Why learn HBase?
The entire Internet is built on the base of databases. There are many types of databases, but there are only a few databases that are indispensable for the development of each Internet business to a certain scale: relational databases represented by MySQL and its distributed The solution is the cache database represented by Redis, the retrieval database represented by ES, and the distributed persistent KV database. In the open source field, especially in China, HBase is almost the preferred solution for distributed and persistent KV databases. There are many business scenarios for HBase applications, such as user portraits, real-time (offline) recommendations, real-time risk control, social feed streams, product history orders, social chat records, monitoring systems, user behavior logs, and so on.
Why write such a book?
The author has written some HBase-related articles in the previous few years. Although each article is self-contained, from the perspective of the HBase system, there seems to be a lack of connection between these blogs. In addition, the blog still lacks many key knowledge modules, such as HBase load balancing mechanism, cluster replication function, etc. In addition, many readers have reported that they want to try to use HBase 2.x on the production line in the later stage, and want to learn more about the new features of the 2.x version. Of course, in the long run, HBase 2.x will definitely take some time in the next period of time. universal. Therefore, I hope to write a book that can first string together the contents of the HBase blog so that readers can understand the working principle of the entire HBase in one go. In addition, the missing core modules are added, more operation and maintenance tuning, case studies are added, and the most important core features of the HBase 2.x version are added at the end.
What can this book bring to readers?
For HBase, a heavyweight core database, understanding its working mechanism will not only help manage and optimize the company's HBase cluster, but also better support more business development. More importantly, it is especially important for the construction of one's own knowledge system. This book can help you establish (or improve) the core technology system from these aspects:
1. How to implement KV database based on LSM storage engine. The LSM storage engine and its variants are the basis of many database implementations, such as the common MyRocks, HBase, InfluxDB, Kudu, and so on. This book will completely analyze the core details of LSM storage engine writing, reading, and compaction based on the HBase kernel perspective. These working principles are similar to other databases based on the LSM storage engine.
2. How to make a single-machine KV database into a distributed cluster. A single-machine KV database into a distributed cluster needs to overcome many technical problems, such as how to split, merge, load balance the region, how to achieve distributed backup and recovery, how to achieve cluster replication, and how to achieve cluster failure recovery. The implementation of these core technologies will be fully explained in this book.
3. HBase cluster operation and maintenance, tuning, and fault location strategies. On the basis of introducing the core working principle of HBase, this book uses a lot of space to introduce the relevant methods of HBase cluster practice in the form of cases. I believe that through the practice of these methods, combined with theoretical knowledge, it is possible to operate and maintain large-scale HBase clusters.
In short, this book takes HBase as an example, introduces a large number of general technical implementation frameworks and details, and arranges more targeted topics. I believe that through the learning and understanding of these knowledge, readers can be more comprehensively enhanced in many aspects of technical competitiveness.
Hudi Principles | Overview of Apache Hudi Core Concepts