Recorded at SpringOne2GX 2015. Speaker: Stephane Maldini, Glenn Renfo, David Turanski Big Data track Slides: http://www.slideshare.net/SpringCentral/high-performance-stream-processing-53163754

Making things go fast is harder than it looks. In this talk we dissect a Spring XD application into three areas - message transport, data serialization, and application processing - to show design patterns and techniques that will help you get the highest throughput and lowest resource utilization for your streaming applications. We will compare and contrast in-memory, RabbitMQ, and Kafka transports; Kryo, Protocol Buffers and C24 Simple Data Objects serialization libraries; and processing patterns that use a traditional shared queue and thread pool dispatcher in contrast to a Reactive Streams approach.