Unlocking Performance: Infrastructure Enhancements for Optimal Data Operations

At Qryptal, we are constantly evolving our infrastructure to deliver the best possible experience to our customers. Our multi-tenant cloud architecture forms the backbone of Qryptal’s secure QR code generation and verification services, supporting a wide range of customer needs. We also provide tailored deployments for customers requiring dedicated instances or on-premise infrastructure.

In late 2024, we launched version 2.5 of our multi-tenant infrastructure—an upgrade that now powers all legacy customers and most new deployments. This release focused on improvements in scalability, reliability, and adaptability, ensuring our platform remains responsive and efficient for diverse use cases.

🔍 Proactive Monitoring & Intelligent Alerting

Beyond traditional reactive methods, our infrastructure is designed with proactive monitoring and alerting mechanisms. We leverage tools like AWS CloudWatch to continuously track critical health indicators such as:

• API latency
• CPU, memory, and disk usage
• Network throughput

Screenshot 1: Alarms by AWS Services.

Screenshot 2: Recent Alarms.

This real-time monitoring enables us to identify and resolve potential issues before they impact customers, helping us maintain a consistently high-quality service.

🌍 Tackling Cross-Region Data Repository Challenges

Recently, we encountered an interesting performance challenge involving a customer who switched to using an AWS S3 bucket in a different region for their encrypted Additional Data Repository (ADR)—used in conjunction with our digitally signed QR codes.

While we had already implemented a custom ADR feature allowing clients to configure their own storage (typically in the same region, whether on AWS or Azure), this was the first major case involving a cross-region ADR.

Performance Impact

This setup resulted in a dramatic performance drop:

• QR code generation latency increased by ~300%

• Data updates—which occurred frequently—suffered a 500–600% increase in latency

To pinpoint the cause of the slowdown, we first confirmed that the size of encrypted data attachments, such as PDFs and images, stored in the ADR had not increased. This allowed us to definitively attribute the performance degradation to the utilization of a remote region.

These issues flagged an urgent need for infrastructure optimization.

🚀 Engineering Response: Persistent Connection Optimization

Our engineering team swiftly diagnosed the root cause: the inherent latency of cross-regional data transfer compounded by the overhead of frequent connection setups.

To address this, we introduced an industrial-grade caching mechanism built around persistent client connection management. Instead of initiating a new connection for each API call, the system now reuses open connections for as long as feasible.

Results

Chart 1: Average API Latency.

Cross-region latency initially caused a 3–6x performance hit, which we halved through persistent connection reuse.

The impact was immediate and significant:

• Generation and update times were reduced by nearly 50%

• Cross-region latency was greatly mitigated, improving overall responsiveness

Significant real-world performance improvements were achieved, even with some remaining latency due to physical distances. This initial enhancement was sufficient for immediate needs and can be further optimized by utilizing S3 Transfer Acceleration or co-locating processing services closer to the data, as circumstances dictate.

🗄️ Eliminating Redundant Updates with Shared Caching

Following this success, we partnered with the customer to further analyze their usage patterns. We discovered that multiple services within their system were triggering identical data writes, leading to unnecessary and redundant updates.

Since changes to their architecture weren’t feasible, we implemented a solution on our side:

Shared Cache Across Nodes

We introduced a shared caching mechanism across our multi-tenant nodes. This system intelligently detects redundant write operations and avoids duplicating them, thereby:

• Reducing compute and storage overhead

• Improving latency and throughput

• Increasing overall system efficiency

🎯 Conclusion - Continuous Evolution

Our engineering team really knocked it out of the park, and we’re so grateful for the partnerships we have with our customers that help us solve these technical puzzles together. This journey really highlights how we build things here at Qryptal:

• We listen to you – real-world challenges help us find things that even our best tests might miss.
• We build for the long haul – optimizations like caching aren’t just quick fixes; they’re the foundation for lasting performance improvements.
• We never stop improving – infrastructure is always evolving, and every new challenge pushes us to get better.

As we keep refining our multi-tenant setup, you can look forward to even better performance, smarter handling of your requests, and an even more dependable platform for your secure, digitally signed QR codes.

Recommendation: You may like to explore more …

Solved: Ability to include Photographs in Secure PDC QR codes

Preventing Counterfeit Degrees: How Secure QR Codes Protect Academic Credentials

Integrating Sustainability and Financial Inclusion to Drive Africa’s Economic Growth

Navigating the Future of Document Security with Qryptal

Whitepaper on Document Security