从Kudu开始对快速数据执行快速分析

Getting Started with Kudu: Perform Fast Analytics on Fast Data
从Kudu开始对快速数据执行快速分析

Jean-Marc Spaggiari, Mladen Kovacevic, Brock Noland, and Ryan Bosshart
让-马克·斯帕加里、姆拉登·科瓦切维奇、布洛克·达拉和瑞安·博哈特

Dedication
献词

To all those data people who day-in and day-out lead hair-pulling, brain-teasing, late-night lives architecting, developing, or consulting on software that appears to have gone rogue and deliberately misbehaves.
致所有这些数据人，日复一日的抓头发、虐大脑、深夜设计，在软件上进行架构设计、开发或咨询，而这些软件看起来已经变得不正常并故意表现不良。

To our families, who may not even care about technology, yet still allowed us to give up time and energy to dedicate to this project with an enormous amount of patience and support, without which none of this was possible. We love you!
对我们的家庭，他们可能甚至不关心技术，但仍然允许我们放弃时间和精力，以极大的耐心和支持致力于这个项目，没有他们，这一切都是不可能的。我们爱你！

Preface
前言

Choosing a storage engine is one of the most important decisions anyone embarking on a big data project makes and is one of the most expensive to change. Apache Kudu is an entirely new storage manager for the Hadoop ecosystem. Its flexibility makes applications faster to build and easier to maintain. As a Hadoop developer, Kudu is a critical skill in your big data toolbox. It addresses common problems in big data that are difficult or impossible to implement on current generation Hadoop storage technologies.
选择存储引擎是任何着手大数据项目的人做出的最重要的决定之一，也是最昂贵的改变之一。Apache Kudu是Hadoop生态系统的全新存储管理器。它的灵活性使应用程序更容易构建和维护。作为一名Hadoop开发人员，Kudu是大数据工具箱中的一项关键技能。它解决了大数据中的常见问题，这些问题在当前一代Hadoop存储技术上很难或不可能实现。

In this book, you will learn key concepts of Kudu’s design and how to architect applications against it, resulting in Kudu applications that are fast, scalable, and reliable. Through hands-on examples, you will learn how Kudu integrates with
other Hadoop ecosystem components like Apache Spark, SparkSQL, and Impala.
在这本书里，你将学习Kudu设计的关键概念，以及如何根据它设计应用程序，从而使Kudu应用程序快速、可伸缩、可靠。通过实际操作的例子，您将了解到Kudu如何与其他Hadoop生态系统组件集成，如Apache Spark、SparkSQL和Impala。

This book assumes some limited experience with Hadoop ecosystem components like HDFS, Hive, Spark, or Impala. Basic programming experience using Java and/or Scala, experience with SQL and traditional RDBMS systems, and familiarity with the Linux shell is also assumed.
这本书假定您具备一些关于Hadoop生态系统组件的有限经验，如HDFS、Hive、Spark或Impala。还需要具备使用Java和/或Scala的基本编程经验、使用SQL和传统关系数据库管理系统的经验，以及熟悉Linux shell。