Scaling Yelp’s Logging Pipeline with Apache Kafka

06/06/2016 - 17:20 to 18:00
Frannz Salon
long talk (40 min)

Session abstract: 

This talk is the story of the transformation of a legacy architecture into a robust multi-region streaming pipeline which relies on Apache Kafka as a core component in order to transport and access terabytes of logs and billions of messages every day. I’ll present the pitfalls and challenges we faced during our journey and how, with proper design, they were overcome in order to meet the expectations of our engineers, who are the final users of this system.