Original address: http://digi.163.com/17/1114/06/D36B05G6001680N8.html
Distributed databases have suddenly become a hot technology this year. What are the environmental changes behind this that have led to the evolution of this technology, and what are the advantages of distributed databases? What is the current status of this distributed database in China?
Database evolution direction
Just like the birth of relational databases in the 1970s, the birth of distributed databases is also based on the background of more and more data and more and more importance of data.
Database evolution history
A database is an electronic filing cabinet that is used to store electronic files. In the context of cloud computing and big data, how to help users better perform operations such as adding, intercepting, updating, and deleting data in files? Databases are a problem faced by IT companies.
The emergence of distributed databases has solved the aforementioned problems to a certain extent. The initial appearance of distributed databases was around 2005, when cloud computing was just emerging. With the rise of the new IT ecosystem, the database has also embarked on a path of evolution to distributed.
Analysis of Distributed Database Advantages
If you don’t know enough about distributed databases, consider that Alibaba’s e-commerce platform adopted this distributed database in the just-concluded Double Eleven.
Judging from the data released yesterday, the total turnover of Double Eleven exceeded 168.2 billion yuan. On that day, the OceanBase distributed database independently developed by Alibaba completed the creation of 256,000 orders and 148,000 payments per second. The advantages of physical separation and logical unity are realized by using distributed databases.
In addition to the characteristics of physical separation and logical unity, distributed databases also have the characteristics of site autonomy. Such basic features enable distributed databases to have data independence and location transparency, and local application response speed upgrades.
Specifically, we all know that the GMV data of Double 11 has been increasing in recent years, and the IT infrastructure of the transaction background has also been increasing. A single database server will eventually not meet the needs of use. The distributed database can well meet the elastic requirements required by the e-commerce platform, and can realize vertical screen expansion, further distribute data by adding multiple servers, and share data processing tasks.
There is no doubt that distributed database systems (including distributed database management systems and distributed databases) are developed on the basis of centralized database systems. Although they are very popular now, they still face some problems.
The domestic open source distributed database is still under fire
As mentioned earlier, OceanBase is a domestic distributed database independently developed by Alibaba. There are also companies such as Baidu and Huawei that also conduct distributed database research and development in China. However, due to the late start of domestic database research and development, there is still a lack of strong technical accumulation, and there is still a lack of enthusiasm.
But it is worth affirming that domestic open source distributed SQL databases such as TiDB have emerged. This more orthodox implementation of Spanner and F1 (the first two are Google distributed databases) supports both traditional RDBMS and NoSQL, but currently It is still in the open source research stage.
In the context of the Internet tide, China's distributed database appeared, which also filled the gap of domestic database. Now, distributed databases are on fire, and this will also develop in the future, and the database will also achieve a "cloud-based" database form.