Three years to go: How far have we come in achieving 100% replacement of domestic databases?
Latest update time：2024-10-19
Reads：
Author | Zhu Xin
Source | First New Voice
In the 1980s,
China's database industry began to take shape;
in the 1990s,
IBM and Oracle monopolized the domestic database market;
at the beginning of this century,
the "Four Golden Flowers" were established one after another;
in
the 10s,
Internet companies started the "de-O" trend;
by the 20s,
domestic databases had already shown a "hundred flowers blooming" trend.
Currently,
domestic databases are still catching up and surpassing.
According to a document released by the State-owned Assets Supervision and Administration Commission in 2022, by 2027, "2+8+N" will achieve 100% domestic substitution of databases.
Now that 2024 is drawing to a close, the industry has different attitudes towards the slogan of de-IOE and replacement schedule for domestic databases. Some companies are already sharing their core system replacement experiences, while others still believe that replacement is too difficult and dare not try it easily...
What are the current capabilities of domestic databases? Do they have the ability to compete in the market? As the countdown to domestic database replacement approaches, what are the progress and results of replacement in various industries?
After a series of surveys, interviews and data collation, First New Voice has figured out the replacement industry portrait of the domestic database industry.
01
The initial development of domestic databases:
From breaking the monopoly to the "Hundred Regiments Campaign"
The development of domestic databases has roughly gone through several key development periods:
The development history of China's domestic database
The embryonic stage of domestic technology in the 1980s:
In the 1980s, Professor Sa Shixuan and Professor Wang Shan
opened the door to China's database field
and cultivated the first generation of Chinese database talents.
The monopoly period of foreign manufacturers in the 1990s:
In 1989, Oracle entered China; in 1991, Sybase entered China; in 1992, IBM entered China; DB2 and Informix took over the financial industry market; in 1997, Oracle took over the telecommunications market.
During this period,
foreign databases almost monopolized the Chinese database market,
accounting for more than 80% of the Chinese database market. However, China also had the first generation of prototype databases, such as Neusoft's Openbase, Chinasoft's Cobase, and Huake's DM Database.
The embryonic stage of domestic manufacturers in the 2000s:
Based on the 863, Core High-tech and 973 programs, a number of domestic manufacturers with university backgrounds such as Renmin University of China Jinkang, Wuhan DAMO, Nanda General, and Shenzhou General were established.
After entering the 21st century,
the national 863 plan established a [major database project]. With the support of national policies, the "four golden flowers of domestic databases" Renmin University Jincang (1999), DAMO Database (2000), Nanda General (2004), and Shenzhou General (2008) began to develop.
However, in the original traditional relational database field, Oracle and IBM had too great a first-mover advantage. The environment at that time required economic development rather than independent control, so domestic databases entered a vicious circle. Without a market, it was impossible to verify whether the database was reliable. If it was impossible to verify whether the database was reliable, no company would dare to use it, and there would be no market.
Rapid development period from 2010 to present:
With the rise of the cloud computing era and the open source community, domestic databases began to overtake others.
Everything needs an opportunity, and external forces are also driving it all. In 2014, the "Prism" incident broke out, and the industry deeply realized the urgency of information security and the necessity of independent control. Domestic databases such as the "Four Golden Flowers" have received widespread attention. With financing for financing and projects for projects, domestic databases are beginning to usher in a golden period of development. In the past, they all relied on the line of industry, academia and research. Since Alibaba shouted the slogan of "de-IOE", Internet and IT giants such as Tencent, Baidu, and Huawei have also participated in the development of domestic databases, and domestic databases have begun their own market practice. At the same time, a series of excellent databases and database companies were born, such as TiDB, OceanBase, etc., which also created the grand occasion of the "Hundred Regiments War" of domestic databases in 2020.
According to data, the scale of my country's database software market was 8.537 billion yuan in 2015 and increased to 12.00 billion yuan in 2017, of which the market scale of domestic database products was 1.715 billion yuan and the market scale of foreign products was 10.307 billion yuan. The domestic market share of domestic database products increased from 4.0% in 2009 to 14.26% in 2017.
Under the wave of "de-O", Oracle's monopoly position has been impacted. In May 2019, it announced a layoff of more than 500 employees in China. This also marked the initial formation of the "domestic advance and foreign retreat" situation in which domestic databases were replaced.
Deepening stage in 2022:
According to Document No. 79 of the State-owned Assets Supervision and Administration Commission in September 2022, by 2027, the "2+8+N" party and government and eight major industries will complete 100% domestic substitution, and the replacement scope will cover chips, basic software, operating systems, middleware and other fields.
Under the catalysis of the information and innovation policy, technologies such as distributed databases and cloud deployment have continued to mature, the development window for domestic databases has been opened, and the historical process of alternating between the old and the new has been accelerated.
02
The overall replacement rate in the financial sector is not high
The "2+8+N" in the SASAC's Document No. 79, "2" refers to the party and government, which are the primary areas for the development of the ICT industry; "8" refers to the eight major industries related to the national economy and people's livelihood, such as finance, electricity, telecommunications, petroleum, transportation, education, medical care, aerospace, etc., which are also key industries for the development of the ICT industry; and "N" refers to the consumer market and various enterprise-level markets, which are potential markets and growth points for the future development of the ICT industry.
According to a survey by the First New Voice, the current domestic substitution rate for databases in the party and government fields is as high as 80%, and is basically in the final stage of replacement.
After the successful trial of domestic database replacement in the party and government fields, it began to gradually and steadily penetrate into the eight major industries.
Whoever controls finance controls the world.
The use of databases in the financial industry has always been a barometer of the development of the database industry. In 1998, IBM DB2 and Informix were put into commercial use in the core financial system, laying the foundation for the pattern of large and small computers in the core system for the next 20 years. Today, the financial industry has become the pioneer in the migration of large and small computers to the lower levels.
There is a saying that has been circulating in the field of domestic database replacement: "Finance crosses the river by feeling the bank, and the bank crosses the river by feeling the stones."
The painless migration of domestic databases in the banking sector has become the key to the successful replacement of domestic databases in the financial industry and even the entire industry.
As basic software, databases require decades of accumulation, and major financial industries have begun to replace their businesses to varying degrees in different scenarios.
Ten years ago, domestic database replacement had already been deployed in non-traditional banking financial scenarios. For example, the domestic database OceanBase successfully took over 10% of Alipay's business traffic during the 2014 Double Eleven Shopping Festival, and migrated all of Alipay's core data to OceanBase in 2016, fully supporting all of Alipay's core business links and five major business segments; following Alipay, the world's first cloud bank, MyBank, which was established in 2015, fully adopted the OceanBase database in 2017, becoming the first institution in the banking industry to achieve 100% de-IOE and self-control. In 2015, the core system of WeBank was officially launched, and TDSQL also carried hundreds of WeBank's core systems and all of the bank's OLTP businesses, becoming the first domestic database used in the core system of a bank.
In addition to pilot projects in financial scenarios of non-traditional banks, the core systems of small financial enterprises are also an area of trial for domestic databases.
In 2019, Zhangjiagang Rural Commercial Bank put its new core system using a domestic database into operation and ran efficiently and stably, becoming the first case in which a distributed domestic database was successfully used to replace a foreign database in the core business system of a traditional commercial bank in my country. In May 2019, Huawei's GaussDB OLTP database was successfully put into operation in the comprehensive payment transaction system of China Merchants Bank and the Industrial and Commercial Bank of China; in 2019, China CITIC Bank adopted the distributed database GoldenDB, which is the first domestic distributed database successfully applied in the core credit card system of a large joint-stock bank in China; Hubei Bank and Meizhou Merchants Bank adopted the bank core solution built by DAMO database in 2019 and 2020 respectively, realizing the full takeover of core business.
More and more successful cases of core banking system replacement have greatly enhanced the industry's recognition of domestic databases and accelerated the replacement process.
The release of national policies at this time is even more powerful. The country has announced the goal of achieving domestic substitution of financial core systems in three years: after three years of financial information innovation pilot from 2019 to 2021, financial information innovation will enter the stage of full promotion in 2022. The state clearly requires the domestic substitution of peripheral systems such as OA in the first year, the domestic substitution of business systems in the second year, and the domestic substitution of core systems in the third year. At the same time, the first batch of localization pilot financial institutions were announced, including major state-owned banks, head securities firms, exchanges, etc., totaling 47. It was required that 5% of new IT purchases in that year be localized, and 15% in the second year. After that, two more batches of institutions entered the pilot, totaling nearly 150.
The timetable has been set, and major banks have begun to steadily advance database migration.
According to incomplete statistics, 46% of the purchasers of domestic databases in 2022 are concentrated in the financial sector, followed by the government, accounting for 18%. This trend has also continued into 2023. In the first half of 2024, more than 130 database projects have been implemented, and the financial industry has concentrated nearly 60% of the projects, followed by operators, accounting for more than 15%.
However, according to the survey and incomplete statistics of the First New Voice, as of now, the eight major industries have higher requirements for the stability, migration difficulty, operation and maintenance difficulty of domestic databases, so the overall replacement rate is still not high. The non-core systems in the financial industry are around 40%. But even with the support of policies and the market, the replacement of domestic databases in my country's overall banking industry is about 20%, and the proportion of foreign databases in the core systems of banks is still over 80%. The energy industry is less than 15%, and many industries such as medical care, manufacturing, and education are even less than 5%. This is still a long way from "2027'2+8+N' party and government and eight major industries to complete 100% domestic substitution".
03
Well-received but not popular:
Database migration is difficult and costly.
Being well received but not popular is the embarrassing situation currently faced by domestic databases.
Especially in the financial sector, domestic database replacement is still attempted at the office system and business system level, while there is still some hesitation about the localization of the database of the core transaction system. After all, the financial industry has always advocated "lossless operation and maintenance", and the risks of business architecture and code adjustment and smooth migration of different databases have made many large financial institutions dare not take too big steps in replacing core systems.
First, at the data level, data consistency, data security, and code security are the most important considerations.
Especially in industries such as finance and government, data security always comes first.
Secondly, at the functional level, it includes compatibility and migration capabilities, transaction processing capabilities, and big data real-time processing capabilities.
The factor that enterprises are most worried about is compatibility. Because after changing the database, they need to worry about the adaptability of the server, chip, and operating system, and also consider the compatibility of OA, ERP and other application systems.
Finally, in terms of effect, it includes stability, reliability and cost performance. Among them, stability is also the factor that manufacturers and institutions care most about when purchasing.
Facing the market demand, domestic database vendors are of varying quality. Looking at the entire database replacement, the transformation and upgrading from traditional centralized databases to domestic databases still faces a series of challenges such as open source data risks, application compatibility, system availability, performance and scalability, and migration costs.
First, database migration is difficult and costly.
This is the first problem in replacing domestic databases. The focus is mainly on compatibility, data security, downtime, data verification and performance assurance during the migration process; the lack of one-stop management of the database, the high complexity of operation and maintenance and backup are also a major problem in database maintenance after migration. The migration cost also includes compatibility issues. Generally speaking, it takes 2 to 3 years for an enterprise to complete the domestic replacement of the database. During this process, it is necessary to evaluate the difficulty of the transformation. Once compatibility problems occur, the losses are significant.
The second is the risk of open source databases.
Open source databases are widely used in the banking industry. According to a survey by the First New Voice, among the domestic databases currently used in China's banking industry, closed source databases account for only 25.6%. Most domestic databases are modified on the basis of existing open sources. Domestic databases are often criticized for their low degree of R&D, patents and code autonomy. However, open source databases have potential risks such as different open source protocol requirements and restrictions, security vulnerabilities, infringement infections, service suspension, policy risks, and insufficient control capabilities.
Third, the domestic database ecosystem is objectively imperfect.
When banks introduce domestic databases, the entire ecosystem is involved, including management, monitoring, backup, optimization, etc., which require the joint promotion of a series of upstream and downstream partners. Currently, there is no de facto standard for banking databases and no leading companies have emerged. The technology selection of each company has not been determined, and ISVs can only be based on Oracle standards or MySQL standards.
Fourth, the performance and functions of domestic databases are not perfect.
Compared with traditional commercial databases such as Oracle, domestic databases are not good enough in terms of high availability and high performance. For example, there is a certain performance pressure during business peak periods, and the ability to discover and repair vulnerabilities in complex situations is also unsatisfactory.
Localization is an important thing for the entire Chinese technology industry, but the entire market is largely forced to mature due to policies. When policies and markets are accelerating, the atmosphere of domestic replacement is full, but the products are not powerful, and the replacement process is still worrying.
04
The financial sector has entered deep waters
It doesn’t matter how slowly you walk, the key is to move forward steadily.
At present, driven by favorable policies and market forces, domestic databases are still taking an active and cautious attitude in various fields and adjusting the replacement rhythm of different industries.
After extensive research, First New Voice summarized two major characteristics of domestic database replacement:
First, the database market has shifted its focus from the provincial and ministerial levels to prefecture-
level cities, districts and counties. The replacement of internal office and electronic documents in the party and government fields at the provincial and ministerial levels is nearing completion, and is now being widely transferred to prefecture-level cities, districts and counties, with local government procurement orders gradually increasing.
Second, in the eight major industries, databases are making breakthroughs from non-core and sub-core systems to core systems.
Unlike the party and government, the eight major industries pay more attention to the stability of domestic databases. In order to prevent the emergence of core business risks, domestic replacements in the past few years have mainly focused on non-core systems. Taking the financial industry as an example, the domestic database replacement rate of the core system of banks is about 15%; the domestic database replacement rate of the core system of securities and insurance does not exceed 20%. Related fields will become the focus of domestic replacement in the next three years.
Both capital and enterprises believe that localization is an excellent opportunity for Chinese databases. According to research firm Gartner, by 2025, only about 30% and 50% of China's analytical databases and transactional databases will come from overseas manufacturers, respectively. The demand for localization is like wind and waves, leading domestic databases into the unknown.
Replacing the old king with a new one is not something that can be done in a day.
"Completing 100% domestic substitution by 2027" is not the end, but a milestone in the development of domestic databases.
As Luo Yun, deputy general manager of Tencent Cloud Database, said, "The road to localization is still difficult and arduous. How domestic databases can fully follow up in terms of product services and ecology requires the joint cooperation of Party A and Party B, and patiently deal with the ups and downs and challenges in the process to ensure the steady progress of the localization strategy."
Note: The cover image is from the Douban movie "Interstellar" stills
-END-
京公网安备 11010802033920号