Vaclav Petricek discusses how to train models, architect and build a scalable system powered by Storm, Hadoop, Spark, Spring Boot and Vowpal Wabbit that meets SLAs measured in tens of milliseconds.
Michael Brunton-Spall shows how DevOps-like patterns can be applied on microservices to give the development teams more responsibility for their choices, and much more.
Diptanu Choudhury discusses the design of Netflix’ distributed scheduler based on Mesos and Titan, focusing on bin packing algorithms, scaling in and out of clusters, fault tolerance, and redundancy.
Small sessions on: Deterministic testing in a non-deterministic world. Hash Spreads and Probe Functions. Typesafe Config on Steroids. Real-Time Distributed Event-Driven Computing at Credit Suisse.
Benjamin Hindman discusses Apache Mesos, focusing on the Mesos API and how the primitives provided by Mesos can make it easier to build new stateful services and frameworks.
Tal Weiss explores five crucial Java techniques for distributed debugging and some of the pitfalls that make bug resolution much harder, and can even lead to downtime.
Piotr Kołaczkowski discusses how they integrated Spark with Cassandra, how it was done, how it works in practice and why it is better than using a Hadoop intermediate layer.
Oliver Gierke summarizes the problems Spring Cloud tries to solve and introduces the individual modules through practical code examples.
Dan Woods discusses the approach to developing a scalable enterprise architecture, and demonstrates implementations based on the variety of technologies available from the Groovy ecosystem.
Reid Draper shows how real world distributed database work, communicate and are tested, trading RPC for messaging, unit-tests for QuickCheck, and micro-benchmarks for multi-week stress tests.
Eugene Mandel discusses challenges of conforming data sources and compares processing stacks: Hadoop+Redshift vs Spark, showing how the technology drives the way the problem is modeled.
Michael Minella uses Spring XD and Spring Batch to orchestrate the full lifecycle of Hadoop processing and uses Apache Mahout to provide the audience with the recommendation processing.