Coursera Flash Sale
40% Off Coursera Plus for 3 Months!
Grab it
Learn about Uber's groundbreaking journey in managing one of the world's largest Apache Kafka clusters through this technical conference talk. Dive deep into the evolution of uReplicator, Uber's custom Kafka replication system developed in 2016 to overcome MirrorMaker 1.0's limitations. Explore the architectural intricacies of a system that handles trillions of daily messages across 20+ clusters while maintaining exceptional availability and minimal replication delays. Master the implementation of advanced features like Auto-Rebalance for autonomous cluster management, intelligent auto-scaling based on traffic patterns, and sophisticated observability tools that simplify operations. Gain valuable insights into overcoming large-scale data streaming challenges through real-world examples from Uber's ecosystem, which processes everything from system metrics and application logs to database changelogs and rider/driver/eats app event data. Discover practical lessons and best practices for implementing robust Kafka replication systems at scale.
