Post-relational database = XML + relational database? (1)

When XML, InfoBase, Webserver, and other concepts are booming, we still see that the current database technology actually only solves 15% of structured data management problems for enterprises, how can we manage the remaining 85% of unstructured data?

Three mountains of pressure on relational databases

Relational databases are originally designed for mainframe applications based on host/terminal methods. Their application scope is relatively limited. After the Client/Server era, relational databases have achieved great development.

Relational database management systems have been dominant in data storage and management technologies for a long time. However, this traditional database management system adopts a two-dimensional data model, there are inherent constraints and restrictions.

With the rapid development of information technology, data processing requires more and more data types, and more unstructured data, it includes various reports, bills, electronic documents, website elements, images, faxes, scanned images, and a large number of multimedia audio and video information. In addition, with the surge in the number of popular website visits, it puts forward higher requirements on the storage mechanism of the database, the usage requirements of a large number of concurrent users, the efficiency of the storage space, and the data integrity and security.

Today, with the Internet and the trend of interconnection and collaboration between IT systems in the network environment, the external soil of databases poses new challenges to information management technologies. These challenges are like the three mountains that place pressure on relational databases.

Information complexity

Today, the complexity of information and the era of relational database theory are no longer the same. The structure of information is increasingly complex. As personalized services and other demands increase, information diversity and differences have greatly increased, however, relational theory is increasingly inadequate in managing information complexity.

In essence, relational databases use a two-dimensional model to describe complex object objects through a series of two-dimensional relationship combinations. All Objects represented by each table have no difference in modeling design, even if only one object has a certain attribute, a field must be created for it. If the differences between individuals are not only manifested in attributes, but also involve structures and relationships, we need to establish different tables and mappings for entities with differences. Therefore, many business systems today usually need thousands or even tens of thousands of tables. The database structure becomes very complex, and the information in the database becomes hard to understand. In addition, many types of real data cannot be processed well by relational databases.

High processing efficiency

In a sense, today's business system database information is a by-product of business system operations. The database is designed to support the operation of specific business applications. The data in the database is only used by a single system, so it is seldom considered whether the information is easy to understand. However, more and more systems need to collaborate and share information. It is increasingly important to make the information readable and easy to understand.

The relational database management system designed for static applications has not been optimized for efficient transaction processing. The result is usually some relational database products, which fail to achieve the expected results in the Web transaction processing process. The two-dimensional table data model of relational databases can process typical multidimensional data in most transaction processing applications. However, the result is that a large number of data tables are usually created and used, it is still difficult to build a data model that can simulate the real world. In addition, when the data needs to be output as a report, it is necessary to, in turn, connect a large number of scattered two-dimensional data tables and use indexes and other technologies to connect the tables to find all the required data, this will inevitably affect the response speed of the application system.

Application flexibility

As IT systems enter all aspects of social life, information is not only increasingly complex, but also the content and structure of its needs are constantly changing over time. the real world requires more and more flexibility and adaptability of information technology. relational Data theory uses a fixed modeling method. Once defined, any relationship and attribute is fixed, and it is difficult to make flexible adjustments as demand changes.

The ability of relational database technology to effectively support applications and data complexity is limited. The standardized design method originally used by relational databases is powerless for the design of complex transaction processing database systems and performance optimization. In addition, high development and maintenance costs also make it difficult for enterprises to afford.

Article content]

Page 1:Three mountains of pressure on relational databases

Page 2:Three types of post-relational databases

Page 3:Four Directions of database technology application

Page 4:Document: Development History of database technology