Book Image

Mastering Apache Spark 2.x - Second Edition

Book Image

Mastering Apache Spark 2.x - Second Edition

Overview of this book

Apache Spark is an in-memory, cluster-based Big Data processing system that provides a wide range of functionalities such as graph processing, machine learning, stream processing, and more. This book will take your knowledge of Apache Spark to the next level by teaching you how to expand Spark’s functionality and build your data flows and machine/deep learning programs on top of the platform. The book starts with a quick overview of the Apache Spark ecosystem, and introduces you to the new features and capabilities in Apache Spark 2.x. You will then work with the different modules in Apache Spark such as interactive querying with Spark SQL, using DataFrames and DataSets effectively, streaming analytics with Spark Streaming, and performing machine learning and deep learning on Spark using MLlib and external tools such as H20 and Deeplearning4j. The book also contains chapters on efficient graph processing, memory management and using Apache Spark on the cloud. By the end of this book, you will have all the necessary information to master Apache Spark, and use it efficiently for Big Data processing and analytics.
Table of Contents (21 chapters)
Title Page
Credits
About the Author
About the Reviewer
www.PacktPub.com
Customer Feedback
Preface
10
Deep Learning on Apache Spark with DeepLearning4j and H2O

Streaming sources


We will not be able to cover all the stream types with practical examples in this section, but where this chapter is too small to include code, we will at least provide a description. In this chapter, we will cover the TCP and file streams and the Flume, Kafka, and Twitter streams. Apache Spark tends only to support this limited set out of the box, but this is not a problem since 3rd party developers provide connectors to other sources as well. We will start with a practical TCP-based example. This chapter examines stream processing architecture.

Note

For instance, what happens in cases where the stream data delivery rate exceeds the potential data processing rate? Systems such as Kafka provide the possibility of solving this issue by caching data until it is requested with the additional ability to use multiple data topics and consumers (publish-subscribe model).

TCP stream

There is a possibility of using the Spark Streaming Context method called socketTextStream to stream...