Shulou(Shulou.com)05/31 Report--

Editor to share with you how the database system should achieve load balancing, I hope you will gain something after reading this article, let's discuss it together!

How to balance the load of database system

Usually no one is doing the Load Balance of the database, even if they do, they just think of it as a hot switch.

If the amount of data is too large for a database to afford, when considering dismantling Database, the first thing to consider is how to deal with data synchronization.

Before doing multiple databases in parallel, we should first determine how high the data consistency needs. If you can tolerate synchronization with time lag, you can consider using the database of Big Table architecture for processing, otherwise it is to speed up the fetch, and try to disperse the task of reading / writing to the database.

In theory, a reasonable approach should be to form a Database Cluster. In a Cluster environment, there can be many Database hosts, but everyone's backend is connected to the same external storage (usually SAN), and all hosts are written to a database in the same memory, and because there is only one storage and one database, there is no need for synchronization between hosts.

How to balance the load of database system

Load balancing equipment can not solve the problem of database overload, but there are many reasons for the insufficient performance of Databse Server. We should explore in detail why the performance is insufficient. Database Cluster can solve some of the problems, but it may not necessarily bring much performance improvement.

Dismantling the Table structure is a method, which is usually recommended for Table with a large amount of data, but in this way, program developers must be very painful. if you really want to adopt this architecture, it is recommended that the program architecture should have an extra layer of data access layer. Commercial objects cannot directly access database data under SQL, but access database data through data access layer components in order to avoid human error by program engineers.

The recommendation is to analyze the database performance bottleneck before determining the architecture. As a rule of thumb, Disk Index O is the biggest problem, and the reason for it is usually that the Index is not set properly, or the SQL instructions written by programmers do not take into account the performance problems after data growth. After these problems are eliminated, consider the backup or mirroring of the database, and point the complex query instructions, BI reports and Data Mining data sources to the Database host that is backed up or mirrored, which should solve your problem.

After reading this article, I believe you have a certain understanding of "how the database system should achieve load balancing". If you want to know more about it, you are welcome to follow the industry information channel. Thank you for reading!

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.