Detailed explanation of distributed database hbase

　　Brother Xinba noticed that with the development of computer technology, the amount of data stored in human beings has undergone great changes, which can be described as massive. Audio, email, etc., so in this case a new type of database represented by Google's BigTable was born and developed rapidly. HBase is the open source implementation of BigTable. Below, Xinba will reveal HBase related knowledge and related applications in detail.

　　The requirements for databases in the Internet era are still different from those of traditional ones.

　　One of more prominent points is the problem of data volume. In today's huge data volume, relational data may not be able to store and process massive data, but distributed databases can achieve these It is required to be able to process data quickly and store massive data.

　　One thing that is more realistic is that distributed databases can cope with the rapid development and change of business needs. Today, with the rapid development of information, when an Internet product is launched, it will definitely encounter an increase or decrease in demand. In this case, the relational database can no longer meet the requirements, because after the data fields of the relational database are determined, some related fields cannot be expanded, and only a new table can be created, which is very poor for the scalability of the program, but the distribution The HBase database is different. It can cope with the continuous business adjustment. You can add fields in the database at will. It has strong scalability and is also the preferred data storage method for many Internet products. Expansion is very helpful.

　　We know that a well-known system will definitely face this problem of high concurrency and high throughput. Some e-commerce websites in our lives are famous for a certain cat, and the number of users that needs to be faced on the day of Double Eleven can be imagined. It is very huge, and the general relational database will definitely collapse and cannot meet people's needs, but the distributed database hbase can handle the high concurrency and high throughput of data very well. It is also one of the databases commonly used by e-commerce websites.

　　Features of Hbase

　　Brother Xinba learned from the official website that the Hbase database runs on top of Hadoop, and designed and implemented a database system with high reliability, high performance, column storage, scalability, and real-time read and write.

　　Column-oriented is a typical feature of HBase, column-oriented storage and permission control, and columns can be retrieved independently. In Hbase, a table can have hundreds of millions of rows and millions of columns, which shows that the amount of data stored in it is huge. For empty columns, it does not occupy storage space, so it can be very sparse when designing the table.

　　Brother Xinba found that Hbase provides data storage downwards and data computing upwards. That is to say, it can not only use the storage capacity of HDFS to provide data storage for users, but also use the MapReduce model for large-scale parallel data processing. Well, the above is some introduction to the distributed database hbase introduced by Xinba. If you still want to understand the Hbase architecture and principles, please pay attention to Xinba.