Validating Backend Modernization Through Front-End Performance Testing with Octoperf
We are launching with this post a new series of blog articles and LinkedIn posts titled "Feedback From the Field".
Defion, our partner based in the Netherlands and Spain (formerly "Computest"), is kicking things off with this project testimonial.
Enjoy the read.
Modernization Is Not Just an Infrastructure Exercise¶
A financial institution recently initiated a backend modernization program, transitioning from legacy components to a more scalable, service-oriented architecture.
The objectives were clear:
- Improve scalability and resilience
- Enable faster feature development
- Reduce operational complexity
However, modernization only delivers value if it improves what customers actually experience. In digital banking, that experience is defined by the responsiveness and stability of the front-end applications.
The key question became:
Does the new backend architecture translate into stable and responsive user-facing systems under real-world load?
The Challenge¶
The digital banking platform consisted of multiple layers:
- Web and mobile front-end channels
- API gateways and backend microservices
- Core banking integrations
- Security layers including authentication and authorization services
The modernization effort introduced several architectural changes:
- New service boundaries
- Updated API endpoints
- Modified data access patterns
- Increased inter-service communication
While these changes are intended to improve scalability and maintainability, they can also introduce additional latency, configuration issues, or unexpected bottlenecks.
Before proceeding with full rollout, the bank required objective validation of system behavior under realistic load conditions.
Why Octoperf¶
We used Octoperf to simulate realistic user behavior and measure both end-to-end and component-level performance.
The reasons were practical:
- Distributed cloud-based load generation, with the option to deploy private internal load generators to reach backend systems inside the private cloud network
- Support for complex, stateful user journeys
- JMeter compatibility, enabling the use of open-source scripting and existing test assets
- Real-time visibility into response time distributions and error patterns
- The ability to quickly execute tests, implement fixes, and rerun validation cycles
Within a modernization program, the ability to run comparable tests against both legacy and modernized environments was particularly valuable.
Test Design: Front-End Driven, Backend Focused¶
Rather than testing isolated services, the scenarios modeled realistic user interactions across the digital banking platform.
These interactions included:
- Secure login with stubbed MFA
- Account overview retrieval
- Transaction execution such as transfers and payments
- High-frequency API calls triggered by front-end components
The testing approach consisted of several phases:
- Baseline performance testing on the legacy backend
- Equivalent load scenarios against the modernized architecture
- Peak-load simulations aligned with projected growth
- Stress testing to observe failure modes
- Endurance testing to verify long-term stability
The primary focus areas were:
- End-to-end response times
- P95 and P99 latency metrics
- Error rates under concurrent usage
- Resource behavior across service layers
What Octoperf Revealed¶
Initial testing highlighted several important behaviors:
- Increased response time variance under peak concurrency
- Error rates appearing at higher load levels
- Improved scalability compared to the legacy architecture
- A new bottleneck at the API gateway layer during burst traffic scenarios
Without realistic front-end simulation, these patterns would likely not have been visible.
Octoperf's reporting capabilities allowed us to correlate user-facing delays with backend service behavior, enabling targeted optimizations rather than broad assumptions.
Final Outcome¶
After remediation and configuration adjustments:
- P99 response times stabilized under peak load
- Transaction throughput (TPS) increased
- Error rates during stress scenarios decreased
- Front-end responsiveness improved measurably
The bank gained confidence that backend modernization translated into tangible improvements in the user experience.
Final Reflection¶
Modernizing backend systems in a financial institution is not simply an architectural exercise. It directly impacts customer experience, transaction reliability, and operational risk.
Performance testing with Octoperf provided the structured validation needed to ensure that architectural changes resulted in measurable performance improvements.
Because in digital banking, "modern" means nothing if it's slower.