Mobile Art: data processing in a three-tier structure

With the further popularization and development of Internet, network-based data processing has become the central link of enterprise informatization. The flexibility, security, and scalability of databases have become the focus of data processing technology. This article will take the data processing structure as the main line, review and look forward to the evolution of network-based data processing methods.

Layer 2 Structure: Yesterday's yellow flower

In network-based data processing, the client/server layer-2 computing mode has a milestone. This model was introduced into the industry in the late 1980s S. It provides unprecedented bi- al communication and flexibility for multi-user systems, revolutionizing the traditional application design and system implementation methods, soon, it was widely used in the design and development of various types of software systems. By the early 1990s S, this computing model has become the mainstream technology in the industry.

The basic structure of the client/server layer-2 computing mode 1 shows that the client provides the user interface and runs the logic processing application. The data server accepts the client SQL statements and queries the database, and then returns the query result. In an image metaphor, the server is like a warehouse. The main function is to store raw materials and finished products, but not to process them. The client is like the final processing workshop, complete the entire manufacturing process from raw materials to finished products, and then store the finished products to the warehouse. Because the business processing logic is mainly executed on the client side, this structure has also become a fat client structure.

　　

 

This structure has a simple structure and a simple development and running environment. However, this structure also produces native problems:

First, it is about scalability and installation and maintenance of the system.

When successful applications at the department level are gradually promoted to key tasks at the enterprise level, these applications run well in small-scale environments. Once they enter the large-scale production system, their performance drops exponentially, as a result, the reliability of the system is affected. After the system is developed, the installation of the entire system is complicated. In addition to installing applications on each client, you must install the corresponding Database Connection Program and complete a lot of system configuration work.

The second is the security of the system.

In the layer-2 structure, a large number of Code-based enterprise business processes reside on the client, which brings a great test to the system security. At the same time, with the increase in the number of users, the maintenance cost of such business logic is also increasing.

Finally, it is a functional obstacle for inter-system communication.

When the second-layer computing mode expands from a department-level application to an enterprise-level application, there is almost no interactive operation between the second-layer applications, so it is difficult to implement the component technology of the distributed system.

These problems are native issues of the two-layer structure. They cannot be solved by patching and developing the two-layer structure. To truly solve these problems, we must fundamentally change this layer-2 structure design.

Three-tier structure: in the ascendant

From the above analysis, we can see that enterprise application systems are constantly expanding and new applications are constantly increasing, the existence of native problems such as system scalability, maintenance cost, data security, and communication functional barriers between applications based on the traditional layer-2 data processing structure is too limited in applications. In this context, the three-layer data computing structure emerged.

What is a three-tier structure? As the name suggests, the data processing process is divided into three parts: the first layer is the client (User Interface), providing user-friendly access to the system; the second layer is the application server, the business logic of the company is implemented. The third layer is the data server, which is responsible for the storage, access and optimization of data information. Because the business logic is extracted to the application server, the burden on the client is greatly reduced, so it also becomes a thin client structure, as shown in Figure 2:

　　

 

We can see that the layer-3 structure adds an application server based on the traditional layer-2 structure and processes the application logic separately, so that the user interface and application logic are on different platforms, the communication protocol between the two is defined by the system. With such a structure design, application logic is shared by all users, which is the biggest difference between two-layer architecture application software and three-layer application software.

What are the advantages of this layer-3 structure data processing? An image metaphor is that it adds a standardized processing factory and transmission route between the two-layer structure to efficiently produce and transmit the products required by the client. Specifically, there are the following points:

First, the entire system is divided into different logical blocks, greatly reducing the cost of application system development and maintenance.

The layer-3 structure separates the parts and business logic parts from the application server based on the customer layer, communication between clients and application servers, application servers and database servers, and data exchange between heterogeneous platforms can all be achieved through middleware or related programs. When the business logic of the database or Application Server changes, the client does not need to change, and vice versa. This greatly improves the reusability of system modules, shortens the development cycle, and reduces maintenance costs.

Second, the scalability of the system is greatly enhanced.

The modular system makes it easy to expand the system in both vertical and horizontal directions: on the one hand, the system can be upgraded to a larger and more powerful platform, at the same time, you can increase the scale to enhance the network application of the system. The distribution data processing is possible because of the restriction of system equivalence.

Three-tier structure: Creating advantages

Previously, we talked about the technical advantages of building a layer-3 data processing structure based on the application server. However, just as "enterprise organizations must follow the strategy", technology can be integrated into the market environment and bring a competitive advantage to the enterprise to truly reflect its value.

The role of the three-layer structure in creating an enterprise's competitive advantage is mainly reflected in the modular design of the three-layer structure, which enables the user to expand the system based on the existing structure, thus improving the speed of Enterprise Informatization and business level; at the same time, the emergence of middleware in the layer-3 structure allows users to directly select appropriate products from the market to build the system, greatly reducing the development cycle and development costs. Currently, the middleware market has been highlighted: BEA's tuxedo and IBM's CICs/6000 Based on C and Java language have already laid a competitive advantage in the market as pioneers in this field, the HP Application Server, which relies on HP's service advantages and bluestone's technical advantages, is fully based on the Java language and based on open standards, highlighting its advantages in cross-platform and reliability, become a powerful Challenger. With the further expansion of the market, more software vendors will be involved, making customers more choices.

In short, in the ever-changing information age of technology and market, the competition rules have evolved from "big fish eat small fish" to "Fast fish eat slow fish", based on middleware, building a three-layer data processing structure will accelerate and better promote the informatization of enterprises, enhance the competitiveness of enterprises, and win advantages and stand out from the fierce market competition.