TUI Enhances Real-Time Performance

Challenges

When Simon Vos (Technology Director) joined TUI, he brought extensive experience in backend development and travel systems. He identified a need for enhanced performance testing, particularly for TUI's new real-time pricing and availability system.

Traditional methods that relied on pre-calculated data and caches were insufficient for the dynamic nature of TUI’s offerings, where prices and availability fluctuate rapidly. The goal was to ensure the system could handle real-time calculations without compromising on speed or accuracy.

Solution

To address these challenges, TUI adopted Gatling Enterprise to integrate performance testing directly into their development pipelines. Simon and his team focused on embedding performance testing within their CI/CD processes, ensuring that every update could be rigorously tested for performance impacts before reaching production.

Implementation

Initially using Gatling Enterprise On-Premise solution, the team later transitioned to Gatling Enterprise on Cloud for greater scalability and ease of use. They integrated performance tests into their GitLab CI/CD pipelines, enabling automated testing at various stages of development and deployment.

“To ensure our new pricing and availability engine is able to scale and can handle the load of our customers' live searches, we chose Gatling Enterprise on Cloud as our performance and load testing solution. We can track our performance over time and allow our development teams to automate the scenarios and runs. Gatling enabled us to automate the performance impact assessment of new feature delivery continuously and before releases.”

Simon Vos, Technology Director, TUI

Results

The integration of Gatling has significantly improved TUI's ability to handle large volumes of real-time data requests. Some of the key benefits include:

• Enhanced performance monitoring: With real-time metrics, TUI can now monitor the performance of their services continuously. This allows for quick identification and resolution of any performance issues.
  
  
• Scalability: Gatling’s ability to generate high loads (over 100,000 requests per second) has proven crucial for stress testing TUI’s systems, ensuring they can handle peak traffic efficiently.
  
  
• Automation and efficiency: The automated nature of Gatling’s tests has reduced the need for manual performance testing, saving time and resources while increasing the reliability of the tests.
  
  
• Response time: The implementation of Gatling has helped reduce response times from initial figures as high as 10-20 seconds to well within the target of less than one second for 95% of requests.
  
  
• System reliability: The focus on performance testing has significantly improved the stability and reliability of TUI’s pricing and availability system, leading to a better customer experience.
  ‍

‍