With the acceleration of digital transformation, enterprises are increasingly demanding data processing and storage. Professional and efficient data management solutions are the key to each enterprise's digital transformation. According to the annual analysis report of China's database industry in 2023 released by motianlun, a mainstream database community in China, the separation architecture of Data Inventory calculation will become an important trend in the development of enterprise storage, provide enterprises with more efficient and flexible data management solutions.

Database has always been the core system supporting bank business processing. In recent years, under the call of the country, the financial industry has accelerated and pushed up the upsurge of database localization transformation. The whole transformation process has realized the process of domestic databases from scratch and from scratch to basically forming a scheme system. In the early stage of database localization transformation, most enterprises adopted the distributed data inventory integration architecture, which emerged from the Internet industry. However, with the continuous expansion of business needs, the mobile Internet is moving towards the era of digital intelligence. The business volume and data scale are increasing rapidly, and the limitations of the integrated storage and computing architecture are becoming increasingly prominent.

It can be observed that some major domestic banks are mainly facing the following problems when using the distributed data inventory integration architecture. First of all, it is difficult to manage and maintain. As business grows, the complexity of distributed database transactions increases, which makes it more difficult to design, develop, manage, and maintain applications and increases resource investment. Secondly, the local disk failure rate of the server is relatively high, which is prone to hard disk sub-health problems and causes business interruption. Enterprises can only solve this problem through multi-copy redundancy and additional data replication, however, this solution will result in a double increase in costs and data center space. In addition, when a fault occurs, primary/secondary failover is required. In a distributed architecture, such a fault is prone to infection, which will cause point-to-surface faults and affect more services, in addition, it is difficult to achieve both performance and reliability in active/standby switching.

Faced with the above problems, the industry's transformation forerunners and head products all begin to explore in the direction of storage and computing separation architecture, for example, by building highly reliable professional storage to avoid local disk failures, reduce the failure rate. Data synchronization at the storage level improves the overall reliability and reduces the overhead at the software layer. In addition, software and hardware can give full play to their respective strengths and give basic capabilities, reliability, disaster recovery, etc. to hardware implementation. Software pays more attention to applications and databases themselves, further improve the overall capabilities of the database system and fundamentally solve the problems encountered in the distributed data inventory integration architecture.

Judging from the current development trend of the entire database industry, the trend of separation of storage and computing is unstoppable. After the transformation of the separation of storage and computing, industry practitioners have begun to experience the benefits of the separation of storage and computing architecture, and more peer enterprises have also begun to follow up. At present, it has been noticed that at the financial industry database technology conference in October 2023, the Financial Information Research Institute took the lead and joined hands with Alibaba, Aoxing bass, Dameng, Nanda GM, Huawei, Tianyi Cloud, ten database manufacturers, including Wanli database, Yunhe enmo, Youxuan database and Yashan technology, jointly launched the financial multi-master database application action plan, on the basis of separation of storage and calculation, we are moving towards a higher level. In addition, the plan achieved the first stage results in April, 2024. Under the leadership of the Institute of Financial Information Technology, it released the best practices for the separation of storage and computing of three manufacturers, namely, Austar bass, Tencent Cloud and Huawei, it has attracted widespread attention in the industry.

In general, as a new trend in the database field, the storage-Computing separation architecture is helping enterprises achieve efficient data processing and management. This is an important development direction for database manufacturers. Providing solutions based on the separation architecture of storage and computing will enable them to have stronger market competitiveness. For enterprises, the separation of storage and computing architecture can improve data processing performance, resource utilization and flexibility, provide support for enterprise digital transformation, and take the lead in the digital era.