House of 1000 Coroutines: How we Debugged 3M Unbounded Coroutines in Agoda’s Kotlin Service

📰 Medium · Data Science

Learn how Agoda debugged 3M unbounded coroutines in their Kotlin service, improving performance and reliability

advanced Published 29 Apr 2026
Action Steps
  1. Identify performance bottlenecks using profiling tools
  2. Apply coroutine debugging techniques to locate issues
  3. Implement bounded coroutine pools to prevent uncontrolled growth
  4. Test and validate changes using load testing and simulation
  5. Monitor and analyze system performance after optimization
Who Needs to Know This

This article benefits backend engineers and developers working with Kotlin and coroutines, as it provides insights into debugging and optimizing complex systems

Key Insight

💡 Unbounded coroutines can lead to performance issues and crashes, but applying bounded coroutine pools and debugging techniques can help mitigate these problems

Share This
🚀 Debugging 3M coroutines in Agoda's Kotlin service: a story of performance optimization and reliability improvement
Read full article → ← Back to Reads