Sharding: One Service’s Tale

Unfortunately, not everyone thinks about scalability when starting to build services — especially when the task seems simple at first. Our team was no exception: over time, the database behind this “simple” service grew to several terabytes, and it became clear that sharding was inevitable.

In this talk, I’ll share how we implemented sharding for Ozon’s business-critical warehouse service — with zero downtime and under 24/7 operational load. I’ll walk through the challenges we faced and explain why sharding is about much more than just schemas and keys. I’ll also share practical solutions, architectural trade-offs, and key lessons learned — so you don’t have to repeat our mistakes.

Whether you’re working with high-load systems or just starting to design with scale in mind, this experience could save you time and headaches.