Big Data processing frameworks like Apache Flink make it easy to write distributed programs, but it can be difficult to go from understanding simple tutorial applications to writing scalable, production-ready code.
BigPetStore, which is a part of Apache BigTop, is a set of interrelated applications that use a sophisticated synthetic data generator to generate realistic workloads for various data processing frameworks, including Hadoop and Spark, and to serve as a nontrivial example application for these. In this talk, we will discuss our efforts to extend BigPetStore to support Apache Flink while taking advantage of Flink’s unique features like its streaming-centric model, iterative stream processing and windowing semantics.
BigPetStore-Flink provides a realistic example application for streaming analytics at large scale, a platform for evaluating the performance of Flink deployments, and a concise exploration of the differences between Flink and other data processing frameworks.
About the speaker
Suneel Marthi is a Senior Principal Software Engineer in the Office of Technology, Red Hat. He is the PMC Chair and Committer on Apache Mahout and a core contributor on DeepLearning4J and Oryx 2.0.
He is the Co-Organizer of the Washington DC Area Apache Flink meetup and founder of the Boston Area Apache Flink meetup.