Case Study: ZeroFlucs achieves worldwide low-latency sports betting with ScyllaDB

How data is replicated to support low latency for ZeroFlucs’ global usage patterns without racking up unnecessary costs

ZeroFlucs, a sports betting technology company, needed to process live in-play events in near real time while handling very high throughput and global demand. Their challenge was to keep latency low for customers anywhere in the world, even as each event update triggered tens of thousands of simulations and about 250,000 in-game events per second.

To meet those needs, ZeroFlucs adopted ScyllaDB on Oracle Cloud and used ScyllaDB Operator to manage their distributed database architecture. They partitioned data into global, regional, and customer-specific keyspaces so replicas could be placed close to services and users, reducing latency while controlling storage costs; ZeroFlucs also created Charybdis to automate keyspace and table management. The result was a scalable, low-latency system that supported their real-time pricing workload and could be extended for disaster recovery and future growth.

ZeroFlucs

Carly Christensen

Director of Software Engineering