Presented by: Bartłomiej Górny
About this talk
When demand spikes, real-time systems are pushed to their limits. Messages must be delivered instantly, and small delays can quickly escalate into system-wide failures.
Bartłomiej Górny explores what happens behind the scenes when platforms come under pressure, from overloaded components and cascading failures to practical limits in infrastructure. He argues that the real challenge is not the technology itself, but how systems are designed, scaled, and operated.
The talk focuses on building architectures that can handle sudden spikes, using approaches like service decoupling, targeted scaling, and back pressure to prevent overload. It also highlights the importance of monitoring and rigorous load testing to ensure systems remain stable under real-world conditions.
What you’ll learn:
- What causes real-time systems to fail under load
- Where bottlenecks and system limits typically appear
- How failures cascade across services
- How to design, scale, and test systems to stay stable under peak demand
