As we all know, since the "digital economy" was first mentioned in the government work report five years ago, it has had a very profound impact on China's economic operation, and has become the core of the current high-quality economic development.
As a key factor of production in the digital economy, the importance of data has become increasingly prominent. During this process, for industry users in various industries, with the continuous expansion and increase of business and applications, their key business systems are also continuously generating a large amount of core data with high value density. How can these data be highly secure, High-performance, high-stability operation and how to better tap the value behind it become crucial.
It is precisely because of this that distributed databases have risen strongly in the past few years. According to IDC data, currently about 26.8% of enterprise-level market users have deployed distributed databases, and more than 90% of enterprises recognize the effect of distributed database deployment. Among them, about 66% of the interviewed enterprises have seen significant improvement in the performance of the database system, effectively solving the pain points and difficulties of database enterprise-level applications.
But we must also see that in the case of continuous and rapid growth of data volume, enterprises have also put forward higher requirements for distributed databases, such as more and more data storage, and higher and higher performance of databases; During the migration process, new goals are also put forward for the compatibility of distributed databases. Similarly, security compliance puts forward new requirements for the security of distributed database libraries; not only that, but the diversification of application scenarios also requires distributed databases to have stronger Adaptability and more.
On August 10, at the 2022 OceanBase annual conference, OceanBase 4.0, code-named "Xiaoyu", was officially released. This is the industry's first single-machine distributed integrated database, which realizes single-machine deployment and takes into account the distributed architecture. The scalability and performance advantages of the centralized architecture not only broke through the bottleneck of the stand-alone performance of the distributed database, but also achieved a historical "leapfrog" in the industry in which the performance of the stand-alone surpassed the centralized database.
More importantly, it also has lower deployment costs and operation and maintenance complexity, flexibly meets the needs of different usage scenarios, and greatly reduces the threshold for small and medium-sized enterprises to use distributed databases, allowing distributed databases to be implemented in thousands of industries. "Inclusive" has laid a solid foundation.
As OceanBase CEO Yang Bing said: "China's basic software industry has been able to 'root innovation', insist on long-term investment, China's database will usher in a big explosion. As a completely self-developed enterprise-level native distributed database, we hope that Through continuous technological innovation, we will lower the threshold for using enterprise databases for different types of customers and help industrial digitization."
Distributed database ≠ distributed system
Since 1970, when EFCodd, the father of relational database, first proposed the concept of relational model in his paper, the history of database has been flowing for 52 years. For a long period of time, the business model and ecosystem built by centralized databases in the information age have made it difficult to shake their market position.
However, in the digital age, with the explosive growth of data generated by various industries, the limitations of centralized databases are gradually revealed in the face of large-scale data processing, which also gave birth to the emergence of distributed databases. In the "Distributed Database Development Trend Research Report" recently released by the Information Security Development Research Center and other institutions, we have seen several key reasons for the strong rise of China's distributed database:
First, from the background of industrial development, we can see that with the emergence of new scenarios and new applications such as mobile payment and big data, the resulting growth in volume and changes in data form require continuous improvement in the processing capabilities of the underlying database. Iterative upgrades, and after the country identified data as a key factor of production, a series of policy drives have allowed the scale of China's distributed database industry to continue to grow.
Among them, in the "14th Five-Year Plan" software and information technology service industry development plan, it is proposed to "accelerate the research and development and application promotion of distributed distributed databases, hybrid transaction analysis and processing databases, and shared memory database clusters." In addition, in In the "National Integrated Big Data Center Collaborative Innovation System Computing Power Hub Implementation Plan", it is also clearly required to "increase the large-scale application of software and hardware products such as server chips, operating systems, databases, distributed computing and storage, and data circulation models." ".
Second, from the perspective of scenario-driven innovation, it is mainly reflected in the fact that due to the huge population base in China, key industries including finance and telecommunications that are related to the "national economy and people's livelihood" are all in the world in terms of the scale of their key applications and the volume of their data. As a result, these industries are facing "world-class" data processing pressure, and the performance of traditional databases often cannot meet the processing needs of this scale, so more and more industry users have moved from centralized databases to distributed databases. database.
Third, from the perspective of technological evolution, the distributed database itself is also continuously iterating and upgrading. For example, the original "distributed middleware + stand-alone database" was mainly transformed on the stand-alone database system to solve the problem of database scalability. , but it also faces "short boards" such as global transaction capabilities, high availability, and the inability of nodes to expand on demand; the "asymmetric computing node + distributed storage" that has emerged since then solves the data problem through a distributed storage engine. The problem of consistency has achieved high availability of writing capabilities, but it also has its own disadvantages, including limited scalability and the need for sub-database processing. At the same time, it is difficult to achieve high availability across regions, and the underlying data center infrastructure There is a strong dependence and so on.
Simply understand, the "distributed system" built on the basis of distributed technology + stitching system is fundamentally different from distributed databases. Whether it is the additional expenses brought about by distributed sub-tables or the cost problem of easy expansion and shrinkage, all need There is a better way to solve it.
Until the emergence of the native distributed database, it can be said that the problems in the above technical route have been solved quite well. The architecture design of the native distributed database itself is based on the underlying design of the distributed consistency protocol, so it supports better consistency, global MVCC, and high availability. Most importantly, it can also achieve unlimited horizontal expansion on ordinary servers. By adding low-cost servers, the computing power can be expanded, and the overall performance of the database cluster can be improved. Therefore, it has become a general trend for distributed databases to move towards native design today.
OceanBase, which was born in 2010, is a typical representative of native distributed databases. At the same time, OceanBase has always adhered to the "original + self-developed" model, accelerated product iteration through diversified business scenarios, and greatly reduced user costs. Make deployment more convenient, and use technology to make the management and use of massive data easier.
At present, OceanBase has helped 400+ industry customers to achieve smooth migration and upgrade of core systems, involving finance, energy, communications and other fields, which also makes OceanBase move from the initial financial industry to industries related to "national economy and people's livelihood", becoming the first choice for core system upgrades. It can be said that in the past twelve years, OceanBase has embarked on a path of innovation in China's original distributed database with continuous technological innovation and rich industry practice.
"Small is big" is the key
In this process, although native distributed databases have various innate advantages, they also face two challenges in the process of landing: On the one hand, in order to ensure absolute high availability, native distributed databases often use "three copies" Formal deployment, which means that if an enterprise wants to deploy a distributed database, it must deploy multiple servers at the same time. It is difficult to deploy "distributed".
On the other hand, in the past, large and medium-sized enterprises usually deployed native distributed database systems, while small and medium-sized enterprises did not need to deploy distributed database systems when the business volume was insufficient due to the small amount of data and relatively simple application scenarios. Instead, they purchased Stand-alone is more cost-effective. However, once a stand-alone server is deployed, structural adjustments will be made when the subsequent business volume is huge, which will further increase the difficulty of deployment.
How to solve this difficult problem of "having both the fish and the bear's paw"? OceanBase 4.0, code-named "Xiaoyu", was "emerged" for this purpose. As the industry's first single-machine distributed integrated architecture, it effectively takes into account the scalability of distributed architecture and the performance advantages of centralized architecture. It is characterized by the fact that it can not only realize single-machine deployment, but also realize the complete functions of distributed deployment under single-machine deployment, which breaks through the single-machine performance bottleneck of distributed database for the first time technically.
Yang Bing said: "Version 4.0 enables OceanBase not only to have large-scale processing capabilities, but also to be able to match the performance of traditional stand-alone databases, and to support industry customers to deploy stand-alone in small-scale scenarios. We gave it a very unusual and full of meaning-little fish, the English name is Paetica, pocket fish, is the smallest fish in the world, this version represents the exploration and innovation of OceanBase in miniaturization, and is also the distribution It is a transition from single-machine distributed to single-machine distributed, which truly allows OceanBase to expand on demand and realize innovation from 0 to 1 and then to N."
So, what are these new features of OceanBase4.0? We can observe from the following dimensions:
First of all, the deployment can achieve "miniaturization". OceanBase4.0 can run on the world's smallest computer (Raspberry Pi). available. In other words, OceanBase4.0 can adapt to businesses of all sizes, from small personal sites to bank core systems and giant e-commerce sites. After enterprises choose it, they can flexibly meet the requirements of cost performance and high availability according to their own characteristics at different stages of business development, and greatly reduce the threshold for enterprises to use distributed databases. The breakthrough of technology and the reduction of usage cost also mean that OceanBase 4.0 has broken through the boundaries of distributed databases, making distributed databases a step closer to general-purpose databases.
Secondly, the system has strong "ease of use", which is specifically reflected in that even in the stand-alone deployment mode, OceanBase4.0 can still realize the complete functions of distributed deployment, including Oracle/MySQL compatibility, TP transaction processing capability, and AP parallelism Analysis and query capabilities, tenant resource isolation, etc. In addition, compared with the previous versions, OceanBase 4.0 has greatly improved the transaction processing capability, improved the analysis and processing capability, realized the coverage of the full scene vectorization capability, and is also the first in the industry to reduce the fault recovery time (RTO) from 30 The second is optimized to 8 seconds, and the smaller the RTO value, the stronger the system failure recovery capability. This is the first time for Chinese databases and the first time for global databases, which also means that the future system failure recovery time will be real Stepping into the second-level era.
Finally, it is the "low cost" of operation and maintenance. The official release of OceanBase Cloud also allows OceanBase to lower the threshold of serving small and medium-sized enterprises with a better billing model. OceanBase Cloud has 100% of OceanBase's capabilities, allowing more small and medium-sized enterprises to have financial-grade high-performance databases on the cloud more conveniently and at a lower cost, and it only takes a few minutes to complete the database setup on the cloud.
It is precisely because of this that in the past year, OceanBase Cloud's annual revenue has increased by 300% year-on-year, which has become a strong second growth curve. It is understood that in addition to supporting Alibaba Cloud, OceanBase Cloud will also officially launch Amazon Cloud Technology (AWS) in September this year, and will continue to support the world's mainstream public cloud infrastructure in succession, which will not only enable OceanBase to better empower thousands of people in the future. In all industries, it can also achieve faster "expansion of territory and territory" in the global market.
In this regard, Yang Bing also said: "Historically, when OceanBase evolved forward, there were two major refactorings. The first major refactoring was in the process from version 0.5 to version 1.0. This major refactoring made OceanBase has become today's peer-to-peer distributed architecture, becoming a real financial-grade distributed database; and the second time is the reconstruction from 3.0 to 4.0, and this upgrade has completely made OceanBase a general-purpose database for the cloud era. .”
From this perspective, OceanBase4.0 truly interprets the concept of "small is big". It not only brings native distributed databases into a new era of "small is new big", but also realizes It not only ensures the reasonable matching of different enterprises' different needs for distributed databases, but also makes "small is big" the key key to leverage the digitalization of the industry, so that thousands of industries can better go to the digital "sea of stars".
"Four in one" to create an innovative engine
It is worth mentioning that, in addition to technically breaking through the stand-alone performance bottleneck of distributed databases for the first time, OceanBase has also released a series of new policies and measures in the fields of service, ecology, and open source. mode, releasing the technological dividend brought by the continuous innovation of OceanBase4.0.
In terms of services, in order to better serve industry customers, OceanBase launched the "Worriless 365" agent operation and maintenance service. This service is aimed at the core scenarios of customers and provides three different levels of The service availability level guarantee scheme, the highest level of service availability reaches 99.99%, that is, it can guarantee that the user's single service failure will not exceed 5 minutes per month. Let the database count, "If you dare to use it, I will pay for it."
In terms of ecology, OceanBase has comprehensively upgraded its business ecology strategy. In the future, it will build a business ecology model centered on "partners" and comprehensively deploy regional markets. To this end, OceanBase plans to launch the "Coral Project", which will target 60 core dealers in key provincial capital cities across the country in the next three years. On this basis, through policy support, OceanBase will help core distributors continue to grow, and realize that the revenue share of partners accounts for more than 60% of the total sales in three years.
In terms of open source, on June 1 last year, OceanBase officially announced its open source. After more than a year of hard work, OceanBase has connected more than 30,000 community developers, and more than 100 small partners have contributed code to OceanBase. Similarly, as the strategic direction of OceanBase, OceanBase's open source policy has also ushered in another upgrade, including OceanBase will release 4.0 community version, further open source OceanBase kernel code, that is, all open for "MySQL compatibility", in addition, the community version will enjoy enterprise version with the same performance, etc. Next, the OceanBase 4.X community edition will also move towards "miniaturization". In the future, it will support 4 cores and 8G small specifications at the minimum. While lowering the application threshold for more developers, it will also contribute more to China's open source database industry. the value of.
Looking forward to the future, Yang Bing finally emphasized: "The four-dimensional integration of products, services, ecology, and developers will be the golden key for OceanBase to continue to move forward; core system upgrades, comprehensive cloud migration, and business expansion will be the three pillars that support OceanBase's future growth. In the future, we will provide high-quality database products and services for customers of different business scales and regions. While creating value for customers, we will also achieve our own development and win-win results, and jointly promote the new process of China's industrial digitalization. "
The full text concludes that based on the stand-alone distributed integrated architecture, this "small step" taken by OceanBase in technological innovation is undoubtedly an important "big step" for the entire database industry. I believe that OceanBase will continue to innovate in the future, And the strong potential energy released by continuous investment in products, services, ecology and open source fields will definitely make OceanBase subvert itself to subvert the industry, and play a more decisive role in reconstructing the new pattern of the global database market. value and role.
Shenyao's Science and Technology Observation was founded by Shenji, a self-media person who crosses the border between technology and automobiles. With 19 years of experience in enterprise-level technology media work, he has long focused on the observation and thinking of enterprise digitalization, industrial intelligence, ICT infrastructure, and automotive technology content.
