How IGI Cut Nodeware® RDS CPU Usage 97.5% by Optimizing AWS RDS and Lambda with Mindex

“As we continue through a period of rapid growth and increased adoption of our product, our focus has shifted from development to scaling our solution. Given our commitment to building on the AWS platform, we needed a partner with an in-depth understanding of our technology. We were fortunate to find that partner locally through Mindex.”

Brian Drake

Director of Technology Development

Infinite Group, Inc. (IGI)

Opportunity

The Challenge: Rising Costs and Slow Database Performance as Nodeware Grew

• Growth in usage increased IGI’s operational costs.
• Nodeware’s AWS Lambda functions faced performance issues and reached their maximum concurrency executions.
• The main AWS Relational Database Service (RDS) instance consistently operated near 100% CPU usage, slowing database operations and adding expense.
• IGI’s PDF builder application was slower and required more upkeep than ideal, and IGI wanted a more efficient, automated solution for generating the HTML and PDF reports its cybersecurity customers rely on for compliance and audits.

Mindex Approach

The Solution: AWS RDS and Lambda Cost Optimization and an Automated PDF Workflow

• Mindex’s discovery showed slow performance in specific RDS queries. Adding a necessary index to the table cut query execution time from seconds to just one second for 1,000 queries and reduced RDS CPU utilization by 75%, which improved Lambda efficiency.
• After resizing the RDS instance to fit the lighter workload, Mindex made additional database structure optimizations based on a review of performance bottlenecks. This AWS RDS and Lambda cost optimization work led to a 90% decrease in CPU usage, allowing the RDS instance to be resized again.
• To automate the PDF process, Mindex’s cloud team integrated IGI’s HTML/PDF builder with the third-party API that IGI had explored and found promising in testing:
  • When an item is added to the HTML bucket, a Lambda function triggers the API to create a new PDF.
  • The API notifies IGI’s API gateway when the job is complete, and the PDF is given a new name for future reference.
  • The PDF’s name, status, and count are recorded in a DynamoDB table (statuses: “delivered,” “creation failed,” and “delivery failed”), so the file can be retrieved and displayed on the web server.

Outcomes

The Results: A 97.5% Reduction in RDS CPU Usage and an Automated PDF Workflow

• Query execution time decreased, improving system responsiveness and user experience.
• Resolving RDS inefficiencies produced a combined 97.5% reduction in CPU utilization, and the resized RDS instance now costs less than 15% of its original expense.
• The PDF builder integration saved time and improved operational efficiency and user experience.
• Mindex’s expertise as an AWS Advanced Tier Services Partner provided insights that led to efficient optimization efforts and valuable time savings.

Ready to Get Started?

Revamp Your Cloud Efficiency

Our cloud services team will guide you through a transformative journey to optimize performance while minimizing costs.