach future challenges with greater confidence.

• Sustainability Goals

• • By reducing CPU usage by 60%, energy consumption decreased, contributing to environmental sustainability.
  • The optimised system aligns with modern best practices for energy-efficient computing.

Conclusion: The DNX Approach to Database Optimisation and Efficiency

This project exemplifies the DNX philosophy: “Stress your code before stressing over your hardware.” It’s a principle that consistently delivers tangible, cost-effective results for our clients.

Key Takeaways:

This engagement reinforced critical lessons about database performance optimisation and how it aligns with business success:

1. Code Efficiency is Paramount: This case reinforces our belief that efficient code can often outperform hardware upgrades.
2. Expert Analysis is Crucial: Our specialised knowledge in database management allowed us to identify issues that weren’t immediately apparent.
3. Cost-Effective Solutions Exist: By avoiding unnecessary hardware upgrades, we saved our customer significant infrastructure costs.
4. Performance and Scalability Go Hand in Hand: The optimised code not only solved immediate issues but also improved overall database performance and scalability.

At DNX, we believe in finding solutions that align with business goals, sustain long-term success, and build a strong foundation for future growth. This is a prime example of how focusing on the right approach and database optimisation can transform a seemingly small change into a powerful, long-lasting impact.

In the end, this story is about more than just fixing a technical problem—it’s about creating a ripple effect of improvement that touches every part of the business, from the team to the end users. And that’s where DNX excels: turning challenges into opportunities for optimisation and growth.