Book Image

Apache Spark 2 for Beginners

By : Rajanarayanan Thottuvaikkatumana
Book Image

Apache Spark 2 for Beginners

By: Rajanarayanan Thottuvaikkatumana

Overview of this book

<p>Spark is one of the most widely-used large-scale data processing engines and runs extremely fast. It is a framework that has tools that are equally useful for application developers as well as data scientists.</p> <p>This book starts with the fundamentals of Spark 2 and covers the core data processing framework and API, installation, and application development setup. Then the Spark programming model is introduced through real-world examples followed by Spark SQL programming with DataFrames. An introduction to SparkR is covered next. Later, we cover the charting and plotting features of Python in conjunction with Spark data processing. After that, we take a look at Spark's stream processing, machine learning, and graph processing libraries. The last chapter combines all the skills you learned from the preceding chapters to develop a real-world Spark application.</p> <p>By the end of this book, you will have all the knowledge you need to develop efficient large-scale applications using Apache Spark.</p>
Table of Contents (15 chapters)
Apache Spark 2 for Beginners
Credits
About the Author
About the Reviewer
www.PacktPub.com
Preface

Monitoring with Spark


The previous chapter covered the details of the installation and development tool setup that is required for developing and running data processing applications using Spark. In most of the real-world applications, the Spark applications can become very complex with a really huge directed acyclic graph  (DAG) of Spark transformations and Spark actions. Spark comes with really powerful monitoring tools for monitoring the jobs that are running in a given Spark ecosystem. The monitoring doesn't start automatically.

Tip

Note that this is a completely optional step for running Spark applications. If enabled, it will give a very good insight into the way the Spark applications are run. Discretion has to be used to enable this in production environments, as it can affect the response time of the applications.

First of all, there are some configuration changes to be made. The event logging mechanism should be turned on. For this, take the following steps:

$ cd $SPARK_HOME 
...