1、2021China Distributed Database Market Report2021中国分布式数据库市场报告2021中国分散市場Tags: Cloud-native, multimodel, distributed, open source, application scenariosAny content provided in the report (including but not limited to data, text, charts, images, etc.) is the exclusiveand highly confidential document of

2、LeadLeo Research Institute (unless the source is otherwise indicated in thereport). Without the prior written permission of LeadLeo Research Institute, no one is allowed to copy, reproduce,disseminate, publish, quote, adapt or compile the contents of this report in any way. If any behaviour violatin

3、gthe above agreement occurs, LeadLeo Research Institute reserves the right to take legal measures and holdrelevant personnel responsible. LeadLeo Research Institute uses “LeadLeo Research Institute” or “LeadLeo” tradename or trademark in all business activities conducted by LeadLeo Research Institut

4、e. LeadLeo Research Instituteneither has other branches other than the aforementioned name nor does it authorize or employ any other thirdparty to carry out business activities on behalf of LeadLeo Research Institute.LeadLeo Research InstituteFrost & Sullivan (China)（Summary Version）China: Database

5、SeriesFrost & Sullivan Market I400-072-5588In present choice of database distributed technology route, the primary goal is to solve theproblem of data capacity expansion. The mainstream solutions are database and table middleware,native distributed, etc. Different technical routes and products have

6、their own advantages anddisadvantages.The booming development of cloud computing has prompted various IT applications to shift tothe cloud, and the unique flexibility of on-demand services and the low cost of on-demand billingor billing by configuration of cloud services are deeply matched with data

7、base users. It isparticularly important to build database services on the cloud and design a cloud-native databasewith the basic cloud first and then fully adapted to cloud characteristics.2The new generation of distributed database meets the core requirements of enterprise users withadvantageous fe

8、atures such as ease of use, scalability, fast update iteration, and relatively low costinvestment.The boom in software applications has created a multi-scenario, multi-ecology, multi-user marketenvironment needed for database technology development. GitHub expects China to become theworlds largest s

9、ource of developers by 2030. 2021 is the most active year for investment andfinancing in Chinas database track, further catalyzing the rapid growth of Chinas database market.Overview of Distributed Database IndustryThe development of distributed database technology should meet the needs of the times

10、 andthe market, and return to the rigid needs of database users. The current distributed databaseneeds to reach the level of centralized architecture products in various dimensions so as to playits performance and cost advantages in various scenarios and penetrate into various industries.In terms of

11、 architecture selection, single database, single database sub-database sub-table +distributed middleware or native distributed database, all have their most advantageousapplication scenarios. Under the trend of distributed database, enterprises should choosedistribution rationally.Development of dis

12、tributed database marketInstructionFrost & Sullivan hereby releases the annual report China Distributed Database Market Report 2021 as part ofthe China Database Series Report. The purpose of this report is to analyze the development status, productcharacteristics and technology trends of distributed

13、 database market in China, and identify the competitionsituation in the market of distributed database in China, and reflect the differentiated competitive advantagesof the leading brands in this market segment.Frost & Sullivan and LeadLeo Research Institute conducted downstream user experience surv

14、eys on coreproducts in the distributed database field. Respondents are of different sizes and in different segments in eachof its industry that includes internet, media, telecommunications, transportation, government and other fields.The performance of distributed database products in the financial

15、field is detailed in the China FinancialDistributed Database Market Report series.Trends in distributed database presented in this market report also reflect trends in the database industry as awhole. The reports final judgment on market ranking and leadership echelon are only applicable to theindus

16、try development cycle of this year.AbstractDevelopment of Distributed Database TechnologyChina: Database SeriesFrost & Sullivan Market I400-072-5588Classification of databasesDefinition and classification of databases3As the infrastructure of most information systems, databases exert hardwarecomputi

17、ng power downward and enable upper-layer applications upward.Various database products meet different business needs respectively. Thespeed, ease of use, stability, scalability, and cost of a database are all critical toan enterprises basic business and growth resilience. Definition and classificati

18、on of databaseDatabase System: basic software that organizes, stores, and manages data according to a specific data structure.Distributed Database: a logically unified database formed by connecting physicallydispersed database units through network.This report focuses on the cutting edge of the data

19、base industry from a distributedarchitecture perspective. The importance of databasesAs the infrastructure of most information systems, database exert hardware computing power and enable upper-layer applications. Various database products meet different business needs respectively. The speed, ease o

20、f use, stability, scalability, and cost of a database are all critical to an enterprises basic business and growth resilience.If databases never existed, programmers would have to deal with massive data and unreliable computer systems. But on the basis of database, programmers do not need to redesig

21、n complex system processes to ensure transactional data processing. Instead, they only need simple operations like C-R-U-D, which greatly reduce the complexity of data storage and processing.Right after the birth of database, applications development exploded and became an important phrase in the hi

22、story of computing.Source: CCF, Frost&Sullivan, LeadleoDocument databaseRelationalNon-relationalOn-premiseCloudData ModelDeployment modelArchitectureStand-aloneCentralizedDistributedKey-value databaseGraph databaseWide column databaseTime series databaseSpatial databaseOne master, multiple SlavesOne

23、 Writes, multiple readMultiple write, multiple readShared-nothing distributed databaseShared-storage Distributed databaseSharding + MiddlewareWorkloadCharacter-isticsHybrid Transaction and Analytical Process(HTAP)On-Line Analytical Processing(OLAP)On-Line Transaction Processing(OLTP)Cloud Native Dat

24、abase(Serverless/DBaaS, flexible multi-tenancy, charge on demand)Cloud Service Database(Enterprise feature database, free operation and maintenance, charge by subscription)Cloud Hosted Database(Infrastructure-as-a-service, single-node cloud deployment, charging by configuration)w/o shardingNative di

25、stributedComputer Science PanoramaTheoryApplicationDatabaseCompilerOperating SystemHardwareInternetSecurity and PrivacySource: Tsinghua university, LeadleoChina: Database SeriesFrost & Sullivan Market I400-072-5588Distributed database concept and technology development evolution4Since 1970, the data

26、base field has been led by the academic and research end. Itentered the stage of rapid development of commercialization after 1995, andcontinued to drive the development of application ecology. After 2015, thedatabase has developed to the direction of distributed architecture and cloud inthe develop

27、ment of Internet infrastructure. The evolution of databaseDatabase development has gone through half a century, experienced academic driven, commercialization landing,paper industry realization, enterprise application demand driven and other technical development stages.From the beginning of the hie

28、rarchicial model, mesh model, relational model, to object model, object-relationalmodel, semi-structured, etc., the data model has been the core and theoretical foundation of the database. Solidtheoretical support and better logical independence will still be the foundation of database in the future

29、.After commercialization, Oracle led the market with MySQL, Microsofts SQL Server and other relational databasesfor many years. From SQL, NoSQL to NewSQL, and even HTAP, business capabilities are driven in iterations.At present, cloud + distribution has become the only solution to the extreme needs

30、of enterprises, and hascreated the current outbreak of the database industry. In the current and ongoing industry cycle, advancedproducts and technologies need to revolve around the market to become the most important competitiveadvantage.Evolution of the classical architecture of distributed databa

31、sesSource: Frost&Sullivan, Leadleo4Patricia G. Selinger pioneered the cost-based query optimization actuator framework, followed by Goetz Graefes volcanic model-based actuator, which is still popular today.19701976After relational model, hierarchical model and mesh model, Dr. Chen Pinshan

32、invented (entity-relation)E-R model, which is widely used as a tool for data modeling.Transaction Processing: Concepts and Techniques by JimGrey from IBM guaranteed data integrity, security and parallelism by transaction processing.Boyce and Chamberlin proposed, originally called Sequel, to implemen

33、t SQL on top of System R, a prototype relational database management System developed by IBM.2Google releases Spanner and F1 papers, solving distributed problems with atomic clocks and the TruetimeAPI. The NewSQL age starts.2017A Relational Model of Data for Large Shared Data Banksby Edga

34、r F. Coddfrom IBM laid the foundation of relational database.Edgar F. Codd proposed the concept of OLAP online analytical transaction processing. Database transactions distinguish between OLAP and OLTP.1993Gartner introduces the concept of HTAP: the desire to enhance the capabilities of analytics on

35、 transactional systems. The need for HTAP is an ongoing future direction for major distributed database vendors.The NoSQL wave, which is represented by HBase and Cassandra that are based on BigTable and Dynamo respectively, is targeting unstructured and massive data expansion, but its transactional

36、support is a defect.MySQL, PostgreSQL relational databases open source to the market and gradually become widely used and the base for many other databases.AWS Aurora proposed The log is The Database in The Sigmod2017 paper. Databases on the cloud will generally replace traditional databases and ste

37、w towards cloud native.China: Database SeriesFrost & Sullivan Market I400-072-5588Colleges and UniversitiesNPU, THU, WHU, RUC, etc.Professional training institutionsENMO Edu, New Century college, Pangucloud class, etc.Database industry support system in ChinaConstruction of distributed database indu

38、stry support system5Source: CAICT, CCF Technical Committee, Company official website, Frost&Sullivan, Leadleo The development of distributed database in China has obtained demographicdividend. Technological innovation needs an advanced academic research system,industry-research integration needs clo

39、se industrial exchanges, and industrypenetration needs a talent training system that keeps up with the needs of thetimes.Research organizations with official backgroundsChina Computer Federation Database Technical Committee, China Communications Standards Association Big Data Technical Standard Prom

40、otion Committee (CCSA TC601)Spontaneous user organization by database technology enthusiastsACDU for DBA, ACOUG for Oracle users, ACMUG for MySQL users, PostgreSQL Branch of China Open Source Software Promotion Alliance for PostgreSQL usersOfficial technical community for specific database products

41、discussion Ali Cloud developer Community, Huawei Cloud openGauss community, PingCAP AskTUG community, PostgreSQL Chinese community, Aikeshengopen Source community, mobile cloud developer communityThird Party Technology CommunityITPUB、MoDB、DBAplusDatabase Industry Support OrganizationAcademic Organiz

42、ation of DatabaseDatabase vendorPingCAP University, Oceanbase College, AUCP of Ali Yun University, OpengAuss community of Tencent Cloud Noah Plan, Giant Sequoia University, etc.Professional CoursesTextbookTeaching serviceCompetitionTest certificationInnovation centerEcological talent marketNumber of

43、 papers in distributed databases ofacademic institutions in ChinaRetrieve institutions of papers with Distributed Database topic based on Web of Science core collection, select China region.Source: Web of Science, Frost&Sullivan, LeadleoTop conferences in the field of databasesVLDB、SIGMOD、ICDE Enter

44、prises Alibaba, Huawei, Tencent, Baidu, PingkaiStars, Dameng, DDA Jincang, etc.Colleges and UniversitiesTHU, UCAS, HUST, ZJU, etc.Numberofpapersindistributeddatabases by countryRetrieve countries and regions of Distributed Database papers based on Web of Science Core Collection.Source: Web of Scienc

45、e, Frost&Sullivan, LeadleoDatabase Personnel Training SystemChina: Database SeriesFrost & Sullivan Market I400-072-55881,0561,4071,4931,4351,9861,9472,0802,2181,8121,09223233537751,183 1,2048292012 2013 2014 2015 2016 2017 2018 2019 2020 2021GlobalChinaGlobal and Chinese distributed datab

46、ase patent applications,2012-20211,9471,396793627509404329239170168Patent analysis of distributed database vendors6From the perspective of patent application data, the number of patentapplications related to distributed database in China climbed from 22% in 2012to 76% in 2021. China has become the t

47、echnological innovation center ofdistributed database in the world.00Ant GroupHUAWEIAmazonIBMSGCCMicrosoftAlibabaJDSAPOracleGooglePinganZTETencentICBCInspur CloudBIDUInspurGBASEKingsoft CloudOceanBasePartofenterpriseswithdistributeddatabasepatentapplication, 2012-2022.2Distributed databas

48、e patent application by Chinese provinces, 2012-2022.2From the perspective of patent application data, this report observes the global innovation status of Chinese distributed database market and the innovation vitality of Chinese database enterprises.Global and Chinese patent application analysisTh

49、e number of patent applications related to distributed database in China climbed from 22% in 2012 to 76% in 2021. China has become the technological innovation center of distributed database in the world.Beijing, Guangdong and Jiangsu are the provinces with a large number of advanced distributed dat

50、abase manufacturers, and the cumulative number of distributed database patent applications leads the whole country.Keywords of distributed database patent topicsThe innovation of distributed database is built on the accumulation of research and application of database technology. The key sections of