Implementation Method for Configuring Modules Based on SNMP / Web
First, outline 1. Network management mode With the development of the Internet, the network system has become increasingly complex, increasingly diversified, and has made higher requirements for the management of network systems. To this end, the Internet's core mechanism IAB (Internet Activities Board) proposes a Simple Network Protocol SNMP (Simple Network Management Protocol) protocol based on computer network, and a telecommunication network-based public management protocol. The network management system is in principle consisting of four parts: a plurality of managed agents, at least one network manager, a generic network management protocol, and one or more management information libraries (MIBs). The network manager and the managed agent obtain network information by exchange management information, where the network manager regularly polls each managed agent, managed the agent to listen, and respond to network management queries and commands from the network administrator. This information exchange is implemented by a network management protocol, information residing in the MIB library of management workstations and managed objects, respectively. We call this model to manage the manager - the agent mode.
2. Network Management Five Features Configuration Management: Configuration Management Initialization Network and configures the network to enable it to provide network services. Configuration Management is a set of related functions necessary for identifying, definition, control, and monitoring of objects that make up a communication network, with the aim of implementing a particular function has achieved optimal network performance. Performance Management: Performance Management Estimated System Performance and Communication Efficiency of System Resources, its capabilities include monitoring and analyzing the performance mechanisms of the service network and their services. Trouble management: Trouble management mainly analyzes the cause of network failure, including three aspects such as fault detection, fault isolation and fault correction. Billing management: Billing management record network resources, the purpose is to control and detect the cost and cost of network operations. Safety management: Safety management includes management of authorization mechanisms, access control, encryption, and encrypted keywords. In addition, this feature also maintains and checks the security log.
Second, regarding the SNMP Network Management Agreement 1. SNMP introduction Simple network management protocol SNMP is a management protocol standard for Internet networks based on TCP / IP protocol. It divides the management part into two parts and specifies the standards of each part. The first part involves information communication, and the protocol specifies how the running management client is communicating with the server. The protocol defines the format and meaning of the client and server exchange packets, and the format of the name and address. The second part about the controlled data, the standard specified the gateway must hold the data item, the name and format of the data item. SNMP design principles are simplicity and extension. Simply look through information type limits, request / response, or protocols; the extension is mainly achieved by separating the management information model with the protocol, the detailed predetermined specification (MIB) of the object being managed. The architecture of SNMP is divided into SNMP managers and SNMP agents. 2. Basic primitive SNMP uses polling and event-driven two access methods. SNMP uses basic five communication primitives to obtain management information. These five basic primitives are: Get-Request, Get-next-Request, Set-Request, Get-Response, Trap. Get-Request is used by Manager to retrieve certain variables from the agent; get-next-request is used by manager to retrieve the next variable of a variable from the agent; set-request is used by manager to set or change the Agent The value of the last variable; get-response is the response sent to Manager to Manager; TRAP is reported to Manager to the Manager. An exception event. Among them, these three operations of Get-Request, Get-next-Request and Set-Request are characteristics with atoms. 3. In SNMP, the PDU format is exchanged between a management station and a proxy in the form of SNMP messages. Each message includes a version number named named SNMP version, a group name for security, and five protocol data One of the units in the unit. This agreement data unit, we call the PDU. Basically, the PDU format consists of PDU type, request ID number, variable domain, and several other aspects. The specific situation is shown in Figure 3.1. Figure 2.1 Five Basic Original PDU Format 4. MIB Library Network Management Information Library (MIB) is an important part of network management. Specifically, MIB is the standard of network management data. This standard specifies the standards of network elements (such as RFC1213 [36]). This standard details the data items, data types, and operations allowed in each data item in detail. Basic network management can be achieved by accessing access to these network elements for the access access of these network elements. Comprehensive analysis of these statistical contents can be achieved. Third, existing network management standards and product 1. The existing network management standard WBEM (Web-Based Enterprise Management) WBEM is launched in July 1996. WBEM is Microsoft's first proposed, including 3COM, Cisco, Intel, etc., 60 suppliers, including 3COM, Cisco, Intel, support this standard. Its purpose is to establish an industrial standard that enables managers to use any browser management distribution network, system, and apps. The architecture of the WBEM is shown in Figure 3.1. Figure 3.1 WBEM architecture
WBEM definition describes architecture, protocols, management mode, and object managers, managers use HTML or other Internet data formats and use HTTP transfer requests. WBEM includes the following three parts: (1) HMMS (HyperMedia Management Schema), it is an extensible, independent public data description mode, which can describe, instantiate, and access various data, is managed for various managed The top of the object is abstract. It consists of two layers of core mode and specific domain mode. The core mode is composed of high-level classes and attributes, and the elements of managed environments are divided into managed system elements, applications, resource components, and network components. The specific domain mode inherits the core mode and uses its basic object to define an object of a particular environment. (2) HMMP (HyperMedia Management Protocol), it is a protocol for access and control mode, with management information between HMMP entities, which belongs to the application layer, and the HMMP client issues a management request to the HMMP server, HMMP The server returns respond after the management task is completed. (3) HMOM (HyperMedia Object Manager), which implements most of HMMP and as a proxy representative HMMP customer request. Among them, HMMP customers mainly communicate with the assigned HMOM, which can directly meet the request, mitigating the burden on multiple devices on the HMMP customer positioning and managing the network. WBEM design goals are: (1) Provide a new framework to overwrite traditional management: configuration, fault, billing, performance, security management, and operation management and planning; (2) Establish this WBEM system in transmission, security, and configuration The standard range and the existing Internet standard; (3) Provide a data model to ensure the modeling and management of the system, network and application uniform; (4) The number of huge and distributed features based on the number of management elements, Provide a scalable solution. WBEM emphasizes the management of network elements and systems, including network devices, servers, desktops, and applications.
JMX (Java Management Extensions) JMX is JMAPI [31]. JMX is committed to solving distributed system management issues, so it is very important to be suitable for a variety of different environments. In order to be able to use the powerful Java computing environment to solve this problem, Sun has expanded the Java underlying class library to develop a dedicated management class library. JMX is an assembly that applies programming interfaces, expansion objects, and methods that can be used to span a series of different heterogeneous operating system platforms, system architecture, and network transport protocols, flexible development seamlessly integrated systems, networks and Service management application [42]. It provides a specification for user interface guidance, Java class, and development integration system, network and network management applications. The management object is the core of the JMX application. JMX Structure [10] includes: supporting the Java web browser user interface, managing running module ARM (Admin Runtime Module) and applications. These three components communicate between the RMI (Remote Method Invocation). It will be described here that RMI is a program that causes a Java virtual machine (JVM) to call another JVM in the remote server. User interfaces are used to publish management operations, which can be excited over a browser or through a separate application. Managing run modules to provide an instantiated management object to the application. It includes an Agent object interface, a notification interface, and a managed data interface. Application refers to those tube device units. JMX is a complete network management application development environment, which provides: vendors need to collect complete feature lists, generate resource checklist tables, graphical user interfaces; access SNMP network API; host remote processes; Database Access Method. JMX This lightweight management infrastructure is to achieve abstraction on the service managed resources. It provides the basic class of low-level, and developers carry out the integrity and consistency of most public management classes. Expand to meet the needs of specific network management applications. WBEM focuses on constructing new protocols and structures to manage networks. JMX is a software framework that focuses on construct management tools and try to use mature techniques. 2. Existing Network Management Open Platform HP OpenView HP OpenView is the first network management open platform, and its core framework provides an open system environment for basic applications. Access to public management services through the API, and make full use of the openness of the network management system. It is not only an open platform, but also provides users with direct management applications. HP will design the structure of the OpenView network management platform into an open distributed architecture and define a comprehensive service and facility environment, divide network and system management issues: communication under the next structure, graphical user interface, management application, management Services and managed objects. These parts are accessed through standard application interface. OpenView has been widely accepted and supported by many third-party application system developers. IBM NetView NetView [33] is developed on the basis of HP OpenView. Its core code has no changes, but a large number of applications, especially the SNA (Systems Network Architecture) network support. NetView does not have the ability to understand the dependence between network devices, but it provides powerful information filtering capabilities, and reducing part redundancy warnings by using a valve value.
Sun NetManager Sun NetManager is a network management system of a distributed management architecture that can manage the DECNET and FDDI network environments, and can cooperate with IBM's NetView. The entrusted agent and collaborative control are fully reflected in its distribution. Its integration tool provides troubleshooting management, configuration, billing, and security services. But it is basically a development platform, requiring third-party manufacturers to develop applications.
4. Configuration Module The role of network management and implementation step communication network is the initialization, naming, and control and modification of the system parameters for system definition, naming, and components. Configuration management can be divided into physical configuration and logical configuration. Physical configuration refers to a description of the physical features of each component (such as the location, model number) of the router, and the circuit connection between the components. Logical configurations are used to control the initialization, operating mode, and data processing functions of these components. Currently changed networks require network management to have the ability to dynamically configure networks. Configuration Management Gets data from the network and uses these data to manage all network devices. Its main function is to master and control the status of the internet network, including the status of each device in the Internet and its connection relationship. The configuration of the network device determines the performance of the computer network. The most important role of configuration management is that it enhances network managers to control the network configuration. Manager can compare the configuration data being used and stored in the system, and modify the data; network configuration management can provide management Taking the visual interface, it is convenient to process the processing of network device information; at the same time, by providing the latest network unit's list, configuration management can provide a further help to network managers, enabling managers to know the network device configuration, Can be used to track network devices, record information about device sources, numbers, etc. The network configuration management function module is also required to have a good connection relationship with performance modules, fault modules, billing modules. The events of other modules are displayed by configuring the visual interface of the module, while configuring the module's backtracking of these events to perfect the entire system. In the configuration module of the network, the following four steps: 1. Collect about current network configuration information; 2. Topological management of the network; 3. Configuration management of network device information; 4. Save data and maintain the latest device list of all network units and generate a variety of reports. 2. Collection Network Information Data Module Figure 4.1 shows the process of describing the collection of network devices discovery and network information data. In the network information data discovery module, it is primarily including the following aspects: 1. Getting the system initialization data system initialization data is required only when the system is running for the first time. During the network topology, you need to get the following aspects from the system: (1) The seed address of the network topology and the topology of the topology of the query; (2) Various information of the MIB library of the network topology: MIB The library file, the read / write common name, port number; (3) system function setting data, if it allows configuration, whether to allow troubleshooting, etc .; (4) variable data required for other functional modules, including polling Time, the scope of the device of the monitor, etc.; Figure 4.1 Network information data discovery structure
2. The data of the main entry is currently in key devices, and there are many search for network topology in many management products, but the degree of implementation is different. .Transcend: Developed by 3COM. The automatic search function of the product is relatively simple. You can only detect the host of the current LAN, which cannot determine the interconnection of each host, and cannot determine the interface type of the host; .CISCO Works: Developed by Cisco. It uses Cisco's expertise - CDP (Cisco Discovery Protocol) to perform accurate search for its company's network products. However, CDP is only suitable for network products for Cisco types, which do not have compatibility for other manufacturers; .Netwalker: Developed by Fujitsu. It is a better network management product. Ability to search for network devices running in the current network, compare network interconnects and network interfaces, and reflect the network current status in real time. But it is only suitable for the UNIX environment, and the search for the network requires a lot of time. Based on the automatic search of many network management software, the target for automatic search for the network management is:. Implement simple, search speed, restrictions on search, can search better The entire park, as far as possible to expand the network search range; accurately determine the interconnection of each subnet in the network; determine the survival of each network device. (1) The main topology main tribute refers to the connection between critical devices including routers, their interfaces and subnets, namely routers, routers, routers - subnets, and interfaces. This information needs to be searched for the network. In the network topology, each subnet communicates with other subnet through the respective routers, which can be a local area network, or a subnet of a local area network, they are connected to a port of the router. One port of the router can be connected to one subnet, or it can be connected to other routers. When a machine of one subnet sends data to other subnets, the packet first reaches the default router of this subnet, the default router detects the destination address in the packet, determines whether the destination address is in line with himself according to its routing table In the connected subnet, if yes, the packets are sent directly to the destination, otherwise forward to the next router specified in the routing table, the next router is performing a similar process, push it, the packet will eventually arrive at the destination . When initializing system data, you can get the starting seed node address of the topology and the number of hops to be queried. When establishing a web map, first, the management personnel are required to understand the overall structure of the network. To create a layered web map, you must determine the composition of each layer, determine the connection between the subnet, and the device. Then use automatic discovery to establish a current list of all network devices in the subnet.
(2) The variable that needs to be used for the main topology determines the type of network device through the MIB library's ipforwarding and sysservices variables. SYSSERVICES information is mainly providing which layer service in the ISO reference model. L is the protocol layer number, each layer is represented by Equation 2 (L-1), returning to the sum of each layer value; ipForwarding is used to determine if the device has a function of forwarding the IP packet forward, thereby determining whether the device is Gateway. Network interface of network devices IFTABLE use. First access the address table of the network device, get the network address and interface index of the various devices of the device interface, then according to the interface index, query the device's interface table, you can get the type of interface to other network devices. Through these interface types, it can be judged to the connection network device. Among them, ifdesr and iftype named the interface, respectively, give it the type. There is a system variable in the router that is sysObjectID, which is different for different types of routers, so it is not enough to marke a router. Referring to the OSPF protocol to make the smallest IP address in the router interface table as its tag. By utilization of variables in the router IP address table iPadDRTable, it is possible to perform a determination of the same router for multiple IP addresses. Router address table ipAddrTable concerning defined as follows: ipAddrTable OBJECT-TYPESYNTAX SEQUENCE OF IpAddrEntryIpAddrEntry :: = SEQUENCE {ipAdEntAddr IpAddress, // IP address ipAdEntIfIndex INTEGER interface, // index ipAdEntNetMask IpAddress interface ...... // interface resides subnet mask Code} The MIB library IP routing table is the most important Table table in the topology. It contains the address of the route, the next hop address, the subnet mask, etc., which is defined as follows: iProuteTable Object-TypeSyntax SequenceOf iProutentryipRouteEntry :: = sequence {iProutedest ipaddress, // Route destination address iProuteiFindex Integer, // Route Local interface index iproutenextHop ipaddress, // Route Next Routing address iProuteType Integer, // Route IPROUTEMASK ipaddress ... // Routing destination Net mask} (3) Main topology needs to use the algorithm master topology traversal query, which follows three related principles: The gateway of the query routing table is adjacent to the iproutenexthop in the routing table; if iProutenexThop is on iProuted In the subnet, iProutenexThop is adjacent to iPrOutEdest; if the gateway being queried is the same as iProutenexThop, the gateway is adjacent to the corresponding iPrOutEDEST. According to these three related principles, there is such a traversal algorithm:
Initialize to access the router queue, initialize the access router binary sorting tree; put the default router's flag IP in the queue to be accessed; WHILE (router queue to be accessed is not empty) {Remove from the router queue to be accessed A router, for the currentrouter; if (CurrenTrouter joins the access route binary sort tree successfully) {Initialize the local connected router queue and local connected subnet queue; Access the CurrenTrouter routing table, for each routing entry: {if (iproutetype = = Indirect) Put the IPROUTENEXTHOP in the routing table to locally connected route queues; if (iProuteType == Direct) puts iProuted & iProuteMask in the locally connected subnet queue;} Traverse local Each router in the router queue, put its flag IP to the route queue to be accessed;}} Description: a. Accessed router is a global non-duplicate queue, its member is the router flag IP, in order to improve its members Do not repeat the efficiency of maintenance, the data structure of the binary sort tree. When the joining a router flag fails, it indicates that the router has been accessed; b. The members to access the router queue are also the flag IP of the router, but can be repeated, the repetition is not repeated; c. This algorithm is a depth Priority algorithm. It retrieves according to the initialized data information.
3. The data obtained by the subnet topology is two ways to topology of a subnet. First, any router uses the ARP protocol to store a map of all subnets to the NIC (MAC) address directly connected in the Cache table in a dynamic update manner. Through the MIB variable IPNetTomedianetAddress in the ARP table, this IP address can be acquired one by one, and these IP addresses match the pattern. If success, the host corresponding to the IP address is proved to be a member of the subnet. The advantage of this method is that the value of the acquisition value is faster, and the deficiencies are because the update of the ARP table has a certain static time interval, and cannot be updated in real time. With this method, if you need to update the ARP table, you can modify the update time via the Telnet simulation according to the user settings. Second, mainly uses the ICMP protocol in TCP / IP. Implementing it mainly experiences the following steps: (1) Determine the gateway, subnet address, and subnet mask of the subnet, subnet address is obtained by "operation" and "operations for IP and Mask; (2) Getting the address of the gateway The type of subnet; (3) obtains the number of machines of the subnet through subnet address and subnet mask; (4) In a given subnet, the IP address of this section performs PING operation, and the IP will be detected. Address Record to the IP address table; (5) According to the Socket request, you can obtain additional information of the survival machine, you can find other information such as the domain name of its IP address. The PING operation is actually issuing an ICMP response request package for the host's IP address, sending and receiving an ICMP response package in an asynchronous manner. Use asynchronous request to improve the speed of detection while not causeing blocking. Because the Java language does not support the ICMP protocol of the IP layer of the TCP / IP protocol, that is, the original sockets in Winsock are not supported, other methods must be processed. There are three ways to solve the operation of the machine PING, as shown in Figure 4.2. One is to generate a .dll file that can be called for Java applications using Java Native Interface. JNI writes a program that implements Winsock Raw Socket with C language, providing header files coupled to Java language. This generation DLL file includes receiving request calls to the Java application to return to the Java main program. The second is to perform ECHO request through the 7th port of Java Socket. The disadvantage of this method is that the information requested is not very sufficient and the amount of information is small. Another way is that the Java program and the underlying C language program are a relatively independent part, they transferred through a database, and the database plays a function of conversion Java and C language. The ICMP Baotou is stored in the data unit stored in the IP packet. The ICMP package is first filled according to the format of the IP packet, and then creates a socket with the SOCK_RAW type, call the function socket (), and use asynchronous mode, use WSaasyncselect () Function Registration FD_READ Event, Final, use Sendto () and Recvfrom () to send and receive ICMP packets. Figure 4.2 Java implementation of ping
4. Further strategy (1) A topological algorithm in front of an heuristic idea is very effective to a certain extent, and the speed of the topology is also very fast. But it has a defect. That is, in a particular domain, all subnet information relies on the characteristics of the device's SNMP. Also, even if the device (such as routing) has a function of supporting SNMP, due to the problem of security access, it is sometimes difficult to obtain the MIB value of the device, there is an incompleteness of the topology. Here is an heuristic algorithm to get subnet addresses and subnet masks of the host from a group of known host IP addresses. Assuming that we are known a group of host addresses A1, A2, A3 connected from a routing, the route address is A. These addresses will have a common subnet address and subnet mask. We assume that these four addresses are: 202.114.33.195, 202.114.33.216, 202.114.33.228, 202.114.33.8. In order to clearly describe this algorithm, we omitted the prefix of these addresses, address is 11000011, 11011110, 11011010,00001000. These values are AND operation and the result is 00000000. Obviously, the address of the subnet is 202.114.33.0. When the AND results are not 0, OR operation is performed. Ensure that the rear end is all in the range of 1 in the result of the OR operation, and the AND operation corresponds to 0. Otherwise, redefine the assumption address and finally obtain its subnet address. (2) Algorithm [43] A DNS Domain Name Server keeps all machine IP addresses and machine names in this domain. This method will be used to initialize the data set. Temporary Data Set = DNS_DOMAIN_TRANSFER (Domain); initializing the hash table of Ands and Ors; while (Node is a temporary dataset) {ping that node; if the node is survive, the permanent data set = this node, otherwise Continue; Traceroute this node; current Router = The most recent hop of this Traceroute; uses DNS queries to find all IP addresses of the Router; the current gateway = current Router's' 1'bit bit xor this node; Ands (current gateway) = Node And and and and and and ands; ORS (current gateway) = this node or ORS (current gateway); new subnet = ands (current gateway); new subnet mask = not (andS (current gateway) xor ORS (Current gateway); store this node;} This algorithm can be obtained in all the subnets and its subnet masks in one domain, and the query is relatively wide. It is not enough to spend a relatively high cost.
3. The main part of the network topology module network topology module is shown in Figure 4.3. 1. The network management system of the network topology module provides a graphical interface that generates a hierarchical view of the network to display the current network topology. Moreover, they generally have automatic discovery tools, as well as tools for establishing a map, which can help network managers when establishing a web map. According to the algorithm mentioned earlier, the graphical map of the current network can be established. This graphical map is the same as above, as well as the display of the main topographic display and the sub-topology. Generally speaking, the automatic build map does not completely reflect the geographic and functional layout of the network, and needs to make handmade modifications. The display of the main topology is a complicated issue. How to objectively reflect the router - router and router - the subnet does not have a relatively unified algorithm. In this system, the network main topography uses the hierarchy obtained by the router depth overhead algorithm in the interface from the upper to the upper-to-bottom display; at the same time, all subnets are distributed in the router for those with subnet connection. around. Any subnet has a certain topology, such as bus type, star ring, etc. Subnets should be displayed according to different subnet types. When a device in the network has failed or other events, the device reports to the network management system, and the corresponding symbols in the web map will change color or flash, and propagate to the upper map to make the manager promptly time. Understand the situation of network devices. According to the international practice, green indicates that the equipment has no error. The yellow representation device may have an error; the red represents the device in an error status; the blue represents the device running, but in an error status; orange indicates that the device is not properly configured; gray indicates no information The purple represents the equipment being queried. Figure 4.3 Network Topology Module 2. Network Topology Topology Server creates initial operations throughout the system, which monitors event services for each function module. Once there is an event related to the entire topology display module, the topology server will process these related events.
3. The map server map server builds a map that reflects the device connection status and geographic location based on the device information obtained by the map discovery module API function. The network environment is constantly changing, the device exists in the network may change, and the configuration parameters of the device will also be adjusted frequently. Therefore, the map server needs to do two works during the entire system: on the one hand, the map manager will and topologies The server performs a message connection to handle various event services; on the other hand, the map manager is used to process the change request of the Browser client to the map, especially for the changes of the new object changes and the object of the object, the manager will be on the map. These dynamic changes are shown, and these changes are filled into the database through information filtering, while the manager will also provide all online client users on other clients connected to the manager. Real-time topology and asynchronous updates.
4. Network Device Configuration Management Module Due to various circumstances, managers may often modify and adjust the configuration of the network. For network device configuration management modules can be divided into two parts. One is a module configuration for the system, and its other is a modification configuration of the network device. The system module configuration, here we only talk about the configuration information required in the configuration module. Mainly includes: information storage database configuration, such as data sources URL, JDBC driver DRIVER name, login name and password of the database, and will also configure those permanent information about topology maps, events, and polling objects in each module; Configuration in the topology module, such as polling discovery intervals, SNMP request timeout, configuration of the SNMP Community Name list, network node information filtering information, etc. Configuration in the topology display, such as the display icon of the device node. The modification configuration of the network device node is mainly the increase in device nodes, deletes, and changes to its properties. Devices that are automatically discovered automatically for users to have manually added topologies need to be manually configured. Using the configuration management tool to change the configuration of the device, these change procedures can be automatically recorded before delivery to the device. In addition, the network management system can also verify whether the changed configuration is suitable for the device, and gives the necessary warning information when the manager is incorrectly configured. Configuration management also considers the synchronization between the objects and real-time updates of each client object instance. 5. Configure the access module of the information, while generating a configuration report A valid management system should store all network configuration information so that the administrator can quickly and efficiently access the configuration data. Relationship Database Management System You can use users to associate various types of data, allowing system servers to be valid, fast access, association, query data, and establish a list of devices. At the same time, the relational database provides the concept of triggers and stored procedures, so that the server can call the stored procedure through the JDBC, encapsulate small modules of commonly used query processing, thereby increasing the speed. The configuration management tool should also generate the necessary reports to make the manager timely understand the configuration of the entire network. The configuration report requires detail the configuration of each network device, including its name, network address, manufacturer and operating system, etc .; requests with all recent network changes; requires a summary report on the entire network equipment.
V. Conclusion 1. The development and working environment of the system software and hardware environment is as follows:. Operating system: Window. Development Tools: Java, Visibroker For Java, SQL Server, C. Hardware Environment: PC. Management Object: Park Computer Network Management 2. In the future, future systems should do a further effort in the following :: Support SNMPv3. Support for higher version of the protocol is inevitable. This requires a more complete writing of SNMP implementation primitives. : The system needs to develop a relatively complete, universal interface topology, to complete the display of the topology of different situations. : The efficiency of the system operation is improved. The system module is required to use more flexible and use rates. This includes two problems, how to improve the efficiency of data acquisition, and due to web-based methods and distributed use, require operational speed, especially messaging, faster. In order to meet the needs of today's corporate network, network management solutions must have 4 key features: scalability, reduce management complexity, and more independently solve problems and accessibility of information. The development requirements of the system are facing the user, facilitates the user purpose. Such a network management software will have a market without development.
Reference 1. Xie Double, Based on the Design and Implementation of the Configuration Management of Network Management System Based on CORBA / Web Park, Cancellation of Huazhong Normal University, 2000.72. Wang Wei, Gu Guanqun, CORBA Technology and its application, Southeast University Press, June 3, 1999.贤 道, Anometown, Network Management Agreement and Application Development, Tsinghua University Press, July 4, 1998. Research, Computer Application, Vol.19, No.2 Feb.19995, Search Algorithm, Computer Application, Vol.19, No.2 Feb.19995. Automatic Top Topology in Network Management System. Journal of Chinese Science and Technology University. R. Totte, P.Patrick, CORBA tutorial, public object request agency architecture, Li Shixian and other translation schools, Tsinghua University Press, October 1999, the author member name: yuannian
