Author: Zoiner Tejada
Publisher: O'Reilly
Pages: 412
ISBN: 978-1491956656
Print: 1491956658
Kindle: B06Y3G2WKN
Audience: Azure developers
Rating: 4.5
Reviewer: Kay Ewbank

This is a good introduction to using Azure Data Lake, HDInsight and Spark, their terminology and ideas. 

I don't think that having read it you'd have mastered Azure Analytics, but you'd know a lot more about it than when you started.

The book opens with a chapter on the fundamentals of enterprise analytics - data lakes, lambda architecture, Kappa architecture, the Azure pipeline. The chapter ends with a section on what you'll need to work through the examples in the book - essentially Visual Studio 2015 and a basic Azure subscription, along with the Azure SDK.

Chapter 2 looks at getting data into Azure, followed by a chapter on storing ingested data in Azure. This chapter covers file-oriented storage (blobs, Data Lake, HDFS) versus queue-oriented storage (event hubs and the IoT hub). As with the rest of the book, there are examples with code throughout the chapters showing how you'd go about it.

The next two chapters cover real-time processing in Azure, and real-time micro-batch processing in Azure. Both look at HDInsight and Storm, with the micro-batch chapter introducing Spark as well. Azure Machine Learning also gets an introduction.

A long chapter (54 pages) on batch processing is next, looking at all aspects including MapReduce, Hive, Pig, Spark, SQL Data Warehouse, Data Lake Analytics, and Azure Batch. Some elements get more coverage than others - Hive on HDInsight is covered quite extensively, Pig gets a page.

A chapter on interactive querying looks at Azure SQL Data Warehouse, Hive used with Tez, and at Spark SQL and USQL. Each option gets a brief introduction, a description of how partitions and distributions are dealt with in that option, and an explanation of the indexes and how they are managed. As with the rest of the book, it's an introduction, not in-depth coverage.

There's an interesting chapter on the hot and cold path serving layer in Azure, mainly looking at how Redis Cache, DocumentDB, and SQL Database can all be used to make data that's both hot and cold available for analysis.

A chapter on intelligence and machine learning briefly introduces Azure Machine Learning, R Server, SQL R Services, and Microsoft Cognitive Services, essentially covering what they do and their individual advantages. This is followed by a chapter on managing metadata in Azure, looking at the Azure Data Catalog, Data Lake Store, Storage Blobs, and SQL Data Warehouse.

 

 

A short chapter on data protection is followed by a chapter on performing analytics using Power BI. It might seem odd in a book called Mastering Azure Analytics to wait until Chapter 12 to actually look at performing analytics, but Azure is a large and complex system and until you've got to grips with everything else, you can't really expect to do much analysis.

In my view, this is a very good introduction to the different elements of Azure and how they all fit together. It won't, despite the title, make you into a 'master' of Azure, but after reading it and working through the examples you'd have a really good grip on what the different parts of Azure do and how to use them.