Giant Sequoia Database: The Future Direction of Financial-Level Databases

introduction

In recent years, the annual investment in global financial technology has exceeded 50 billion US dollars, and China's financial technology development is leading the world trend. Today, with the continuous development of financial technology, the development of China's financial Internet and retailing is becoming more and more intense, making the organic combination of my country's financial business and technology application model attracting worldwide attention.

 

Corresponding to the rapid development of business models and innovations, the modern financial system urgently needs the innovation of technical architecture to meet the growing business needs. This includes business system agility, risk control, cost control, performance and elastic growth capabilities corresponding to business development. In today's information age, databases carry the core data of financial enterprises. As the hub of the new technology architecture, the financial-grade database is the foundation for the transformation and innovation of the modern financial system.

 

 

In 2017, in the " Database Report 2017 " by Gartner, an internationally renowned technology and market analysis agency , three Chinese database vendors were included in the report for the first time in history, namely Alibaba Cloud Database, SequoiaDB and NTU General Gbase Database. . The three databases are completely self-developed by Chinese teams. Although the application scenarios are not identical, they all have good application cases in the financial industry, which can be said to meet the requirements of "financial level". This also marks that in the field of database, the power of Chinese manufacturers has gradually risen, marking that in the field of basic software outside the open source field, Chinese manufacturers have gradually entered the world-renowned stage.

 

 

Also at the end of 2017, Gartner and Jushan Database jointly released the report "Future Development Direction of Financial-Grade Databases", which expounds the future needs of the financial industry for databases and the future development direction of financial-grade databases.

 

In this article we'll also look at the future direction of "financial-grade" databases .

 

 

Financial-grade database development

For a long time, the financial-grade database market has been monopolized by traditional relational databases such as Oracle, IBM DB2, and Microsoft SQL Server. After decades of development and iteration, traditional relational databases have been unable to meet the needs of new financial technology development. Therefore, replacing the traditional architecture with a new distributed database has become the mainstream trend in the financial-grade database market.

 

In the past few decades, limited by the storage and computing capabilities of traditional databases, data between different business departments in an enterprise is often stored in an independent manner. With the continuous development of new financial technology businesses, cross-departmental and cross-business data access has become the core requirement of enterprises. However, the data stored independently by each business system often forms "data islands", which makes the internal data management of enterprises face great challenges.

 

The emergence of new distributed databases aims to break the traditional data management system, manage and maintain cross-business and multi-type data in a unified manner, and integrate various departments and business lines within the enterprise from the data level.

 

To achieve this goal, the new financial-grade database needs to redefine the traditional database architecture from the perspectives of distributed architecture, multi-modal data management, standardized data access, data reliability, and mixed workloads.

 

 

Distributed Architecture

With the Internetization of financial technology, the traditional database architecture has been unable to carry the explosive growth of massive data. At the same time, the large-scale introduction of Internet channels has created new demands on the concurrency and performance of databases for fintech applications.

 

Since the single-point architecture of traditional databases cannot meet the data volume and concurrency requirements of new financial technology applications, the new generation of financial-grade databases must adopt a distributed architecture to meet such challenges.

 

In the traditional database architecture, enterprises must improve the storage and processing capabilities of the database by continuously enhancing the processing performance of a single hardware device. However, in today's information explosion, the improvement of hardware performance is far behind the growth of data volume. Therefore, the new database adopts a distributed architecture to store massive data evenly in multiple physical devices to avoid bottlenecks caused by a single device.

 

At the same time, the flexible expansion capability of distributed databases provides elastic capacity and performance support for financial business growth, and has obvious technical advantages in large-scale data applications.

 

In addition, using a PC server or cloud environment, the new distributed database can effectively reduce TCO and improve development and operation and maintenance efficiency under the premise of ensuring safety and reliability.

 

 

Multimodal Data Management

Today, under the trend of "Internetization" and "retailization" of financial business, financial institutions are beginning to provide users with more personalized and customized products and services. At the same time, with the increasing complexity of each business system, the correlation between the systems is also increasing. Therefore, the application system puts forward new standards and requirements for data storage management.

 

For a long time, traditional relational databases only support structured data storage and access capabilities of form types, but are powerless for semi-structured and unstructured data management such as hierarchical objects, pictures and images.

In order to realize the unified management and data fusion of financial business data, the new database needs to have the ability of multi-model data management and storage, so as to meet the management requirements of applications for structured, semi-structured and unstructured data.

 

Generally speaking, structured data refers to the data storage structure of form type, and typical applications include traditional businesses such as banking core transactions; while semi-structured data is obtained in scenarios such as user portraits, IoT device log collection, and application clickstream analysis. Large-scale use; unstructured data corresponds to a large number of pictures, videos, and document processing services, which are growing rapidly under the development of financial technology.

 

The multi-modal data management capability enables financial-level databases to store and manage data across departments and businesses in a unified manner, realize multi-business data fusion, and support diversified financial services.

 

 

 

Standardized data access

With the unification and integration of multi-service and multi-modal data, the ever-increasing diversity and complexity of services make data access methods face new challenges.

 

In traditional databases, SQL is almost the only way to access the database. With the development of business diversification, the proportion of unstructured and semi-structured data in fintech applications continues to increase. Therefore, in addition to providing standard SQL language support for structured data, the new distributed database also needs to provide access capabilities such as JSON and object storage management for semi-structured and unstructured data.

 

 

The standardized data access capability not only meets the needs of multi-type data management, but also effectively improves the efficiency of development and operation and maintenance for enterprises. Therefore, financial-grade databases, as the hub of new fintech architectures, need to provide standardized data access capabilities for applications.

 

data security

With the continuous improvement of the internal value of enterprises, data has become the lifeline and core assets of financial enterprises. As a database carrying key enterprise data, its security, reliability and stability have always been the core values ​​of financial-grade databases.

 

At the same time, whether in China or overseas, data security in the financial industry has become the primary requirement of regulators. For example, the security of the core system of banks has always been the focus of the my country Banking Regulatory Commission. Most bank data centers already have high availability and "three centers in two places" capabilities.

 

However, the perfect realization of high availability and disaster tolerance in a distributed architecture faces many technical challenges. Generally speaking, the distributed database for statistical analysis relatively weakens the function of this part, while the distributed database for online and transaction business maintains high standards and strict requirements for data security.

 

For example, data disaster recovery and active-active are the last guarantees for financial enterprise data security. Disaster recovery requires real-time data backup in multiple centers. Once a major disaster occurs in the data center, all online production services can switch centers in time to continue running. Active-active is based on disaster recovery, allowing the active and standby data centers to undertake production services at the same time, making full use of the active-active capability to improve business performance and further reducing downtime when disasters occur.

 

 

mixed load

With business diversification and data integration, different businesses have different functional requirements for data management. Due to the single data storage and access methods of traditional databases, users usually divide applications into two categories: online services and offline services.

 

Online business generally refers to end-user-oriented business systems such as bank core transaction systems. Generally speaking, such systems need to meet the characteristics of high concurrency, low latency, and high reliability. The corresponding offline business focuses on batch jobs. Generally, such services have the characteristics of high throughput, low concurrency, and high latency.

 

With the continuous development and integration of fintech businesses, the data requirements of each business line are no longer completely independent. Under this trend, financial-grade databases need to support mixed loads of online and offline services at the same time.

 

 

According to Gartner's latest definition, hybrid workload (HTAP Hybrid Transactional/Analytical Processing) not only retains the original online transaction function, but also emphasizes the ability of the database to calculate and analyze natively. Databases that support mixed loads can avoid a large amount of data interaction between online and offline databases in traditional architectures, and can also perform real-time statistical analysis on the latest business data.

 

In order to avoid resource interference between online real-time read and write and batch jobs, mixed-load databases are usually implemented using read-write separation or in-memory processing technology. Generally speaking, the multi-copy architecture of distributed databases naturally supports read-write separation technology, while databases based on traditional architectures are often implemented using in-memory processing technology.

 

 

Conclusion: About the development of financial-grade databases in China

For a long time, the financial industry has accounted for more than 50% of all enterprise-level IT investment. Under the requirements of decades of business development and strong supervision, financial institutions generally have the most stringent requirements in the industry for the security, reliability, and stability of databases. Therefore, financial-grade database products that meet the needs of the financial industry have become the benchmark in all industries.

 

At the same time, the number of users of China's commercial banks has always been at the forefront of the world. With the rapid development of China's economy and the promotion of inclusive finance, transaction banking and other businesses and policies, China's commercial banking business is also transforming to "Internet" and "retail". This enables banks to be closer to end users, interact more frequently, and have more diverse business scenarios. These new demands have accelerated the technological transformation of China's financial and banking industries, and led the world in a variety of technologies and business models.

 

On the other hand, the development and maturity cycle of database products is long. Generally speaking, a database product that is used on a large scale in the financial industry requires a long period of accumulation in technology, products, engineering, after-sales support and industry experience before it can gradually mature.

 

In addition, unlike application software, general database products as basic software should meet various business needs in various customers, not just serve a single specific scenario. This requires database manufacturers to firmly grasp the core code and development direction of the product, so that they can quickly respond to various needs of customers while ensuring a high degree of productization and standardization. At the same time, the banking, securities, insurance and other leading industries faced by financial-grade database products have extremely high requirements on the quality and stability of products, which makes users raise more questions about the complexity and maturity of financial-grade database products. high level requirements.

 

In this context, Jushan Database, as a financial-level database product independently developed by my country, has a leading technology direction, diverse application scenarios, and can be applied on a large scale in financial enterprises. Therefore, the continuous and vigorous development of the giant sequoia database has also attracted the attention and recognition of the international industry.

Guess you like

Origin http://43.154.161.224:23101/article/api/json?id=325078146&siteId=291194637