Building Resilient Asynchronous and Event-Driven Systems

Asynchronous and event-driven architectures promise scalability and durability, but they often mask complex failure modes that only surface under unpredictable situations like large traffic spikes or dependency failures. These problems become even more complicated when you are serving multiple customers, i.e, in multi-tenant systems, which is often the case. This training dives into real-world design challenges - drawing on over a decade of experience operating multi-tenant, high-throughput systems. We will explore less talked about architecture patterns like shuffle sharding, retry storms, back-pressure, and concurrency limiting that allow systems to degrade gracefully instead of collapsing under pressure. Attendees will leave with mental models, metrics, and playbooks for building systems that are prepared to fail—and recover—predictably.

By the end of the training session, you will have confidence in building resilient asynchronous and event-driven systems that don’t collapse and lead to gradual recovery when unexpected situations hit.