WO2004051927A1 - Procede et systeme de gestion en grappes d'installations du reseau - Google Patents

Procede et systeme de gestion en grappes d'installations du reseau Download PDF

Info

Publication number
WO2004051927A1
WO2004051927A1 PCT/CN2003/000720 CN0300720W WO2004051927A1 WO 2004051927 A1 WO2004051927 A1 WO 2004051927A1 CN 0300720 W CN0300720 W CN 0300720W WO 2004051927 A1 WO2004051927 A1 WO 2004051927A1
Authority
WO
WIPO (PCT)
Prior art keywords
cluster
network
management
devices
management device
Prior art date
Application number
PCT/CN2003/000720
Other languages
English (en)
French (fr)
Inventor
Jiewen Luo
Anping Hu
Haiyin Ma
Original Assignee
Huawei Technologies Co., Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Huawei Technologies Co., Ltd filed Critical Huawei Technologies Co., Ltd
Priority to EP03812116A priority Critical patent/EP1575212B1/en
Priority to JP2004555961A priority patent/JP4081472B2/ja
Priority to AU2003257805A priority patent/AU2003257805B2/en
Priority to US10/537,488 priority patent/US7904535B2/en
Priority to DE60317705T priority patent/DE60317705T2/de
Publication of WO2004051927A1 publication Critical patent/WO2004051927A1/zh

Links

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L61/00—Network arrangements, protocols or services for addressing or naming
    • H04L61/09—Mapping addresses
    • H04L61/25—Mapping addresses of the same type
    • H04L61/2503—Translation of Internet protocol [IP] addresses
    • H04L61/2514—Translation of Internet protocol [IP] addresses between local and global IP addresses

Definitions

  • the present invention relates to a network device management method, and in particular, to a method and device for managing transmission control protocol / Internet protocol (TCP / IP) network devices in a cluster manner.
  • TCP / IP transmission control protocol / Internet protocol
  • the first method is to directly connect the network management device to the managed network device through a serial port to perform configuration management on the network device.
  • the main advantage of this management method is that the management method is relatively simple, no additional equipment is needed, and because the managed network device is directly connected through the serial port, there is no need to assign a corresponding management IP address when configuring the network device, saving IP resources.
  • network equipment may be distributed in a large range, especially in telecommunication networks, the number of data equipment is very large and the deployment points are very scattered. Therefore, this management method will inevitably bring maintenance difficulties and a large workload. Puzzle.
  • the second method is based on a single network management protocol (SIP) or a remote login protocol (TELNET) terminal emulation protocol, which can manage TCP / IP network equipment from remote.
  • SIP network management protocol
  • TELNET remote login protocol
  • the remote management of the managed network equipment is implemented through the network management terminal or the TELNET simulation terminal.
  • each managed network equipment must be configured with a public network IP address, which will cause the waste of IP address resources, especially in the In telecommunications networks with a large number of devices, this defect is particularly prominent.
  • the last method is proposed on the basis of the second method.
  • the core idea is to implement the management of network devices by using the serial number uniformly assigned by the network device manufacturer as the device address, thus saving IP address resources, but this This method is only applicable to equipment of a specific manufacturer.
  • the network management program runs on a private protocol stack. Therefore, the management protocol has almost no openness and cannot implement unified management of network equipment manufactured by various manufacturers.
  • the object of the present invention is to provide a cluster management method for network equipment, which can realize centralized management of a large number of TCP / IP network equipment by using only a small amount of IP resources, thereby reducing maintenance workload and facilitating network management.
  • Another object of the present invention is to provide a cluster management apparatus for network equipment.
  • a plurality of network devices form a cluster, and at least one of the network devices is set as a cluster management device and a public IP address is configured, and other network devices are configured by the cluster.
  • the management device configures and updates the private IP address and route.
  • the network management device manages network devices in the cluster according to the following steps: (A) the cluster management device establishes an IP data channel between the network device in the cluster and the network management device via the cluster management device; and
  • the network management device manages network devices in the cluster via the IP data channel.
  • the cluster management device configures and updates private IP addresses and routes for other network devices according to the network topology structure information and device information of the network devices in the cluster. Even better, the cluster management device dynamically configures a private IP address for another network device.
  • the cluster includes a plurality of the cluster management devices, and one of the cluster management devices is responsible for managing the configuration and update of private IP addresses and routes of the network devices in the cluster, and the network management device. Communication with network devices in the cluster.
  • the cluster management device fails, one of the other cluster management devices can be designated according to a predetermined policy to manage the configuration and update of private IP addresses and routes of network devices in the cluster, and network management. Communication between the device and network devices in the cluster.
  • the network management device designates a device in the network as the management device of the cluster, and configures the device accordingly;
  • the cluster management device starts the topology collection process to collect network topology structure information within a specified number of hops in the network
  • the network management device specifies candidate devices in the topology that need to be added to the cluster, and notifies the cluster management device to start the set. Joining process of group member devices;
  • the cluster management device adds the designated candidate device to the cluster, so that it can be configured accordingly and become a member device of the cluster;
  • the member devices in the cluster are managed by the cluster management device.
  • Management packets from outside the cluster and targeted at the member devices are forwarded to the cluster management device through a standard network address translation process (NAT).
  • NAT network address translation process
  • the corresponding member device processes, and the member device processes the management message according to a normal processing procedure.
  • the corresponding configuration of the cluster management device as described in step (1) includes configuring the following content on the device: the cluster name, the enabled state of the cluster, the management IP address pool of the cluster, the effective retention time of the cluster state, and the member devices Handshake interval, the role of the management device in the cluster, and the IP address of the management device.
  • the process of adding candidate network devices to the cluster as described in step (4) includes: (A1) The cluster management device sends a cluster addition request to the candidate network devices that can join the cluster;
  • the candidate device judges whether it can join the cluster according to its own situation. If not, it returns a refusal to join the response and ends the process of joining the cluster; otherwise, it returns a confirmation response to the cluster management device that it agrees to join the cluster;
  • step (A3) After the cluster management device receives the confirmation response from the candidate device, if the candidate device agrees to join, it sends configuration information including the private IP address, member number, handshake interval, and state retention time to the candidate device, and the candidate device receives After the information is obtained, perform corresponding settings, and send a confirmation response to the cluster management device to the end of the cluster after the settings are completed.
  • step (A2) the candidate device determines whether it can join the cluster. It is determined whether the candidate device is already in another cluster and whether the software version in the device supports cluster management.
  • step (A2) before the candidate device feeds back to the cluster management device a confirmation response to agree to join the cluster, it is also necessary to determine whether a privileged user password has been set for the candidate device. If it is not set, directly feedback the confirmation response operation that agrees to join the cluster. ; If it has been set, then feedback to the cluster management to request an authentication operation, and then perform the authentication operation of this candidate device according to the authentication information issued by the management device. After the authentication is passed, then feedback the confirmation response operation that agrees to join the cluster Otherwise, a response to the refusal to join the cluster is fed back to the cluster management device.
  • the configuration for each member device includes the following: the number of the member device, the private IP address of the member device, the name of the member device, and the Status, status of member device operations, password for cluster management.
  • each member device joining the cluster as described in step (4) is performed using a data structure including the following fields:
  • Network type used to identify the type of network where the device is located
  • Physical address The physical address used to identify the network where the device is located.
  • the cluster management device uses stream conversion technology or address conversion technology to establish an IP between the network device in the cluster and the network management device via the cluster management device. Data channel.
  • a cluster management device for a network device includes: a cluster device manager and a member device manager, wherein: The cluster device manager includes:
  • a conversion module is used to perform network address translation on management messages of member devices;
  • a DHCP-like module is used to complete the assignment of private IP addresses of member network devices;
  • a cluster member management module is used to perform centralized management of member network devices. Management messages from outside the cluster that are targeted for member devices are forwarded to the corresponding member devices for processing through a standard network address translation module, so that the member devices process the management messages according to the usual processing procedure;
  • a topology information processing module is used to discover the topology of the network and collect information about the topology of the network within a specified number of hops in the network;
  • Member device managers include:
  • the cluster member management module is used to complete the cluster management tasks on the member device side.
  • the topology information processing module is used to complete the discovery of adjacent devices and the response / forwarding processing of the topology collection request.
  • the present invention since the present invention adopts the configuration including assigning a private IP address and identifying each member device that joins the cluster, it can be implemented on the existing IP protocol stack, which makes it relatively easy to implement and has good availability. Openness, which is convenient for the management of equipment from different manufacturers. Since the management of network equipment is achieved by using private IP addresses, precious public IP address resources can also be saved. In addition, the present invention uses standard network address translation to forward equipment management messages. It is convenient to implement by hardware, thereby improving the versatility of the present invention, greatly reducing the CPU load of the management device, and achieving centralized management of a large number of network devices. Therefore, the present invention enables effective cluster management of network devices. BRIEF DESCRIPTION OF THE DRAWINGS
  • FIG. 1 is a schematic diagram of a cluster management system networked according to the method of the present invention
  • FIG. 2 is a flowchart of a method for forming a cluster by using multiple network devices according to the present invention
  • FIG. 3 is a flowchart of adding a candidate network device to a cluster according to the method described in FIG. 2; Block diagram.
  • the core idea of the present invention is to group multiple network devices into a cluster, in which the network device set as the cluster management device is configured with a public IP address, and other network devices use a private IP address segment configured and updated by the cluster management device.
  • the device is the only external management interface of the entire cluster. In other words, when a network management device accesses or manages any network device in the cluster, a device must be established through the cluster management device.
  • the IP data channel realizes centralized management of network devices in the cluster.
  • the above IP public address can be set by the network management device outside the cluster, or by the user on the cluster management device.
  • FIG. 1 is a schematic diagram of a cluster management system networked according to the method of the present invention.
  • the IP address of the network management device station (that is, the network management device) is 69. 110. 1. 100
  • the network administrator sets a network device in the cluster as a command switch (generally a Layer 3 switch). Or a more powerful network device) and assign a public IP address 69. 110. 1. 1 (for example, through a network management device or a command-line interface of a command switch), and the cluster also includes a backup switch (usually also Layer 3 switches or more powerful network devices) and member switches 1 to 3 (generally two-layer switches or low-end network devices that implement similar functions).
  • a backup switch usually also Layer 3 switches or more powerful network devices
  • member switches 1 to 3 generally two-layer switches or low-end network devices that implement similar functions.
  • the command switch is equivalent to the management device in the cluster.
  • the management device responsible for the unified management of the entire cluster network and the establishment and maintenance of the cluster, including functions such as adding / deleting member devices, maintaining state, and collecting topology information of the cluster network, such as assigning IP private addresses and routes to member switches.
  • a backup switch or a backup cluster management device should also be set.
  • the backup switch can replace the command switch and take over the entire cluster network.
  • Centralized management After member devices join the cluster, users can remotely manage them through command switches, including configuration, query, and other operations.
  • a switch with a cluster management function can also be configured to decide whether to join the cluster, for example, a candidate switch with a cluster management function in FIG. 1.
  • IP public address In order to save IP public address resources to the greatest extent, for a cluster, you can only assign an IP public address to the command switches in the cluster. Other network devices with cluster management functions in the cluster do not temporarily assign IP public addresses. Instead, they act as command switches. When a fault occurs, the IP public address of the command switch is assigned to the backup switch that works instead, or a new IP public address can also be assigned to the backup switch that replaces the command switch. After such a switchover occurs, the specific cluster configuration update will be initiated and organized by the new command switch.
  • step 1 a network management device or a user designates a device as a cluster management device in the network through a command line.
  • the device is usually a Layer 3 switch or a network device with better performance, and then the device is correspondingly processed.
  • Configuration The configuration performed in this example includes: Cluster name: The name used to identify the management cluster;
  • Cluster enable status Used to identify whether the management cluster is in a valid state;
  • the cluster's management IP address pool A private IP address segment used to configure member devices;
  • the cluster effective retention time Used to indicate how long no member has been received After the handshake, it is considered that the contact between the member and the management device is interrupted;
  • Handshake time interval used to configure the interval for sending a handshake between the member device and the management device;
  • the role of the management device in the cluster used to identify whether the device is a management device or a member device;
  • Management IP address of the management device It is used to identify the IP address used by the management device for internal cluster communication;
  • step 2 the scope of the management cluster is first determined.
  • the network management device needs to determine the size of the network topology collection hops, and then the cluster management device initiates the topology collection process to collect the network topology structure within the specified hops in the network.
  • Information to obtain information about the managed devices that can join the management cluster including obtaining the MAC addresses and interconnect port numbers of the managed devices.
  • the network management device / command line user can decide whether to establish a device management cluster according to the topology information collected by the cluster management device and other related circumstances. For example, when there are fewer devices, the scope of establishing a management cluster needs to be re-selected. Instead of setting up a management cluster immediately. If a cluster can be established, the cluster management device is notified to start the joining process of the cluster member device.
  • the cluster management device determines candidate devices that can join the cluster in the network, and adds the determined candidate devices to the cluster to make it a member device of the cluster.
  • Each member device entering the cluster is configured including assigning private IP addresses and member numbers.
  • a device does not need to be connected to the Internet or another proprietary network. At this time, it is not necessary to comply with the requirements for application and registration of IP addresses. Address, such as using a private IP address.
  • RFC 1597 Address Allocation for Private Internet s
  • the following IP address segments are specified as private addresses: Class A address: 10. 0 0. 0 to 10. 255. 255. 255
  • Class C address 192. 168. 0. 0 to 192. 168. 255. 255
  • the above private IP address can be used for device management in the cluster.
  • the configuration described in this example includes the assignment of a private IP address and a member number for each member device that joins the cluster.
  • the configuration for each member device includes the following: Cluster name: The name of the cluster used to identify the current switch. ;
  • Cluster password A unified cluster management password is used to authenticate the management process in the cluster.
  • Member device number Used to uniquely identify the device in the cluster. This is an internal number and is used as an index for easy implementation.
  • Private IP address of the member device The network address used for member devices to communicate based on the IP network.
  • Management device IP address Used for IP-based communication between member devices and management devices in the cluster.
  • each member device that joins the cluster is identified using the following words: Segment data structure: type (2byte) Reserved (2byte) device network physical address (6byte) network type (Type): used to identify the type of network where the device is located;
  • Physical address It is used to identify the physical address of the device in the network. It is represented by characters.
  • step 4 above the process of adding candidate network devices to the cluster is implemented using the following steps, referring to FIG. 3:
  • step 11 the cluster management device delivers the candidate network devices that can join the cluster.
  • step 12 the candidate device judges whether it can be added according to its own situation.
  • the cluster For example, you can decide whether to join the cluster according to whether the candidate device is in another cluster and whether the software version in this device supports cluster management. If you cannot join the cluster, end the process of joining the cluster, and feedback to the cluster management device to refuse to join. The response of the cluster; otherwise, it is judged in step 13 whether the candidate device has set a privileged user password. If it is not set, then no authentication operation is required, and a response to the confirmation of agreeing to join the cluster is performed directly. Password, then to the management device in step 14 After the authentication is passed, a confirmation response operation for agreeing to join the cluster is fed back at step 15, otherwise the rejection information is fed back to the cluster management device, and the operation for joining the cluster ends.
  • the authentication operation of the above device refers to the following process: First, the candidate device returns a message containing a random number (chal lenge) for authentication to the management network device. After receiving the message, the cluster management device uses the random number Encrypt the authentication information to be transmitted.
  • the authentication information includes the candidate network device and the authentication password (which may be a cluster management password issued by the cluster network device), and then uses the authentication information to encapsulate a response message. After the message is ready, the message is sent to the corresponding candidate device. After the candidate device confirms the identity of the management network device after authentication, it returns a confirmation message to the management network device that it agrees to join.
  • the cluster management device After receiving the response from the candidate network device that agrees to join the cluster, the cluster management device allocates a cluster member identification number, a private IP used for management, and some other configuration information to the candidate network device. Using this information, Add the password (which can be encrypted), encapsulate it into a configuration message and send it to the candidate device. After receiving the message, the candidate device parses out the password, configuration information such as the cluster management serial number, and management private IP. After recording these configuration information issued by the cluster management device, the candidate network device first changes its role to a member switch, and then returns a confirmation of joining completion to the cluster management device. The cluster management device receives a confirmation of joining the candidate network device. After that, the candidate network device is identified as a cluster member, and thus the joining process of the candidate device ends.
  • FIG. 4 is a structural block diagram of a device embodiment according to the present invention.
  • the cluster management device for a network device shown in the figure includes a cluster device manager A1 and a cluster device manager set on a command switch side. Set the member device manager A2 on the member switch side, where:
  • the cluster device manager A1 is set in the cluster management device and is used to implement cluster management of member devices, including:
  • a conversion module A11 is used to perform network address translation on management messages of member devices;
  • a class DHCP module A12 is used to complete the assignment of private IP addresses of member network devices;
  • a cluster member management module A13 is used to Centralized management. Management messages from outside the cluster that are targeted for member devices are forwarded to the corresponding member devices for processing through a standard network address translation module, so that the member devices process the management message according to the usual processing procedure;
  • the topology information processing module A14 is used for discovering the topology of the network, and collects the topology information of the network within a specified number of hops in the network;
  • the member device manager A2 is set in the managed member device and is used to complete the management of the member device side in the cluster management, including:
  • the cluster member management module A21 is used to complete the cluster management tasks on the member device side; the topology information processing module A22 is used to complete the discovery / recovery processing of the neighboring devices and the topology collection request.
  • the topology information processing module A14 collects the network topology structure information within a specified number of hops in the network through the topology information processing module A22 on the candidate device side, and transmits this information to the cluster members.
  • the management module A13 the cluster member management module A13 issues a cluster join request to the cluster member management module A21 of the candidate device that can join the cluster, and the cluster member management module A 21 judges whether it can join the cluster according to its own situation, which is about to join or reject The added response is fed back to the cluster Member management module A13;
  • the cluster member management module A13 receives a response from the candidate device to confirm joining, the DHCP-like module A12 allocates the private IP addresses of the member network devices, and the assigned private IP address is sent through the cluster member management module A13
  • the cluster member management module A21 is given to the candidate device, and configuration information such as the member number, handshake interval, and state holding time is also sent to the cluster member management module A21.
  • the cluster member management module A21 uses the information to set the device accordingly. And send a confirmation response to the cluster management device to the end of joining the cluster after the setting is completed.
  • the management packets from the outside of the cluster that are targeted for the member device undergo standard network address translation through the conversion module A 11 and are then forwarded to the corresponding member device by the cluster member management module A 13
  • the processed cluster member management module A21 enables a member device to process the management message according to a normal processing procedure.
  • FIG. 4 A preferred embodiment for implementing cluster management in a cluster shown in FIG. 1 is described below with reference to FIG. 4.
  • a conversion module A11, a DHCP-like module A12, a cluster member management module A13, and a topology information processing module A14 are set in the command switch side, and a cluster member management module A21 and Topology information processing module A 22.
  • the following describes the functions of each module and the flow of collaborative work.
  • the topology information processing module in FIG. 4 is used to obtain network topology structure information and network device information. Specifically, the topology information processing module implements a device discovery function, and collects and processes the obtained device information (such as the device type and software and hardware). Version, etc.) to get the topology information of the network.
  • the obtained device information such as the device type and software and hardware. Version, etc.
  • the topology information processing modules in the command switch and member switches When powered on, the topology information processing modules in the command switch and member switches periodically send topology information containing their own device information to other nearby devices. Processes messages, and at the same time continuously receives topological information processing messages containing device information from neighboring neighboring devices and processes them accordingly. For example, if a message is received from a new device that has not been recorded, the neighboring device A new entry is added to the information buffer area to store the information of the new device, and the new entry is timed. If no new message of the recorded device is received after the timeout, the corresponding entry is aged. If the information contained in the message is different from the previously stored device information of the recorded device, the previously stored device information is updated and the timekeeping is restarted.
  • the topology information processing module periodically sends device information, and the topology information processing message can be implemented by a timer. For example, when the timer reaches a set value, the setup information processing topology information processing message is sent, and Reset the timer when it is finished to restart timing.
  • the sending of device information topology information processing messages does not depend on the Standard Spanning Tree Protocol (STP) tree. As long as there is a physical connection between network devices, the topology information processing messages will pass through the ports blocked by STP.
  • STP Spanning Tree Protocol
  • the basic information processing message is not allowed to be forwarded, that is, the topological information processing message is valid only within the range of 1 hop, so that the device can always accurately know the status and topology of the surrounding equipment.
  • a topology information processing module can be triggered to collect the information.
  • the collection process can take the following steps: First, collect the initiating device (the device is not necessarily a command switch, if necessary, The member device may also be the initiator of the collection.) The collection range is set, and then the topology information processing module of the device sends a collection request message to neighboring devices. The topology information processing module of the neighboring device receives the request. After the request message, the response message containing the device information is sent back to the initiating device and the hop value is decreased by 1. If the hop value after the subtraction is not equal to 0, the message is forwarded to the adjacent network device, otherwise the forwarding is stopped.
  • the sending or forwarding of the request message by the topology information processing module should depend on the STP tree.
  • the topology information processing module of the collection initiating device processes the received response message containing the device information to obtain network topology structure information within the collection range.
  • the collection and processing of the topology information processing module should be performed at the data link layer.
  • the command switch can start the cluster establishment process. Specifically, the cluster member management module on the command switch side first obtains information about relevant candidate switches (that is, network devices that have joined but not yet joined the cluster) from the topology information processing module, and then works in cooperation with the cluster member management module on the candidate switch side to: Automatic mode (that is, all candidate switches are automatically added to the cluster) or manual mode (that is, the user specifies the candidate switches that are added to the cluster from the candidate device list) to add the candidate switches to the cluster. If the user already knows the information of the candidate network device, such as the MAC address of the known device, the topology structure information collection step of the topology information processing module can also be omitted, and the device can be directly added to the cluster by manual operation.
  • relevant candidate switches that is, network devices that have joined but not yet joined the cluster
  • the cluster member management module on the candidate switch side works in cooperation with the cluster member management module on the candidate switch side to: Automatic mode (that is, all candidate switches are automatically added to the cluster) or manual mode (that is, the
  • the member management module on the switch side is instructed to treat the managed device (candidate member switch) as a special user, and a DHCP-like module is used to dynamically assign IP private addresses and configure related routes for the candidate member devices. Therefore, an IP data channel between a command switch and a member switch can be established in the cluster. After you configure private IP addresses for member devices in the cluster, you must also establish IP data channels for communication inside and outside the cluster. To this end, the cluster member management module on the command switch side also sets a conversion policy for the private IP addresses of member devices through the conversion module during the member joining process. Establish IP data channels for network management device stations outside the cluster to access member devices in the cluster or for member devices to access network management device stations.
  • the conversion method can use stream conversion technology or network address translation technology.
  • an IP data channel through the command switch can be established between the network management device and the member switches, so users can use a remote public network management device station to use a public IP address to Unified management of multiple devices.
  • the member switch and the command switch can start the handshake process through their respective cluster member management modules to maintain the state of the cluster.
  • the backup function of the command switch can also be implemented by means of a backup module (not shown), that is, by designating the backup switch, when the command switch fails, the management of the command switch through the specified policy
  • the function is transferred to a designated backup command switch, thereby avoiding a system crash due to a single point of failure of the command switch.

Landscapes

  • Engineering & Computer Science (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Signal Processing (AREA)
  • Data Exchanges In Wide-Area Networks (AREA)
  • Small-Scale Networks (AREA)

Description

一种网络设备的集群管理方法及装置 技术领域
本发明涉及网络设备管理方法，特别涉及一种以集群方式管理传 输控制 协议 /互联网协议（TCP/IP ) 网絡设备的方法和装置。
背景技术
随着网络规模和范围的急剧扩大 ,人们各项活动对网络服务的依 赖程度日益增强， 网络所承担的任务也越来越重， 无论从网絡用户的 角度，还是从网络管理者的角度来看，有效地进行网络管理都成为曰 益迫切的要求。
在目前的网络环境下，对网络设备的管理主要采用三种方式， 以 下分别加以描述。
第一种方式是将网络管理设备通过串口直接连接至被管理的网 络设备以对网络设备进行配置管理。这种管理方式的主要优点是管理 手段相对比较简单， 无需其它附加设备， 而且由于通过串口直接连接 被管理网络设备， 因此在配置网絡设备时不需要分配相应的管理 IP 地址， 节省了 IP资源。 但是由于网络设备可能分布在较大的范围之 内， 特别是在电信网络中， 数据设备的数量非常大并且布放点十分分 散， 因此这种管理方式将不可避免地带来维护困难和工作量大的难 题。
第二种方式基于筒单网络管理协议（S丽 P)或者远程登陆协议 ( TELNET )终端仿真协议， 可从远端管理 TCP/IP网络设备。 与第一 种方式相比， 由于网络管理程序是在标准的 IP协议栈上实现的， 因 此实现方式比较筒单，工作量较小，而且与其它设备的相关性比较小， 管理员可以在远端通过网络管理终端或者 TELNET仿真终端实现被管 理网络设备的远程管理，但是主要缺点是必须为每个被管理网络设备 都配置一个公共网 IP地址， 这将造成 IP地址资源的浪费， 特别是在 设备数量巨大的电信网络中， 这种缺陷显得尤为突出。
最后一种方式是在第二种方式的基础上提出的，核心思想是将网 络设备制造厂商为设备统一分配的序列号作为设备地址来实现网络 设备的管理， 因此节约了 IP地址资源， 但是这种方式只适用于特定 厂商的设备， 网络管理程序运行在私有协议栈上， 因此管理协议几乎 没有可开放性， 无法对各家厂商制造的网络设备实现统一管理。 发明内容
本发明的目的是提供一种网络设备的集群管理方法，它仅利用少 量的 IP资源即可实现对大量 TCP/IP网络设备的集中管理，从而减少 了维护工作量， 方便了网络管理。
本发明的另一个目的是提供一种网络设备的集群管理装置。 为达到上述目的， 在本发明提供的网络设备集群管理方法中， 多 个网络设备组成一个集群，其中至少一个网络设备被设定为集群管理 设备并配置公有 IP地址， 其它网络设备由所述集群管理设备配置和 更新私有 IP地址和路由， 所述网络管理设备按照下列步驟管理集群 内的网络设备： (A)所述集群管理设备在集群内的网络设备与所述网络管理设备 之间建立经由该集群管理设备的 IP数据通道； 以及
(B)所述网络管理设备经所述 IP数据通道对集群内的网络设备 进行管理。
比较好的是， 在上述网络设备管理方法中， 所述集群管理设备根 据集群内网络拓朴结构信息和网络设备的设备信息为其它网络设备 配置和更新私有 IP地址和路由。 更好的是， 所述集群管理设备为其 它网络设备动态配置私有 IP地址。
比较好的是， 在上述网絡设备管理方法中， 所述集群内包含多个 所述集群管理设备，其中一个集群管理设备负责管理集群内网络设备 私有 IP地址和路由的配置和更新以及网络管理设备与集群内网絡设 备之间的通信， 当该集群管理设备出现故障时， 可按照预定的策略指 定其它集群管理设备中的一个负责管理集群内网络设备私有 IP地址 和路由的配置和更新以及网络管理设备与集群内网络设备之间的通 信。
本发明中， 采用下述步骤将多个网络设备组成一个集群：
(1)网络管理设备在网络中指定一台设备作为集群的管理设备， 并对该台设备进行相应的配置；
(2)集群管理设备启动拓朴收集过程收集网络中规定跳数内的网 络拓朴结构信息；
(3)网络管理设备根据从集群管理设备取来的拓朴结构信息， 指 定拓朴结构中需要加入集群的候选设备，并通知集群管理设备启动集 群成员设备的加入过程；
(4)集群管理设备将指定的候选设备加入集群， 使其进行相应的 配置， 成为集群的成员设备；
(5)集群建立后， 通过集群管理设备对集群中的成员设备进行管 理， 来自集群外部、 目标为成员设备的管理报文， 在集群管理设备上 经过标准的网絡地址转换过程 ( NAT )转发到相应的成员设备处理， 成员设备按通常的处理过程处理该管理报文。
步骤（1 ) 中所述对集群管理设备进行相应的配置， 包括在该设 备上配置下述内容： 集群名称、 集群的使能状态、 集群的管理 IP地 址池、 集群状态有效保留时间、 成员设备的握手时间间隔、 本管理设 备在集群中的角色、 本管理设备的 IP地址。
步骤（4 ) 中所述将候选网络设备加入集群的过程包括： ( A1 )集群管理设备向可以加入集群的候选网络设备下发集群加 入请求；
( A2 )候选设备根据自己的状况判断是否可以加入集群， 如果不 可以， 则反馈拒绝加入响应， 结束加入集群的过程； 否则向集群管理 设备反馈同意加入集群的确认响应；
( A3 )集群管理设备收到候选设备的确认响应后， 如果候选设备 同意加入， 就向所述候选设备发送包括私有 IP地址、 成员编号、 握 手间隔、状态保持时间等的配置信息， 候选设备收到该信息后进行相 应的设置，并在设置结束后向集群管理设备发出加入集群结束的确认 响应。 步驟（ A2 )所述候选设备判断自己是否可以加入集群， 是本候选 设备是否已经在其它集群中以及本设备中的软件版本是否支持集群 管理决定的。
步骤（A2 )中， 在候选设备向集群管理设备反馈同意加入集群的 确认响应前， 还要判断本候选设备是否已设定特权用户密码， 如果没 有设定， 直接反馈同意加入集群的确认响应操作； 如果已经设定， 则 向集群管理反馈要求鉴权操作，然后根据管理设备下发的鉴权信息进 行本候选设备的鉴权操作， 当鉴权通过后， 再反馈同意加入集群的确 认响应操作， 否则向集群管理设备反馈拒绝加入集群的响应。
步驟（4 ) 中所述为加入集群的每个成员设备进行必需的配置， 对每个成员设备配置包括下述内容： 成员设备的编号、成员设备的私 有 IP地址、 成员设备名称、 成员设备的状态、 成员设备操作的状态、 集群管理的密码。
步骤（4 ) 中所述为加入集群的每个成员设备进行标识采用包括 下述字段的数据结构进行：
网络类型： 用于标识设备所在的网络类型；
物理地址： 用于标识设备所在网络的物理地址。
在上述网络设备管理方法中， 在步驟（1)中， 所述集群管理设备 利用流转换技术或地址转换技术在集群内的网络设备与所述网络管 理设备之间建立经由该集群管理设备的 IP数据通道。
本发明提供的网络设备的集群管理装置， 包括： 集群设备管理器 和成员设备管理器， 其中： 集群设备管理器包括：
转换模块， 用于对成员设备的管理报文进行网络地址转换； 类 DHCP模块， 用于完成成员网络设备的私有 IP地址的分配； 集群成员管理模块， 用于对成员网络设备的集中管理，对来自集 群外部、 目标为成员设备的管理报文， 在经过标准的网络地址转换模 块转发到相应的成员设备处理，使成员设备按通常的处理过程处理该 管理报文；
拓朴信息处理模块， 用于网络拓朴结构的发现， 收集网络中规定 跳数内的网络拓朴结构信息；
成员设备管理器包括：
集群成员管理模块， 用于完成成员设备端的机群管理任务； 拓朴信息处理模块，用于完成邻接设备的发现和拓朴收集请求的 响应 /转发处理。
由上可见，由于本发明采用对加入集群的每个成员设备进行包括 分配私有 IP地址和进行标识的配置，可以在现有的 IP协议栈上实现， 这样使得实现相对容易， 而且具有良好的可开放性， 便于不同厂商设 备的管理； 由于利用私有 IP地址实现对网络设备的管理， 还可以节 约宝贵的公有 IP地址资源； 另外， 由于本发明采用标准的网络地址 转换进行设备管理报文的转发，便于通过硬件实现， 因此提高了本发 明的通用性， 能够大大减轻管理设备的 CPU负担， 实现大量网络设备 的集中管理， 因此， 采用本发明， 能够对网络设备进行有效的集群管 理。 附图说明
图 1为按照本发明方法组网的集群管理系统的示意图；
图 2为本发明采用的将多个网络设备组成一个集群的方法流程图； 图 3是图 2所述方法采用的将候选网络设备加入集群的流程图； 图 4为本发明所述装置实施例的结构框图。
具体实施方式
本发明的核心思想是将多个网络设备组成一个集群，其中设定为 集群管理设备的网络设备配置公有 IP地址， 其它网络设备使用一个 由集群管理设备配置和更新的私有 IP地址段， 集群管理设备是整个 集群对外唯一的管理接口， 换句话说， 当网络管理设备对集群内任一 网络设备进行访问或管理时， 都必须建立一条经由集群管理设备的
IP数据通道， 由此实现了对集群内网络设备的集中管理。 上述 IP公 有地址可以由集群外部的网络管理设备设置，也可以由用户在集群管 理设备上设置。
图 1为按照本发明方法组网的集群管理系统的示意图。如图 1所 示， 网络管理设备站（即网络管理设备）的 IP地址为 69. 110. 1. 100, 网络管理员将集群内的一台网絡设备设定为命令交换机 (一般为三层 交换机或性能更强的网絡设备）并分配 IP公有地址 69. 110. 1. 1 (例如 通过网络管理设备或者命令交换机的命令行接口进行设定）， 集群内 还包含一台备份交换机（一般也为三层交换机或性能更强的网络设备) 和成员交换机 1 ~ 3 (一般为两层交换机或实现类似功能的低端网络设 备）。 在图 1所示的集群中， 命令交换机相当于集群内的管理设备， 负责对整个集群网络的统一管理和集群的建立和维护，包括成员设备 的加入 /删除、 状态维护以及集群网络的拓朴结构信息收集等功能， 例如为成员交换机分配 IP私有地址和路由等。 为了提高网絡管理的 可靠性， 避免单点故障对集群管理功能的影响， 还应该设置备份交换 机或备份的集群管理设备， 当命令交换机出现异常时，备份交换机可 替代命令交换机，接管对整个集群网络的集中管理。 成员设备加入集 群后， 用户即可通过命令交换机来对它们进行远程管理， 包括配置、 查询等各种操作。 此外， 具有集群管理功能的交换机也可以通过配置 来决定是否加入集群， 例如图 1中具有集群管理功能的候选交换机。
为了最大程度地节省 IP公有地址资源， 对于一个集群， 可以只 为集群中的命令交换机分配一个 IP公有地址， 集群内其它具有集群 管理功能的网絡设备暂时不分配 IP公有地址， 而是当命令交换机出 现故障时， 将命令交换机的 IP公有地址分配给替代工作的备份交换 机， 或者也可以为替代命令交换机的备份交换机分配新的 IP公有地 址。 在发生这种切换后， 具体的集群配置更新将由新的命令交换机启 动和组织完成。
本发明在通过使用标准协议栈及私有 IP地址段实现对网络设备 的集中管理的过程中 , 采用下述步驟将多个网络设备组成一个集群， 具体实施过程参考图 2。 首先在步骤 1 , 由网络管理设备或者用户通 过命令行在网络中指定一台设备作为集群的管理设备，该设备通常为 3层交换机或性能较好的网络设备，然后对该台设备进行相应的配置， 在本例中所进行的配置内容包括： 集群名称： 用于标识本管理集群的名称；
集群的使能状态： 用于标识本管理集群是否处于有效状态； 集群的管理 IP地址池： 用于给成员设备配置的私有 IP地址段； 集群有效保留时间： 用于表示在多久没有收到成员握手后， 认为 成员与管理设备联系中断；
握手时间间隔： 用于配置成员设备与管理设备间发送握手的间隔； 本管理设备在集群中的角色：用于标识设备是管理设备还是成员 设备；
管理设备的管理 IP地址： 用于标识管理设备用于集群内部通讯 的 IP地址；
在步骤 2, 首先确定本管理集群的范围， 为此网络管理设备需要 确定网络拓朴收集跳数的大小，然后由集群管理设备启动拓朴收集过 程收集网絡中规定跳数内的网络拓朴结构信息，以获取可以加入管理 集群的被管理设备的信息，包括获取被管理设备的 MAC地址和互连端 口号。
在步骤 3, 网络管理设备 /命令行用户可以根据集群管理设备收 集来的拓朴信息， 和其他相关情况决定是否建立设备管理集群， 例如 当设备较少时需要重新选定建立管理集群的范围，而不是立即进行管 理集群的建立。 如果可以建立集群， 则通知集群管理设备启动集群成 员设备的加入过程。
在步骤 4, 集群管理设备确定网络中可以加入集群的候选设备， 并将确定的候选设备加入集群，使其成为集群的成员设备， 同时对加 入集群的每个成员设备进行包括分配私有 IP 地址和成员编号等配 置。 （在具体的通信网络中， 在有些情况下， 一个设备并不需要连接 到互联网或另一个专有的网络上， 此时无须遵守对 IP地址进行申请 和登记的规定， 该设备可以使用任何的地址， 如使用私有 IP地址。 在 RFC 1597 (专用国际互联网络地址分配， Address Al locat ion for Private Internet s ) 中， 规定以下 IP地址段是用作私用地址的： A类地址: 10. 0. 0. 0 到 10. 255. 255. 255
B类地址: 172. 16. 0. 0 到 172. 31. 255. 255. 255
C类地址: 192. 168. 0. 0 到 192. 168. 255. 255
因此可以使用上述私有 IP地址用于集群中的设备管理。
本例中所述为加入集群的每个成员设备进行包括分配私有 IP地 址和成员编号的配置， 对于每个成员设备的配置包括下述内容： 集群名称： 用于标识当前交换机所处的集群名称；
集群口令： 统一的集群管理口令， 用于集群内管理过程的鉴权。 成员设备的编号： 用于在集群中唯一标识该设备。 这是一个内部 编号， 便于实现时作为索引。
成员设备的私有 IP地址： 用于成员设备基于 IP网絡通讯的网 络地址。
管理设备的 IP地址： 用于集群内成员设备和管理设备之间基于 IP的通信。
成员设备名称： 用于标识该设备的名称。
在本例中，为加入集群的每个成员设备进行标识采用包括下述字 段的数据结构： type (2byte) Reserved ( 2byte ) 设备网络物理地址（ 6byte ) 网络类型 （Type ): 用于标识设备所在的网絡类型；
物理地址： 用于标识设备在网络中的物理地址， 用字符表示。
在上述结构中共 10字节， 其中网络类字段 2个字节， 物理地址 字段 6个字节， 保留（Reserved ) 2个字节另做他用。 采用上述结构， 不再需要厂商自己定义设备的标识方法， 有利于维护设备的唯一性， 而且是借用设备物理地址的唯一性来保证。同时这样标识不局限在某 种物理网络之上， 例如当 TYPE为 G时， 设备网络物理地址表示为以 太网地址。 TYPE可以随设备所在的物理网络不同而进行扩充。
在上述步驟 4中，所述将候选网络设备加入集群的过程采用下述 步骤实现， 参考图 3:
在步骤 11， 集群管理设备向可以加入集群的候选网络设备下发
'集群加入请求。 在步骤 12候选设备根据自己的状况判断是否可以加
入集群，例如可以根据本候选设备是否在其它集群中以及本设备中的 软件版本是否支持集群管理决定是否加入集群； 如果不可以加入集 群， 结束加入集群的操作过程， 向集群管理设备反馈拒绝加入集群的 响应； 否则在步骤 13判断本候选设备是否已设定特权用户密码， 如 果没有设定， 则不需要进行鉴权操作， 直接进行反馈同意加入集群的 确认响应操作， 如果本设备已经设定密码， 则在步骤 14对管理设备 的身份进行鉴权操作， 当鉴权通过后， 在步骤 15反馈同意加入集群 的确认响应操作， 否则向集群管理设备反馈拒绝信息， 结束加入集群 的操作。
上述设备的鉴权操作参考下述过程：首先候选设备向管理网络设 备返回一个包含一个用于认证的随机数（chal lenge ) 的报文， 集群 管理设备在收到报文后， 利用该随机数， 对要传递的鉴权信息进行加 密， 鉴权信息包括有候选网络设备、 认证口令（可以是集群网络设备 的下发的集群管理口令）， 然后利用该鉴权信息封装一个响应报文， 封装好报文后， 将该报文下发到相应的候选设备，候选设备在经过鉴 权后确认该管理网络设备的身份后，向管理网絡设备返回一个同意加 入的确认4艮文。
在步驟 16， 集群管理设备在收到候选网络设备返回的同意加入 集群的响应后， 为该候选网络设备分配集群成员标识号、 管理使用的 私有 IP以及一些其他配置信息等， 利用这些信息， 再加上口令（可 以经过加密处理）， 封装成一个配置报文下发到候选设备， 候选设备 在收到该报文后， 解析出包含的口令， 以及集群管理序号和管理私有 IP 等配置信息， 记录下集群管理设备下发的这些配置信息后， 该候 选网络设备先将自己的角色改变成成员交换机，然后向集群管理设备 返回一个加入结束确认，集群管理设备在收到候选网絡设备的加入确 认后 ,标识该候选网络设备为集群成员，至此候选设备加入过程结束。
图 4为本发明所述装置实施例的结构框图。图中所示的网络设备 的集群管理装置， 包括设置在命令交换机端的集群设备管理器 A1和 设置在成员交换机端的成员设备管理器 A2 , 其中：
集群设备管理器 A1设置在集群管理设备中， 用于实现成员设备 的集群管理， 包括：
转换模块 Al l , 用于对成员设备的管理报文进行网络地址转换； 类 DHCP模块 A12 ,用于完成成员网络设备的私有 IP地址的分配； 集群成员管理模块 A13 , 用于对成员网络设备的集中管理， 对来 自集群外部、 目标为成员设备的管理报文，在经过标准的网络地址转 换模块转发到相应的成员设备处理，使成员设备按通常的处理过程处 理该管理艮文；
拓朴信息处理模块 A14 , 用于网络拓朴结构的发现， 收集网络中 规定跳数内的网络拓朴结构信息；
成员设备管理器 A2设置在被管理的成员设备中， 用于完成集群 管理中的成员设备侧的管理， 包括：
集群成员管理模块 A21 , 用于完成成员设备端的集群管理任务； 拓朴信息处理模块 A22 , 用于完成邻接设备的发现和拓朴收集请 求的响应 /转发处理。
使用上述装置进行网络设备的集群管理时 ,首先拓朴信息处理模 块 A14通过候选设备侧的拓朴信息处理模块 A22收集网络中规定跳数 内的网络拓朴结构信息， 将该信息传给集群成员管理模块 A13 , 由集 群成员管理模块 A13 向可以加入集群的候选设备的集群成员管理模 块 A21下发集群加入请求， 集群成员管理模块 A 21根据自己的状况 判断是否可以加入集群，即将可以加入或拒绝加入的响应反馈给集群 成员管理模块 A13; 当集群成员管理模块 A13收到候选设备的确认加 入的响应后，由类 DHCP模块 A12进行成员网络设备的私有 IP地址的 分配， 分配的私有 IP地址通过集群成员管理模块 A13发给候选设备 的集群成员管理模块 A21 , 同时发给集群成员管理模块 A21的还有成 员编号、 握手间隔、 状态保持时间等的配置信息， 集群成员管理模块 A21使用该信息后对设备进行相应的设置， 并在设置结束后向集群管 理设备发出加入集群结束的确认响应。当候选设备成为集群的成员设 备后， 对来自集群外部、 目标为成员设备的管理报文， 在经过转换模 块 A 11进行标准的网络地址转换， 然后经集群成员管理模块 A13转 发到相应的成员设备处理的集群成员管理模块 A21 , 使成员设备按通 常的处理过程处理该管理报文。
以下借助图 4描述在图 1所示,集群内实现集群管理的一个较佳实 施例。 如图 4所示， 在命令交换机端内设置了转换模块 Al l、 类 DHCP 模块 A12、 集群成员管理模块 A13和拓朴信息处理模块 A14 , 而在每 个成员交换机内设置集群成员管理模块 A21 和拓朴信息处理模块 A 22。 以下分别描述各个模块的功能以及协同工作的流程。
图 4 中的拓朴信息处理模块用于获取网络拓朴结构信息和网络 设备信息， 具体说拓朴信息处理模块实现设备的发现功能， 通过收集 处理获取的设备信息（例如设备的类型和软硬件版本等）得到网络的 拓朴结构信息。
当上电后，命令交换机和成员交换机内的拓朴信息处理模块周期 性地向周围直接相邻的其它设备发送包含自己设备信息的拓朴信息 处理报文，与此同时也不断接收来自周围邻接设备的包含设备信息的 拓朴信息处理报文并作相应处理，例如如果接收到的是未记录过的新 设备的报文，则在邻接设备信息緩存区内添加新的表项以存储该新设 备的信息， 并对新表项开始记时； 如果在超时后仍未收到已记录设备 的新报文， 则老化相应表项； 如果收到报文的内容所含信息与已记录 设备先前存储的设备信息不同， 则更新原先存储的设备信息， 并重新 开始记时。 拓朴信息处理模块的周期性发送设备信息， 拓朴信息处理 报文可以用定时器实现，例如当定时器到达设定值时开始进行设^言 息拓朴信息处理 ^艮文的发送， 并在完成后复位定时器以重新开始计 时。 此外，设备信息拓朴信息处理报文的发送不依赖于标准生成树协 议（STP )树， 只要网络设备间存在物理连接即可， 拓朴信息处理报 文将透过 STP阻塞的端口， 但是拓朴信息处理报文不允许被转发， 即 拓朴信息处理报文只在 1跳的范围内有效，这样就保证了设备始终能 够准确知道自己周围设备的状态和拓朴结构。
当需要获取网络拓朴结构信息或者要建立集群时，可以触发拓朴 信息处理模块来收集信息， 该收集过程可以采用以下步骤： 首先在收 集发起设备 (该设备不一定是命令交换机， 必要时， 成员设备上也可 以是发起收集方）设定收集范围， 然后由该设备的拓朴信息处理模块 向周围邻接的设备发送收集请求报文，周围的邻接设备的拓朴信息处 理模块在接收到该请求报文后，向发起设备回送包含设备信息的响应 报文并使跳数值减 1， 如果减 1后的跳数值不等于 0， 则将报文向相 邻的网络设备转发， 否则停止转发。 为了避免出现因错误信息重发和 环路而造成的泛洪现象，拓朴信息处理模块对请求报文的发送或者转 发请求报文应依赖于 STP树。收集发起设备的拓朴信息处理模块将接 收到的包含设备信息的响应报文处理后即得到收集范围内的网络拓 朴结构信息。 为了保证网络拓朴结构信息的完整性和可用性， 拓朴信 息处理模块的收集处理过程应该在数据链路层上进行。
在利用拓朴信息处理模块获得网络的拓朴结构信息后，命令交换 机即可开始集群的建立过程。 具体而言， 命令交换机端的集群成员管 理模块首先从拓朴信息处理模块获得相关的候选交换机（即可加入但 尚未加入集群的网絡设备)信息， 然后与候选交换机端的集群成员管 理模块协同工作， 以自动方式（即自动将所有的候选交换机都加入到 集群中）或者手动方式（即由用户从候选设备列表中指定加入到集群 中的候选交换机)使候选交换机加入到集群中。 如果用户已经知道候 选网络设备的信息， 例如已知设备的 MAC地址， 则也可省略拓朴信息 处理模块的拓朴结构信息收集步骤，直接通过手动操作将该设备加入 到集群中。
在将候选交换机加入集群的过程中，命令交换机端的成员管理模 块将被管理设备 (候选的成员交换机)视为一个特殊用户，利用类 DHCP 模块为候选的成员设备动态分配 IP私有地址并配置相关路由， 从而 可在集群内建立命令交换机与成员交换机之间的 IP数据通道。 在为 集群内的成员设备配置私有 IP地址后， 还必须为集群内外的通信建 立 IP数据通道。 为此， 命令交换机端的集群成员管理模块在成员加 入过程中还通过转换模块设定对成员设备 IP私有地址的转换策略， 为集群外部的网络管理设备站访问集群内的成员设备或者成员设备 访问网络管理设备站建立 IP数据通道。 转换方式可以采用流转换技 术或者网络地址转换技术。 总之， 经过上述的配置后， 即可在网络管 理设备与成员交换机之间建立一条经由命令交换机的 IP数据通道， 因此用户就可通过远程的网络管理设备站， 利用一个 IP公有地址对 集群内的多个设备实现统一管理。
在将成员交换机加入集群后，成员交换机与命令交换机即可通过 各自的集群成员管理模块开始握手过程以维护集群的状态。
除了上述基本功能模块外，还可借助备份模块（未画出）实现命令 交换机的备份功能， 即通过对备份交换机的指定，使得在命令交换机 发生故障时 ,·通过指定的策略将命令交换机的管理功能转移至指定的 某台备份命令交换机上，从而避免了由于命令交换机的单点故障而导 致的系统崩溃。

Claims

权 利 要 求
1.一种网络设备的管理方法， 其特征在于， 多个网格设备组成一 个集群， 其中至少一个网络设备被设定为集群管理设备并配置公有
IP地址， 其它网络设备由所述集群管理设备配置和更新私有 IP地址 和路由， 所述网络管理设备按照下列步骤管理集群内的网络设备：
(A)所述集群管理设备在集群内的网络设备与所述网絡管理设备 之间建立经由该集群管理设备的 IP数据通道； 以及
(B)所述网络管理设备经所述 IP数据通道对集群内的网络设备 进行管理。
2.如权利要求 1所述的方法， 其特征在于， 所述集群管理设备根 据集群内网絡拓朴结构信息和网络设备的设备信息为其它网络设备 配置和更新私有 IP地址和路由。
3.如权利要求 2所述的方法， 其特征在于， 所述集群管理设备为 其它网络设备动态配置私有 IP地址。
4.如权利要求 1 - 3中任意一项所述的方法， 其特征在于， 所述 集群内包含多个所述集群管理设备，其中一个集群管理设备负责管理 集群内网络设备私有 IP地址和路由的配置和更新以及网络管理设备 与集群内网络设备之间的通信， 当该集群管理设备出现故障时， 可按 照预定的策略指定其它集群管理设备中的一个负责管理集群内网络 设备私有 IP地址和路由的配置和更新以及网络管理设备与集群内网 络设备之间的通信。
5.如权利要求 4所述的方法， 其特征在于， 在步骤（A)中， 所述 集群管理设备利用流转换技术在集群内的网络设备与所述网络管理 设备之间建立经由该集群管理设备的 IP数据通道。
6.如权利要求 4所述的方法， 其特征在于， 在步骤 (A)中， 所述 集群管理设备利用网络地址转换技术在集群内的网络设备与所述网 络管理设备之间建立经由该集群管理设备的 IP数据通道。
7. 如权利要求 4所述的方法， 其特征在于， 采用下述步骤将多 个网络设备组成一个集群：
( 1 )网络管理设备在网络中指定一台设备作为集群的管理设备， 并对该台设备进行相应的配置；
(2)集群管理设备启动拓朴收集过程收集网络中规定跳数内的网 絡拓朴结构信息；
(3)网络管理设备根据从集群管理设备取来的拓朴结构信息， 指 定拓朴结构中需要加入集群的候选设备，并通知集群管理设备启动集 群成员设备的加入过程；
(4)集群管理设备将指定的候选设备加入集群， 使其进行相应的 配置， 成为集群的成员设备；
(5)集群建立后， 通过集群管理设备对集群中的成员设备进行管 理， 来自集群外部、 目标为成员设备的管理报文， 在集群管理设备上 经过标准的网络地址转换过程 ( NAT )转发到相应的成员设备处理 , 成员设备按通常的处理过程处理该管理报文。
8. 如权利要求 7所述的方法， 其特征在于， 步骤（1 )中所述对 集群管理设备进行相应的配置， 包括在该设备上配置下述内容： 集群 名称、 集群的使能状态、 集群的管理 IP地址池、 集群状态有效保留 时间、 成员设备的握手时间间隔、 本管理设备在集群中的角色、 本管 理设备的 IP地址。
'
9、 如权利要求 7所述的方法， 其特征在于， 步骤（4 ) 中所述将 候选网络设备加入集群的过程包括：
( A1 )集群管理设备向可以加入集群的候选网络设备下发集群加 入请求；
( A2 )候选设备根据自己的状况判断是否可以加入集群， 如果不 可以， 则反馈拒绝加入响应， 结束加入集群的过程； 否则向集群管理 设备反馈同意加入集群的确认响应；
( A3 )集群管理设备收到候选设备的确认响应后， 如果候选设备 同意加入， 就向所述候选设备发送包括私有 IP地址、 成员编号、 握 手间隔、状态保持时间等的配置信息，候选设备收到该信息后进行相 应的设置，并在设置结束后向集群管理设备发出加入集群结束的确认 响应。
10、 如权利要求 9所述的方法， 其特征在于， 步骤 （A2 )所述 候选设备判断自己是否可以加入集群，是本候选设备是否已经在其它 集群中以及本设备中的软件版本是否支持集群管理决定的。
11、 如权利要求 9所述的方法， 其特征在于， 步骤（A2 )中在候 选设备向集群管理设备反馈同意加入集群的确认响应前，还要判断本 候选设备是否已设定特权用户密码， 如果没有设定， 直接反馈同意加 入集群的确认响应操作； 如果已经设定， 则向集群管理反馈要求鉴权 操作 , 然后根据管理设备下发的鉴权信息进行本候选设备的鉴权操 作， 当鉴权通过后， 再反馈同意加入集群的确认响应操作， 否则向集 群管理设备反馈拒绝加入集群的响应。
12、 如权利要求 7所述的方法， 其特征在于， 步骤（4 ) 中所述 为加入集群的每个成员设备进行必需的配置，对每个成员设备配置包 括下述内容： 成员设备的编号、 成员设备的私有 IP地址、 成员设备 名称、 成员设备的状态、 成员设备操作的状态、 集群管理的密码。
13、 如权利要求 7所述的方法， 其特征在于， 步骤（4 ) 中所述 为加入集群的每个成员设备进行标识采用包括下述字段的数据结构 进行：
网络类型： 用于标识设备所在的网絡类型；
物理地址： 用于标识设备在网络中的物理地址。
14、 一种网络设备的集群管理装置， 包括： 集群设备管理器和成 员设备管理器， 其中：
集群设备管理器包括：
转换模块， 用于对成员设备的管理报文进行网络地址转换； 类动态主机配置协议 ( DHCP )模块， 用于完成成员网络设备的私 有 IP地址的分配；
集群成员管理模块， '用于对成员网络设备的集中管理， 对来自集 群外部、 目标为成员设备的管理报文， 在经过标准的网络地址转换模 块转发到相应的成员设备处理，使成员设备按通常的处理过程处理该 管理报文； 拓朴信息处理模块， 用于网络拓朴结构的发现， 收集网络中规定 跳数内的网络拓朴结构信息；
成员设备管理器包括：
集群成员管理模块， 用于完成成员设备端的集群管理任务； 拓朴信息处理模块，用于完成邻接设备的发现和拓朴收集请求的 响应 /转发处理。
PCT/CN2003/000720 2002-12-04 2003-08-26 Procede et systeme de gestion en grappes d'installations du reseau WO2004051927A1 (fr)

Priority Applications (5)

Application Number Priority Date Filing Date Title
EP03812116A EP1575212B1 (en) 2002-12-04 2003-08-26 Method and system for cluster managing of network facilities
JP2004555961A JP4081472B2 (ja) 2002-12-04 2003-08-26 ネットワーク装置のクラスタ管理方法及びその装置
AU2003257805A AU2003257805B2 (en) 2002-12-04 2003-08-26 Method and system for cluster managing of network facilities
US10/537,488 US7904535B2 (en) 2002-12-04 2003-08-26 Method of cluster management of network devices and apparatus thereof
DE60317705T DE60317705T2 (de) 2002-12-04 2003-08-26 Verfahren und system zur cluster-verwaltung von netzwerkeinrichtungen

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN02151018.0 2002-12-04
CNB021510180A CN1266882C (zh) 2002-12-04 2002-12-04 一种网络设备的管理方法

Publications (1)

Publication Number Publication Date
WO2004051927A1 true WO2004051927A1 (fr) 2004-06-17

Family

ID=32400069

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2003/000720 WO2004051927A1 (fr) 2002-12-04 2003-08-26 Procede et systeme de gestion en grappes d'installations du reseau

Country Status (8)

Country Link
US (1) US7904535B2 (zh)
EP (1) EP1575212B1 (zh)
JP (1) JP4081472B2 (zh)
CN (1) CN1266882C (zh)
AU (1) AU2003257805B2 (zh)
DE (1) DE60317705T2 (zh)
ES (1) ES2295696T3 (zh)
WO (1) WO2004051927A1 (zh)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113282391A (zh) * 2021-05-21 2021-08-20 北京京东振世信息技术有限公司 集群切换方法、集群切换装置、电子设备及可读存储介质

Families Citing this family (39)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US7631100B2 (en) * 2003-10-07 2009-12-08 Microsoft Corporation Supporting point-to-point intracluster communications between replicated cluster nodes
US7188167B2 (en) * 2004-03-19 2007-03-06 Motorola, Inc. Method and system for registering multiple communication devices of a user in a session initiation protocol (SIP) based communication system
US20050256935A1 (en) * 2004-05-06 2005-11-17 Overstreet Matthew L System and method for managing a network
FI20045234A0 (fi) * 2004-06-21 2004-06-21 Nokia Corp Datan lähetys viestintäjärjestelmässä
US7760695B2 (en) 2006-09-29 2010-07-20 Symbol Technologies, Inc. Methods and systems for centralized cluster management in wireless switch architecture
CN101175096B (zh) * 2006-11-01 2010-06-02 中国电信股份有限公司 基于源路由的可扩展ip网络的实现
CN101043442B (zh) * 2006-11-17 2011-05-25 神州数码网络(北京)有限公司 一种在以太网交换机上实现urpf的方法
US7860099B2 (en) * 2006-12-21 2010-12-28 Alpha Networks Inc. Method for managing and setting many network devices
US20080205295A1 (en) * 2007-02-02 2008-08-28 Lights Together Llc Creation of organizational hierarchies in a group-centric network via handshake mechanisms
CN101436981B (zh) * 2007-11-13 2011-12-07 中国电信股份有限公司 在扩展的IPv4网络中的域名服务器系统
CN101183984B (zh) * 2007-12-14 2011-11-09 华为技术有限公司 网管系统、管理方法及设备
CN101197720B (zh) * 2007-12-27 2010-06-23 华为技术有限公司 配置以太网设备的方法和装置
FR2931970B1 (fr) * 2008-05-27 2010-06-11 Bull Sas Procede de generation de requetes de manipulation d'une base de donnees d'initialisation et d'administration d'une grappe de serveurs , support de donnees et grappe de serveurs correspondants
WO2010071888A2 (en) * 2008-12-19 2010-06-24 Watchguard Technologies, Inc. Cluster architecture and configuration for network security devices
CN101621528B (zh) * 2009-06-30 2012-04-18 迈普通信技术股份有限公司 基于以太交换机集群管理的会话系统及会话通道实现方法
US8103795B2 (en) * 2009-07-09 2012-01-24 International Business Machines Corporation TCP/IP host name resolution on a private network
US8578055B2 (en) * 2009-07-09 2013-11-05 International Business Machines Corporation Propogation of DNS server IP addresses in a private network
US8140669B2 (en) * 2009-08-31 2012-03-20 International Business Machines Corporation Resolving hostnames on a private network with a public internet server
US8868764B2 (en) * 2009-12-18 2014-10-21 Centurylink Intellectual Property Llc System and method for management of ethernet premise devices
CN101841574B (zh) * 2010-04-09 2012-12-26 江苏东大金智建筑智能化系统工程有限公司 网络视频监控系统中ip地址自动分配及管理方法
CN102195809A (zh) * 2011-03-25 2011-09-21 杭州再灵电子科技有限公司 一种网络传输及控制系统
CN102769867B (zh) 2011-05-05 2017-08-11 北京三星通信技术研究有限公司 网络接入方法
CN102170371B (zh) * 2011-05-18 2013-07-10 武汉烽火网络有限责任公司 一种基于集中式管理的收集网络设备重要信息的方法
CN102546267B (zh) * 2012-03-26 2015-06-10 杭州华三通信技术有限公司 网络设备的自动配置方法和管理服务器
JP5740652B2 (ja) * 2012-03-28 2015-06-24 株式会社日立製作所 計算機システム及びサブシステム管理方法
CN102638472B (zh) * 2012-05-07 2015-04-15 杭州华三通信技术有限公司 一种Portal认证方法和设备
CN103516531A (zh) * 2012-06-21 2014-01-15 中兴通讯股份有限公司 一种自动发现网元的方法及对应网元、网管装置
CN102983988B (zh) * 2012-10-26 2016-04-06 杭州迪普科技有限公司 一种设备代理装置以及网络管理装置
JP5974911B2 (ja) * 2013-01-21 2016-08-23 日立金属株式会社 通信システムおよびネットワーク中継装置
US10237252B2 (en) * 2013-09-20 2019-03-19 Oracle International Corporation Automatic creation and management of credentials in a distributed environment
CN104753702B (zh) 2013-12-27 2018-11-20 华为技术有限公司 一种集群系统中的集群处理方法、装置及系统
US20160344717A1 (en) * 2014-01-31 2016-11-24 Hewlett-Packard Development Company, L.P Communicating between a cluster and a node external to the cluster
CN106411588B (zh) * 2016-09-29 2019-10-25 锐捷网络股份有限公司 一种网络设备管理方法、主设备及管理服务器
CN107566544B (zh) * 2017-08-30 2020-07-07 苏州浪潮智能科技有限公司 一种用于存储集群的存储设备部署变更方法
US20200295997A1 (en) * 2019-03-14 2020-09-17 Nokia Solutions And Networks Oy Device management clustering
US11579949B2 (en) 2019-03-14 2023-02-14 Nokia Solutions And Networks Oy Device application support
US11579998B2 (en) 2019-03-14 2023-02-14 Nokia Solutions And Networks Oy Device telemetry control
CN113794593B (zh) * 2021-09-14 2023-05-26 新华三信息安全技术有限公司 一种集群故障处理方法及装置
CN114124902B (zh) * 2021-11-12 2023-07-04 迈普通信技术股份有限公司 多激活检测方法及系统

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US6058431A (en) * 1998-04-23 2000-05-02 Lucent Technologies Remote Access Business Unit System and method for network address translation as an external service in the access server of a service provider
US20020040397A1 (en) * 2000-10-02 2002-04-04 Samsung Electronics Co., Ltd. IP based network system and networking method thereof
CN1411214A (zh) * 2002-11-15 2003-04-16 烽火通信科技股份有限公司 基于以太网技术的网络设备管理方法

Family Cites Families (12)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPH11122301A (ja) * 1997-10-20 1999-04-30 Fujitsu Ltd アドレス変換接続装置
US6167052A (en) * 1998-04-27 2000-12-26 Vpnx.Com, Inc. Establishing connectivity in networks
JP2000059385A (ja) 1998-08-07 2000-02-25 Ntt Data Corp Ｉｐアドレス重複時の複数システム管理方法
WO2000079733A2 (en) * 1999-06-23 2000-12-28 At & T Wireless Services, Inc. Methods and apparatus for reducing traffic over a communication link in a computer network
US6654796B1 (en) * 1999-10-07 2003-11-25 Cisco Technology, Inc. System for managing cluster of network switches using IP address for commander switch and redirecting a managing request via forwarding an HTTP connection to an expansion switch
US6917626B1 (en) * 1999-11-30 2005-07-12 Cisco Technology, Inc. Apparatus and method for automatic cluster network device address assignment
US6725264B1 (en) * 2000-02-17 2004-04-20 Cisco Technology, Inc. Apparatus and method for redirection of network management messages in a cluster of network devices
JP3574372B2 (ja) * 2000-03-14 2004-10-06 Ｋｄｄｉ株式会社 Ｄｎｓサーバ、端末および通信システム
US7334098B1 (en) * 2000-06-06 2008-02-19 Quantum Corporation Producing a mass storage backup using a log of write commands and time information
US7370223B2 (en) 2000-09-08 2008-05-06 Goahead Software, Inc. System and method for managing clusters containing multiple nodes
US7159016B2 (en) * 2001-12-18 2007-01-02 Avaya Technology Corp. Method and apparatus for configuring an endpoint device to a computer network
US7035858B2 (en) * 2002-04-29 2006-04-25 Sun Microsystems, Inc. System and method dynamic cluster membership in a distributed data system

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US6058431A (en) * 1998-04-23 2000-05-02 Lucent Technologies Remote Access Business Unit System and method for network address translation as an external service in the access server of a service provider
US20020040397A1 (en) * 2000-10-02 2002-04-04 Samsung Electronics Co., Ltd. IP based network system and networking method thereof
CN1411214A (zh) * 2002-11-15 2003-04-16 烽火通信科技股份有限公司 基于以太网技术的网络设备管理方法

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
See also references of EP1575212A4 *

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113282391A (zh) * 2021-05-21 2021-08-20 北京京东振世信息技术有限公司 集群切换方法、集群切换装置、电子设备及可读存储介质
CN113282391B (zh) * 2021-05-21 2023-09-26 北京京东振世信息技术有限公司 集群切换方法、集群切换装置、电子设备及可读存储介质

Also Published As

Publication number Publication date
US20060041650A1 (en) 2006-02-23
ES2295696T3 (es) 2008-04-16
JP4081472B2 (ja) 2008-04-23
US7904535B2 (en) 2011-03-08
EP1575212B1 (en) 2007-11-21
CN1266882C (zh) 2006-07-26
EP1575212A1 (en) 2005-09-14
DE60317705D1 (de) 2008-01-03
EP1575212A4 (en) 2006-08-09
AU2003257805A1 (en) 2004-06-23
AU2003257805B2 (en) 2008-02-21
DE60317705T2 (de) 2008-10-30
CN1505322A (zh) 2004-06-16
JP2006509386A (ja) 2006-03-16

Similar Documents

Publication Publication Date Title
WO2004051927A1 (fr) Procede et systeme de gestion en grappes d'installations du reseau
Sharma et al. In-band control, queuing, and failure recovery functionalities for openflow
US7380025B1 (en) Method and apparatus providing role-based configuration of a port of a network element
CN102084638B (zh) 计算机网络中接入服务器的确定性会话负载平衡和冗余
JP5144444B2 (ja) Ｉｐｏａチャンネルベースのデフォルト維持管理チャンネルを確立するための方法
WO2016177030A1 (zh) Sdn网络设备建链方法、设备和系统
US20050108432A1 (en) Automatic address management method
US20080080438A1 (en) Methods and systems for centralized cluster management in wireless switch architecture
CN102801623B (zh) 一种多接入数据转发方法及设备
WO2007019798A1 (fr) Procede de sauvegarde de liaison ascendante
CN107769939B (zh) 数据通信网中网元管理方法、网管、网关网元及系统
CN112769614B (zh) 一种按需vpn的自动管理方法和异构网络的互通系统
WO2003101124A1 (fr) Procede d'etablissement automatique d'une voie speciale d'entretien de fonctionnement dans une station de base de troisieme generation
WO2009059505A1 (fr) Procédé et système d'initialisation à distance
WO2019237683A1 (zh) 一种协议报文以及虚拟客户终端设备的管理方法
CN107547403B (zh) 报文转发方法、协助方法、装置、控制器及主机
WO2014180213A1 (zh) 一种tcp会话建立方法、装置、宿主节点和卫星节点
WO2007030970A1 (fr) Systeme de gestion par grappe dans une couche de commutation ethernet et procede associe
EP3583751A1 (en) Method for an improved deployment and use of network nodes of a switching fabric of a data center or within a central office point of delivery of a broadband access network of a telecommunications network
CN106131177B (zh) 一种报文处理方法及装置
WO2015090035A1 (zh) 网络资源的共享处理、共享方法及装置、系统
US10686752B2 (en) Methods for configuring and managing an IP network, corresponding devices and computer programs
JP4495049B2 (ja) パケット通信サービスシステム、パケット通信サービス方法、エッジ側ゲートウェイ装置、およびセンタ側ゲートウェイ装置
CN101971522A (zh) 使用多个通用即插即用因特网网关设备提供因特网网关服务的方法和装置
WO2012155571A1 (zh) 一种基站自配置的方法及装置

Legal Events

Date Code Title Description
AK Designated states

Kind code of ref document: A1

Designated state(s): AE AG AL AM AT AU AZ BA BB BG BR BY BZ CA CH CO CR CU CZ DE DK DM DZ EC EE ES FI GB GD GE GH GM HR HU ID IL IN IS JP KE KG KP KR KZ LC LK LR LS LT LU LV MA MD MG MK MN MW MX MZ NI NO NZ OM PG PH PL PT RO RU SC SD SE SG SK SL SY TJ TM TN TR TT TZ UA UG US UZ VC VN YU ZA ZM ZW

AL Designated countries for regional patents

Kind code of ref document: A1

Designated state(s): GH GM KE LS MW MZ SD SL SZ TZ UG ZM ZW AM AZ BY KG KZ MD RU TJ TM AT BE BG CH CY CZ DE DK EE ES FI FR GB GR HU IE IT LU MC NL PT RO SE SI SK TR BF BJ CF CG CI CM GA GN GQ GW ML MR NE SN TD TG

121 Ep: the epo has been informed by wipo that ep was designated in this application
ENP Entry into the national phase

Ref document number: 2006041650

Country of ref document: US

Kind code of ref document: A1

WWE Wipo information: entry into national phase

Ref document number: 10537488

Country of ref document: US

WWE Wipo information: entry into national phase

Ref document number: 2004555961

Country of ref document: JP

WWE Wipo information: entry into national phase

Ref document number: 2003257805

Country of ref document: AU

WWE Wipo information: entry into national phase

Ref document number: 2003812116

Country of ref document: EP

WWP Wipo information: published in national office

Ref document number: 2003812116

Country of ref document: EP

WWP Wipo information: published in national office

Ref document number: 10537488

Country of ref document: US

WWG Wipo information: grant in national office

Ref document number: 2003812116

Country of ref document: EP