Book Image

Learning Apache Apex

By : Thomas Weise, Ananth Gundabattula, Munagala V. Ramanath, David Yan, Kenneth Knowles
Book Image

Learning Apache Apex

By: Thomas Weise, Ananth Gundabattula, Munagala V. Ramanath, David Yan, Kenneth Knowles

Overview of this book

Apache Apex is a next-generation stream processing framework designed to operate on data at large scale, with minimum latency, maximum reliability, and strict correctness guarantees. Half of the book consists of Apex applications, showing you key aspects of data processing pipelines such as connectors for sources and sinks, and common data transformations. The other half of the book is evenly split into explaining the Apex framework, and tuning, testing, and scaling Apex applications. Much of our economic world depends on growing streams of data, such as social media feeds, financial records, data from mobile devices, sensors and machines (the Internet of Things - IoT). The projects in the book show how to process such streams to gain valuable, timely, and actionable insights. Traditional use cases, such as ETL, that currently consume a significant chunk of data engineering resources are also covered. The final chapter shows you future possibilities emerging in the streaming space, and how Apache Apex can contribute to it.
Table of Contents (17 chapters)
Title Page
Credits
About the Authors
About the Reviewer
www.PacktPub.com
Customer Feedback
Preface

Fault-tolerance components and mechanism in Apex


In Chapter 2, Getting Started with Application Development, we looked at the deployment of an Apex application when it is executing on a YARN cluster. Let's revisit the diagram to see which type of failures may occur and how they are handled by the system:

The client is only required for launching the application; it is not involved in the execution of the DAG on the cluster, and failure of the client node does not affect the pipeline. Since Apex is running on YARN, let's first see how YARN supports resilient applications (from a user's perspective).

YARN consists of a resource manager (RM) and node managers (NM). Each YARN cluster node has a node manager service running, which communicates with the resource manager.

  • Failure of an NM: When a node manager fails (regardless of software or hardware failure), the RM will detect this and ensure that the affected containers can be allocated on different machines. The RM itself cannot recover those...