Logging at Scale in Production - Derry O'Sullivan & David Corley

Derry O'Sullivan and David Corley were at the core of building our internal real time logging system based on the ELK (Elasticsearch/Logstash/Kibana) stack & Apache Kafka. They'll talk about the history of researching/deploying the solution, the architecture used, the challenges they faced scaling the system to billions of log events a day as well as the benefits it now provides to Workday

Running ZooKeeper at Scale - Daniel McGrogan & Igor Belsa

Daniel & Igor will explain how the Workday development team ensures continued scaling of the Workday platform though innovative applications of ZooKeeper. We will discuss the techniques we apply to achieve high volume reads and writes beyond traditional Service Discovery and configuration management patterns.

Workday has recently expanded ZooKeeper usage, beyond Service Discovery, in support of its feature-toggle management. Improvements in this area focus on write-performance and enable clients to write hundreds of thousands of logical nodes to ZooKeeper while experiencing 30x performance improvements over typical ZooKeeper writes.