Building Resilient Asynchronous Commerce Systems
summary
Claim: observability is the linchpin that makes async event‑driven commerce reliable. Concrete moves include decoupling payments and building an async BOPIS flow, instrumenting trace context with OpenTelemetry, and monitoring Kafka with Burrow plus Grafana/Jae ger/Kibana dashboards. Surprising scale: 100k orders at NPI spikes, 7M monthly logins, 130k carts vs 35k orders baseline, and measured 97% availability under load. Outcomes: lower abandonment, higher conversion, and actionable visibility via queue depth monitoring and event replay.
AI-written summary · about
chapters (8)
- 0:00Introduction and Speaker Background
- 0:16The Importance of Observability in Async Systems
- 1:49Challenges and Solutions in Async Commerce
- 3:12Case Study: T-Mobile's Async Implementation
- 5:51Tools and Techniques for Observability
- 18:14Testing and Monitoring Async Systems
- 20:05Lessons Learned and Best Practices
- 21:43Conclusion and Final Thoughts
related talks
54mReacting to an Event-Driven World
25mAccelerate Event Enabling the API Economy
44mUnlocking Business Agility with Event-Driven Architectures
45mE-commerce platform. Core build principles, problems and guidelines.
52mSo You Want to Build An Event Driven System?
01h 02mAsynchronous Distributed Systems
28mEvent Driven API Management - Increase Reusability with AsyncAPI
33mHigh availability by offloading work