Microservices promise independent deployments and scalability, but many teams end up with systems that are fragile, tightly coupled, and difficult to operate.

In this talk, we look at the most common failure modes seen in real-world microservices architectures. From availability coupling and cascading failures to communication patterns that look good on diagrams but fail under load, we’ll explore why these issues emerge and how teams unintentionally design themselves into trouble.

The session focuses on architectural thinking rather than blame, showing how small decisions compound over time. We’ll also look at practical ways to reduce risk and improve resilience, with concrete examples grounded in AWS-based systems.

Target audience: backend developers and software architects working with microservices.
Session format: architecture-focused talk with real-world examples.
Duration: available as a 30 or 45-minute session.
Level: intermediate. Familiarity with microservices concepts assumed.