Analysis on the Reliability optimization technology of computer network

in the information age, the life of the network lies in its security and reliability. The most important aspect of the computer network is the information service which it provides to the user and the information resources it possesses, and the network connection brings convenience to the user at the same time, it also brings convenience to the network intruder. Therefore, the future computer network should have high security and reliability, can withstand high IQ network intruders, make users more reliable and more convenient to have a large number of personalized customer service.

A study on computer reliability model

computer network reliability as a system engineering science, after 5 more than 0 years of development, has been formed a more complete and sound system. Our computer network reliability is defined as: the computer network in the specified conditions, within the specified time, the network to maintain connectivity and meet the requirements of communication capabilities, known as computer network reliability. It reflects the ability of computer network topology structure to support the normal operation of computer network.
the reliability problem of computer network can be modeled as a graph reliability problem. The computer network model is represented by probability graph G (v,e), where the node set V represents the user terminal of the computer network, the host or server, and the edge set E represents the link of the computer network. The probability map of computer network model is a graph obtained after assigning a certain probability value to each side of the graph and the normal operation State of the node. The reliability problem of graphs contains two aspects: one is to analyze the problem, that is to calculate the reliability of a given graph, and the other is to design a graph with maximum reliability after given all the elements. When the reliability of graphs is not easy to solve, the failure degree (reliability + failure degree =1) can be obtained first, and then the reliability of the graph is obtained. The node and link failure models of graphs can be divided into three types, such as link failure model, node failure model, node and link mixed failure model, among which "node and link mixed failure model" are most commonly used.

Second, the design principle of computer network reliability

in the engineering practice of computer network design and construction, researchers have summed up a lot of specific design experience and principles, which has played a better standard and guiding role in the optimization design of computer network reliability. The following principles should be followed when building a computer network:
adhere to international standards, the use of open Computer network architecture, so as to support heterogeneous systems and heterogeneous equipment effective interconnection, has a strong ability to expand and upgrade.
Advanced and mature , practical, universal combination, the choice of sophisticated and mature computer network technology, select a practical and general computer network topology structure. Computer network should have strong interconnection ability, can support a variety of communication protocols. The security and reliability of computer network are high, and have strong redundancy and fault-tolerant ability. Computer network management is stronger, should choose advanced network management software and support SNMP and CMIP network equipment. Should choose a better computer network link of the medium, to ensure that the backbone network has enough bandwidth, so that the entire network has a faster response speed. Make full use of the existing computer network resources, reasonable deployment of existing hardware facilities, network cabling, the mature network operating system software and network application software. The reliability design of computer network should be as high as possible.
　　
Analysis of main optimization design method of computer network reliability
　　
improving the reliability of computer network related components and attaching corresponding redundant parts are the two main ways to improve the reliability of computer network. On the premise of satisfying the expected function of computer network, the redundancy technology (increasing the number of spare chain flagging) can improve the reliability of the local segment of computer network, on the other hand, it also improves the construction cost of computer network. Because each computer network link has the reliability and the cost, therefore the number of links in the computer network is less, correspondingly, the computer network reliability is higher. We will discuss the following aspects:
(a) fault tolerance design strategy of computer network
the general guiding principle of computer network fault tolerance design is: Parallel backbone, dual Network Center. The principles of the design of the computer network fault tolerance design can be referred to the following points:
A parallel computer network and a redundant network center are used to connect each user terminal and server to two computer network centers.

The interconnection of data links and routers within a wide area network. In computer network, the boundary network to the Network Center adopts multi-data link, multi-route connection, which can guarantee that the fault of any data link does not affect the normal use of local network users.

Posted in free paper Download Center http://www.hi138.com computer network design, should adopt a modular structure, hot-plug hot-dial function of networking equipment. This can not only have a flexible network mode, but also can replace the fault module in time without cutting off the power supply, in order to improve the computer network system long-time continuous work ability, thus can greatly improve the whole computer network system fault-tolerant ability.
network servers should adopt new technologies, such as dual-machine hot backup, dual-machine mirroring and fault-tolerant storage to enhance the server's fault tolerance and reliability.
in the fault-tolerant design of network management software, it should be implemented by multi-processor and specially designed fault-tolerant network operating system, which provides the basic fault recovery function with checkpoint.
(ii) dual network redundancy design strategy of computer network
The redundancy design of computer network is to add a kind of backup network on the basis of a single computer network, to form a dual network structure and to realize the fault tolerance of computer network by redundancy of computer network. In the dual network structure of computer network, each network node is connected by a double network. When a node needs to transmit messages to other nodes, it can be sent over a network in a dual network. Under normal circumstances, the dual network can transmit data simultaneously, or it can be used as a backup of the computer network system in the primary alternate way. When a network is disconnected due to some reasons, another computer network can quickly replace the work of the error network, so as to ensure the reliable transmission of data, so that the computer network in the physical hardware facilities to ensure the overall reliability of the computer network.
(iii) Adoption of multi-layered network structure system
the multi-layer network structure of computer network can most effectively utilize the business functions of the 3rd layer, such as the segment of network traffic, load sharing, failure recovery, reducing the general network problems caused by improper configuration or faulty equipment. In addition, the multi-layer network structure of the computer network can isolate the faults of the networks very well and support all common network protocols. The multi-layer mode of computer network makes porting of computer network more simple, because it preserves the network addressing scheme based on routers and hubs, and has good compatibility with the former computer network. The multilayer network structure of a computer network consists of three hierarchies:
Access Layer: The access layer of the computer network is the starting point for the end user to be granted access to the computer network. The access layer can provide further control of user traffic through filtering or access control lists. In the local network environment, the access layer mainly focuses on providing service functions through low-cost, high-port density devices, the main functions of the access layer are as follows: to provide the terminal network users with the access port of the computer network, to provide the bandwidth of exchange for the computer network, to provide a second layer of computer network services, such as VLAN membership and data flow filtering based on interface or MAC address.
Distribution layer: The distribution layer of computer network is the dividing point between the access layer and the core layer of computer network. The distribution layer also helps define and differentiate the core layers of the computer network. The hierarchy provides a boundary definition, where a potentially laborious packet operation is preprocessed. In the LAN environment, the most functions of the distribution layer are: The aggregation of V L A N, the access of department level or workgroup in computer network, the determination of network mode of broadcast domain network or multipoint broadcast domain network in computer network ;
(iv) core level
Computer core layer is the backbone of computer network. The primary function of the core layer is to exchange data as quickly as possible. This hierarchical structure of the computer network should not be involved in laborious packet operations or any slow processing of data exchange. When partitioning the logical functions of a computer network, you should avoid using functions such as access control lists and packet filtering in the core layer. For the hierarchical structure of the computer network, the core layer is responsible for the following tasks: Providing a connection between swap blocks, providing access to other chunks (such as server chunks), and exchanging data frames or packets as quickly as possible.
Looking at the future development of computer networks, people will be more and more demanding of the network. They want to create a "Click to everything" world, although this simple idea makes it a reality is not a very easy thing, but once recognized the good development prospects of computer networks, by virtue of human wisdom, we have reason to believe that our world will receive its unprecedented freedom.

Analysis on the Reliability optimization technology of computer network