[go: up one dir, main page]
More Web Proxy on the site http://driver.im/

WO2024230474A1 - 去中心的网络设备系统、方法、网络云报文转发器及可读存储介质 - Google Patents

去中心的网络设备系统、方法、网络云报文转发器及可读存储介质 Download PDF

Info

Publication number
WO2024230474A1
WO2024230474A1 PCT/CN2024/089374 CN2024089374W WO2024230474A1 WO 2024230474 A1 WO2024230474 A1 WO 2024230474A1 CN 2024089374 W CN2024089374 W CN 2024089374W WO 2024230474 A1 WO2024230474 A1 WO 2024230474A1
Authority
WO
WIPO (PCT)
Prior art keywords
ncp
management
ncps
message
primary
Prior art date
Application number
PCT/CN2024/089374
Other languages
English (en)
French (fr)
Inventor
赖利根
江祥栋
林丹宁
黄国华
Original Assignee
锐捷网络股份有限公司
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 锐捷网络股份有限公司 filed Critical 锐捷网络股份有限公司
Publication of WO2024230474A1 publication Critical patent/WO2024230474A1/zh

Links

Definitions

  • the present application relates to the field of network communication technology, and in particular to a decentralized network device system, method, network cloud message forwarder and readable storage medium.
  • the decentralized Network Cloud Controller (NCC) and the decentralized Disaggregate Chassis (DDC) network system can perfectly solve the problem of NCC loss of connection in the DDC network system.
  • a decentralized network device system comprising:
  • One or more network cloud message forwarders NCP including one or more candidate management NCPs and a primary management NCP, wherein the one or more candidate management NCPs are determined by the one or more NCPs according to received configuration messages, the primary management NCP is determined from the one or more candidate management NCPs by an election algorithm, and the primary management NCP is configured to centrally process management service messages of the system; and
  • One or more network cloud switching fabrics NCF configured to receive and forward messages from the one or more NCPs, the messages including messages notifying the results of the election algorithm and management services Message.
  • the one or more NCPs further include a backup management NCP, where the backup management NCP is determined from the one or more candidate management NCPs through an election algorithm and is different from the primary management NCP, and the primary management NCP is further configured as:
  • the management data used to manage the one or more NCFs and the one or more NCPs is backed up to the backup management NCP.
  • the management service message includes configuration information of the one or more NCFs and the one or more NCPs, and the main management NCP is further configured as:
  • the configuration information is distributed to the new NCP or NCF.
  • the management service message includes version files of the one or more NCFs and the one or more NCPs, and the main management NCP is further configured as:
  • the version file is received, and the version file is sent to an NCP to be upgraded and/or an NCF to be upgraded, wherein the NCP to be upgraded is at least one of the one or more NCPs, and the NCF to be upgraded is at least one of the one or more NCFs.
  • the backup management NCP is further configured to:
  • the backup management NCP is further configured as:
  • a re-election message is sent to the one or more candidate management NCPs through the NCF connected to the backup management NCP, so that the one or more candidate management NCPs determine a new backup management NCP from the one or more candidate management NCPs through the election algorithm.
  • the primary management NCP is further configured as:
  • the corresponding relationship between the IP address corresponding to the primary management NCP and the primary management NCP is updated into the mapping table between the IP address and the NCP;
  • the corresponding relationship between the IP address corresponding to the new primary management NCP and the new primary management NCP is updated in the mapping table between the IP address and the NCP.
  • the one or more NCPs are further configured as:
  • control service messages of the one or more NCPs are processed respectively, wherein the control service messages include messages used to configure protocols or messages for controlling the one or more NCFs and the one or more NCPs in the system.
  • the one or more NCPs include a first NCP, and the first NCP is further configured as:
  • the external message is sent to the NCP corresponding to the destination IP address for processing.
  • a second aspect of the present application provides a decentralized network construction method, wherein a network cloud message forwarder NCP applied to a decentralized network device system, the method comprising:
  • the election result is obtained, and when it is determined that the system belongs to the primary management NCP according to the election result, the management service message of the system is processed centrally.
  • the method further includes:
  • the management data for managing the one or more NCFs and the one or more NCPs in the system is backed up to a backup management NCP, wherein the backup management NCP is selected from the one or more candidate management Elected by the NCP.
  • the method further includes:
  • the version files of the one or more NCFs and the one or more NCPs are received, and the version files are sent to corresponding NCPs or NCFs, so that the one or more NCPs or one or more NCFs complete an upgrade operation based on the version files.
  • the method further includes:
  • the method further includes:
  • a re-election message is sent to the one or more candidate management NCPs through the NCF corresponding to itself, so that the one or more candidate management NCPs determine a new backup management NCP from the one or more candidate management NCPs through the election algorithm.
  • the method further includes:
  • the corresponding relationship between the IP address corresponding to the new primary management NCP and the new primary management NCP is updated in the mapping table between the IP address and the NCP.
  • the method further includes:
  • control service messages of the one or more NCPs include messages for configuring protocols or messages for controlling the one or more NCFs and the one or more NCPs in the system.
  • the method further includes:
  • the external message is sent to the NCP corresponding to the destination IP address for processing.
  • a network cloud message forwarder NCP comprising:
  • the memory is used to store the processor executable instructions
  • the processor is configured to execute the instructions to implement the decentralized network construction method as described in the second aspect above.
  • a computer-readable storage medium which stores a computer program executable by a computer.
  • the program runs on the computer, the computer executes the decentralized network construction method as described in the second aspect above.
  • a decentralized network device system comprising: a plurality of network cloud switching boards NCF, and a plurality of network cloud message forwarders NCP;
  • Each of the NCFs is used to receive and forward messages from the NCP, wherein the messages include election result messages and management service messages;
  • Each of the NCPs is used to, when determining that it is a candidate management NCP based on a configuration message from a management server, execute an election algorithm through an election module to determine a primary management NCP among multiple candidate management NCPs; and, when determining that it is the primary management NCP based on an election result from the election module, centrally process management service messages of the system through a protocol processing module.
  • the primary management NCP is further used for:
  • the primary management NCP is further used for:
  • a version file is received from the management server, and the version file is sent to a corresponding NCP or NCF, so that the NCP or NCF completes an upgrade operation based on the version file.
  • each of the NCPs is further used for:
  • the backup management NCP is further used for:
  • a re-election message is sent to the remaining candidate management NCPs except the primary management NCP and the backup management NCP through NCF, so that the remaining candidate management NCPs execute an election algorithm through an election module to determine a new backup management NCP among the remaining candidate management NCPs.
  • each of the NCPs is further used for:
  • control service message of each NCP is processed by the protocol processing module, and the control service message is a message used to configure the protocol or control the NCF and NCP in the system.
  • each of the NCPs is further used for:
  • the external message is processed by the protocol processing module
  • the external message is sent to the NCP corresponding to the destination IP address for processing.
  • each of the NCPs is further used for:
  • the primary management NCP corresponding to the primary management NCP is elected by the election module.
  • the IP is sent to the management message relay module, so that the management message relay module sets the NCP corresponding to the primary management IP in the mapping table of IP address and NCP as the primary management NCP;
  • the new active management IP corresponding to the new active management NCP is sent to the management message relay module through the election module, so that the management message relay module updates the NCP corresponding to the active management IP in the mapping table between IP address and NCP.
  • a decentralized network construction method which is applied to a network cloud message forwarder NCP of a decentralized network device system, and the method comprises:
  • the election result is obtained from the election module, and when it is determined that the system belongs to the primary management NCP according to the election result, the management service message of the system is centrally processed through the protocol processing module.
  • the method further includes:
  • a decentralized network construction device which is applied to a decentralized network device system network cloud message forwarder NCP, and the device includes:
  • an election module configured to receive a configuration message from a management server, and when determining that the management NCP belongs to a candidate management NCP according to the configuration message, execute an election algorithm through the election module to determine a primary management NCP among a plurality of candidate management NCPs;
  • the management module is configured to obtain the election result from the election module, and when it is determined that it belongs to the primary management NCP according to the election result, centrally processes the management service messages of the system through the protocol processing module.
  • an electronic device including:
  • the memory is used to store executable instructions of the processor
  • the processor is used to execute the executable instructions to implement the decentralized network construction method as described in the second aspect above.
  • a computer-readable storage medium is provided.
  • the instructions in the computer-readable storage medium are executed by a processor of an electronic device, the electronic device is enabled to execute the decentralized network construction method as described in the second aspect above.
  • the present application provides a computer program product, including a computer program:
  • FIG1 is a schematic diagram of a DDC system in the related art
  • FIG2 is a schematic diagram of a framework of a decentralized network device system provided in an embodiment of the present application.
  • FIG3 is a schematic diagram of the structure of the NCP provided in an embodiment of the present application.
  • FIG4 is a schematic diagram of a flow chart of a step of executing message relaying by NCP according to an embodiment of the present application
  • FIG5 is a schematic diagram of a message relay process provided in an embodiment of the present application.
  • FIG6 is a schematic diagram of the structure of a decentralized network construction device 600 provided in an embodiment of the present application.
  • FIG. 7 is a schematic diagram of the structure of an electronic device provided in an embodiment of the present application.
  • “/” means or, for example, A/B can mean A or B; the “and/or” in the text is merely a description of the association relationship of associated objects, indicating that three relationships may exist, for example, A and/or B can mean: A exists alone, A and B exist at the same time, and B exists alone.
  • “multiple” refers to two or more than two.
  • first and second are used for descriptive purposes only and should not be understood as suggesting or implying relative importance or implicitly indicating the number of technical features indicated.
  • features defined as “first” and “second” may explicitly or implicitly include one or more of the features.
  • “multiple” means two or more.
  • DDC is a distributed separated rack network device system based on Application Specific Integrated Circuit (ASIC) chip design.
  • ASIC Application Specific Integrated Circuit
  • NCC replaces the original management board
  • NCF replaces the original switching network board
  • NCP Network Cloud Packet Forwarder
  • NCP replaces the original management board.
  • MMD network management device
  • NCF network cloud fabric
  • NCP Network Cloud Packet Forwarder
  • DDC forms a rack system that can be flexibly expanded and deployed.
  • Traditional rack switches are limited by the number of chassis slots, cabinet power supply, and cabinet size, and currently support a maximum of 16 line cards, while the DDC system can support 96 or even more line cards (the line card is called NCP in the DDC architecture).
  • DDC uses ASIC chips and takes advantage of the chip's credit scheduling, backplane cell switching, VOQ (Virtual Output Queue) and other features to achieve a network architecture with stable latency and high bandwidth utilization.
  • NCP1 needs to know the specific exit of NCP2 in order to forward across NCPs, which also makes it impossible to manage them completely as a single box device. Therefore, traditional solutions usually logically abstract the DDC system into a rack-mounted device for management.
  • the number of failure points increases. For example, the network between the NCC and the NCP/NCF is disconnected due to multiple single-point failures. Once the NCC loses connection, the entire DDC system will be out of control.
  • the DDC system needs to integrate 100 to 500 NCPs, and the network scale is dozens of times that of traditional rack equipment. The failure radius is too large, which is unacceptable to most customer scenarios.
  • the decentralized DDC network system includes the following features:
  • NCC is removed and only NCP and NCF are retained.
  • Each NCP provides management and configuration interfaces to the outside world and processes control protocol messages in a distributed manner, such as Address Resolution Protocol (ARP) messages.
  • ARP Address Resolution Protocol
  • NCP devices such as global table entries, NCP devices and ports, etc.
  • the decentralized (NCC removed) DDC network system can solve the problem of NCC disconnection.
  • NCC since there is no NCC to aggregate the information of each device, it is necessary to log in to different NCPs and NCFs to perform equipment maintenance and management, which greatly increases the complexity of maintenance and management, which is mainly reflected in the following aspects:
  • each exemplary embodiment of the present application provides a decentralized network equipment system, a decentralized network construction method and equipment to solve the problem of complex maintenance and management of decentralized DDC network systems in related technologies.
  • a plurality of network cloud switching fabrics NCF and a plurality of network cloud message forwarders NCP are combined into a decentralized network device system; wherein each NCP includes an election module and a protocol processing module, and each NCF is used to receive and forward messages from the NCP; each NCP is used to determine that it belongs to a candidate management NCP according to a configuration message from a management server (such as the aforementioned network management device) (that is, the candidate management NCP can be determined by the one or more NCPs according to the received configuration message, that is, the configuration message is used for the NCP that receives the configuration message to determine whether it belongs to a candidate management NCP), execute an election algorithm through the election module to determine the main management NCP among the multiple candidate management NCPs; and, when it is determined that it belongs to the main management NCP according to the election result from the election module, the management service messages of the system are centrally processed through the protocol processing module (that is, centralized management can
  • the embodiment of the present application is based on the decentralized DDC system, and the main management NCP obtained by election aggregates the management service messages of each device in the DDC system, without logging in to different
  • the NCP and NCF of the equipment are used for maintenance and management, which reduces the complexity of maintenance and management of the decentralized DDC system and improves the availability of system maintenance and management.
  • the embodiment of the present application provides a decentralized network device system, which belongs to a DDC network.
  • the network device system includes multiple network cloud switching boards NCF and multiple network cloud message forwarders NCP, wherein each NCF is used to receive and forward messages from the NCP; each NCP includes an election module and a protocol processing module, which is used to determine that it belongs to a candidate management NCP according to the configuration message from the management server, and execute an election algorithm through the election module to determine the main management NCP among multiple candidate management NCPs; and when it is determined that it belongs to the main management NCP according to the election result from the election module, the management service message of the system is centrally processed through the protocol processing module.
  • the candidate management NCP can directly obtain the election result of the election module to determine whether it belongs to the main management NCP; and forward the election result in the form of an election result message through the NCF, so that all NCPs except the candidate management NCP can know the election result.
  • each NCP is further configured to:
  • the control service message of each NCP is processed by the protocol processing module.
  • each NCP in the embodiment of the present application includes: a command line interface (CLI) module, an election module, a management message relay module, a packet transceiver module, a protocol processing module and an ASIC chip, wherein the CLI module is used to configure the IP address of the NCP, the election module is used to elect the main management NCP, and the packet transceiver module is used to
  • the ASIC chip is used to receive and forward messages within the NCP.
  • the protocol processing module is used to process the management service messages of the system.
  • the management message relay module is used to relay external messages, which can be messages sent by the management server.
  • the ASIC chip is used to receive and forward messages from other NCPs and management servers.
  • the difference from the NCP in the related art is at least that, in the embodiment of the present application, an election module is added to the NCP to elect a main management NCP, and the management service messages of each device in the DDC system are aggregated and managed by the elected main management NCP. There is no need to log in to different NCPs and NCFs to perform equipment maintenance and management, thereby reducing the complexity of maintenance and management of the decentralized DDC system.
  • the present application solves the contradiction between the reliability and availability of the decentralized DDC network system by separating control and management, that is, centrally processing the management services (such as configuration delivery, fault diagnosis, and upgrade) that are less affected by the failure of the centralized control point, corresponding to the above-mentioned steps of centrally processing the management service messages of the system through the main management NCP, the management service messages are messages used for configuration delivery, fault diagnosis, system upgrades, or messages for managing the NCF and NCP in the system; the control plane service messages that are greatly affected by the failure of the centralized control point (such as messages for the address resolution protocol ARP and the configuration of the border gateway protocol BGP) are distributedly processed.
  • the control service messages are messages used for configuring the protocol and/or controlling the NCF and NCP in the system.
  • the management service message of the primary management NCP centralized processing system includes the following contents:
  • Configuration information is information that enables the device to work normally, such as global table entry information.
  • the backup management NCP is selected from multiple candidate management NCPs.
  • the embodiment of the present application provides three flexible upgrade methods, including specifying a single NCP/NCF device for upgrade, specifying multiple devices for upgrade, and upgrading all devices in the entire network with one click. Three functional options meet the upgrade needs in different scenarios.
  • the specific process of each NCP selecting a backup management NCP from multiple candidate management NCPs includes:
  • the election module executes an election algorithm to determine a backup management NCP among multiple candidate management NCPs;
  • the candidate management NCP can also directly obtain the election results of the election module to determine whether it belongs to the backup management NCP; and forward the election results in the form of election result messages through the NCF, so that all NCPs except the candidate management NCP can know the election results.
  • the embodiment of the present application needs to first determine the candidate management NCP.
  • the process of determining the candidate management NCP is as follows: First, determine the candidate management NCPs participating in the election, such as selecting 3 NCPs from 100 NCPs to be configured as candidate management NCPs. In order to avoid the problem of network-wide interruption during system version upgrade, it is necessary to upgrade the NCPs in batches. Therefore, the candidate management NCPs need to be distributed to different upgrade batches.
  • the embodiment of the present application can automatically poll and select (or randomly select) the candidate management NCP from the median of the NCP device number to both sides greater than the median and less than the median, or the user can specify the candidate management NCP through configuration.
  • An example of automatic polling selection is as follows: Assume that there are 100 NCPs numbered from 1 to 100. Starting from the median NCP numbered 50, NCP numbered 49 is selected first, then NCP numbered 51 is selected, then NCP numbered 48 is selected first, then NCP numbered 52 is selected, and so on, until the number of candidate management NCPs reaches the expected number. If the expected number is 3, NCPs numbered 49, 50, and 51 are selected.
  • the candidate management NCP election module automatically selects the candidate management NCP.
  • the NCPs are determined to be the primary management NCP and the backup management NCP respectively, and the remaining NCPs continue to serve as candidate management NCPs.
  • the embodiment of the present application solves the problem of difficulty in decision-making for all NCPs to participate in the election by designating a small number of candidates to manage NCPs.
  • the backup management NCP is further used for:
  • a re-election message is sent to the remaining candidate management NCPs except the primary management NCP and the backup management NCP through the NCF, so that the remaining candidate management NCPs execute the election algorithm through the election module to determine a new backup management NCP among the remaining candidate management NCPs.
  • the backup management NCP switches to the new main management NCP, and the election module re-elects an NCP from the remaining candidate management NCPs as the backup management NCP, avoiding the problem of system paralysis caused by the failure of the main management NCP and improving the availability of the system.
  • the management server may access NCP2, but the main management NCP is NCP1. At this time, the message that the management server communicates with the main management NCP will be sent to the CPU of NCP2 first.
  • the embodiment of the present application adds a management message relay module to each NCP. As shown in Figure 3 above, the problem of cross-NCP in-band remote management can be solved by message relay, that is, the message is relayed to NCP1 through the management message relay module of NCP2.
  • message relay that is, the message is relayed to NCP1 through the management message relay module of NCP2.
  • each NCP performs a message relay step, the process of which is shown in FIG. 4 and includes the following steps.
  • step 401 an external message is received, and a mapping table between an IP address and NCP is queried according to the destination IP address of the external message through a management message relay module.
  • the external message may be a management message.
  • step 402 if it is determined that the NCP corresponding to the destination IP address of the external message is the local device, the external message is processed by the protocol processing module.
  • step 403 if it is determined that the NCP corresponding to the destination IP address of the external message is not the local machine, Incoming packets are sent to the NCP corresponding to the destination IP address for processing.
  • NCP2 receives a message from the management server requesting access to NCP1, and sends the request message to the CPU of NCP2 (the CPU includes modules such as an election module and a management message relay module).
  • NCP2 finds that the destination IP address of the access is the IP address of another NCP.
  • the management message relay module queries the mapping table of IP addresses and NCPs according to the destination IP address of the request message, determines that the NCP corresponding to the destination IP address of the request message is the IP address of NCP1, and relays the message to the CPU of NCP1.
  • NCP1 finds that the accessed IP address is the management address of the local machine, and processes the request message through the protocol processing module.
  • each NCP is further configured to:
  • the primary management IP corresponding to the primary management NCP is sent to the management message relay module through the election module, so that the management message relay module sets the NCP corresponding to the primary management IP in the mapping table of IP addresses and NCPs as the primary management NCP;
  • the new active management IP corresponding to the new active management NCP is sent to the management message relay module through the election module, so that the management message relay module updates the NCP corresponding to the active management IP in the mapping table between IP address and NCP.
  • the NCP corresponding to the primary management IP is empty by default.
  • the election module selects the primary management NCP, the election module notifies the management message relay module to update the NCP corresponding to the primary management IP to the primary management NCP.
  • the election module notifies the management message relay module to update the NCP corresponding to the new primary management IP in the mapping table of IP address and NCP to the new primary management NCP.
  • the embodiment of the present application further provides a decentralized network construction method, which is applied to the network cloud message forwarder NCP of the above-mentioned decentralized network device system, and includes the following contents:
  • the election module Obtain the election results from the election module and determine that it is the primary administrator based on the election results.
  • the system management service messages are centrally processed through the protocol processing module.
  • the method further includes:
  • the embodiment of the present application further provides a decentralized network construction device, which is applied to the network cloud message forwarder NCP of the above-mentioned decentralized network device system.
  • the device 600 includes:
  • the election module 601 is configured to receive a configuration message from the management server, and when it is determined that it belongs to a candidate management NCP according to the configuration message, execute an election algorithm through the election module to determine a primary management NCP among multiple candidate management NCPs;
  • the management module 602 is configured to obtain the election result from the election module, and when it is determined that it belongs to the primary management NCP according to the election result, centrally process the management service messages of the system through the protocol processing module.
  • the present application provides an optional embodiment, which, based on a decentralized DDC system, aggregates and manages the management service messages of each device in the DDC system through the elected main management NCP. There is no need to log in to different NCPs and NCFs to perform equipment maintenance and management, thereby reducing the complexity of maintenance and management of the decentralized DDC system and improving the availability of system maintenance and management.
  • the electronic device 130 according to the embodiment of the present application is described below with reference to Fig. 7.
  • the electronic device 130 shown in Fig. 7 is only an example and should not bring any limitation to the function and scope of use of the embodiment of the present application.
  • the electronic device 130 is in the form of a general electronic device.
  • the components of the electronic device 130 may include but are not limited to: at least one processor 131, at least one memory 132, and a bus 133 connecting different system components (including the memory 132 and the processor 131).
  • Bus 133 represents one or more of several types of bus structures, including a memory bus or a storage
  • a bus may be a processor controller, a peripheral bus, a processor, or a local bus using any of a variety of bus architectures.
  • the memory 132 may include a non-volatile storage medium or a readable medium in the form of a volatile memory, such as a random access memory (RAM) 1321 and/or a cache memory 1322 , and may further include a read-only memory (ROM) 1323 .
  • RAM random access memory
  • ROM read-only memory
  • the memory 132 may also include a program/utility 1325 having a set (at least one) of program modules 1324, such program modules 1324 including but not limited to: an operating system, one or more application programs, other program modules, and program data, each of which or some combination may include an implementation of a network environment.
  • program modules 1324 including but not limited to: an operating system, one or more application programs, other program modules, and program data, each of which or some combination may include an implementation of a network environment.
  • the electronic device 130 may also communicate with one or more external devices 134 (e.g., keyboards, pointing devices, etc.), may also communicate with one or more devices that enable a user to interact with the electronic device 130, and/or communicate with any device that enables the electronic device 130 to communicate with one or more other electronic devices (e.g., routers, modems, etc.). Such communication may be performed via an input/output (I/O) interface 135.
  • the electronic device 130 may also communicate with one or more networks (e.g., a local area network (LAN), a wide area network (WAN), and/or a public network, such as the Internet) via a network adapter 136.
  • networks e.g., a local area network (LAN), a wide area network (WAN), and/or a public network, such as the Internet
  • the network adapter 136 communicates with other modules for the electronic device 130 via a bus 133. It should be understood that, although not shown in the figure, other hardware and/or software modules may be used in conjunction with the electronic device 130, including but not limited to: microcode, device drivers, redundant processors, external disk drive arrays, RAID systems, tape drives, and data backup storage systems.
  • the present application further provides a computer-readable storage medium including instructions, such as a memory 132 including instructions, and the above instructions can be executed by the processor 131 of the electronic device 130 to complete the above decentralized network construction method.
  • the computer-readable storage medium can be a non-temporary computer-readable storage medium, for example, the non-temporary computer-readable storage medium can be a ROM, a random access memory (RAM), a CD-ROM, a tape, a floppy disk, an optical data storage device, etc.
  • a computer program product including a computer program, which, when executed by the processor 131, implements the decentralized network construction method provided in the present application. Law.
  • the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented in one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that include computer-usable program code.
  • a computer-usable storage media including but not limited to disk storage, CD-ROM, optical storage, etc.
  • These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer-readable memory produce a manufactured product including an instruction device that implements the functions specified in one or more processes in the flowchart and/or one or more boxes in the block diagram.
  • These computer program instructions may also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, whereby the instructions executed on the computer or other programmable device provide steps for implementing the functions specified in one or more processes in the flowchart and/or one or more boxes in the block diagram.

Abstract

本申请实施例公开了一种去中心的网络设备系统、方法、网络云报文转发器及可读存储介质。所述系统包括:一个或多个网络云报文转发器NCP,包括一个或多个候选管理NCP及主管理NCP,其中,所述一个或多个候选管理NCP由所述一个或多个NCP根据接收到的配置报文来确定,所述主管理NCP是通过选举算法从所述一个或多个候选管理NCP中确定,以及所述主管理NCP配置为集中处理所述系统的管理业务报文;以及一个或多个网络云交换网板NCF,配置为接收以及转发来自所述一个或多个NCP的报文,所述报文包括通知所述选举算法的结果的报文和管理业务报文。

Description

去中心的网络设备系统、方法、网络云报文转发器及可读存储介质
相关申请的交叉引用
本申请要求于2023年05月08日提交中国国家知识产权局、申请号为202310511065.1、发明名称为“去中心的网络设备系统、去中心的网络构建方法及设备”的中国专利申请的优先权,其全部内容通过引用结合在本申请中。
技术领域
本申请涉及网络通信技术领域,尤其涉及去中心的网络设备系统、方法、网络云报文转发器及可读存储介质。
背景技术
目前,去中心网络云控制器(Network Cloud Controller,NCC),化的分布式分离机架网络设备系统(Distributed Disaggregate Chassis,DDC)网络系统,可以完美解决DDC网络系统存在NCC失联的问题。
发明内容
在本申请的第一方面,提供一种去中心的网络设备系统,包括:
一个或多个网络云报文转发器NCP,包括一个或多个候选管理NCP及主管理NCP,其中,所述一个或多个候选管理NCP由所述一个或多个NCP根据接收到的配置报文来确定,所述主管理NCP是通过选举算法从所述一个或多个候选管理NCP中确定,以及所述主管理NCP配置为集中处理所述系统的管理业务报文;以及
一个或多个网络云交换网板NCF,配置为接收以及转发来自所述一个或多个NCP的报文,所述报文包括通知所述选举算法的结果的报文和管理业务 报文。
在一种可能的实施方式中,所述一个或多个NCP还包括备份管理NCP,所述备份管理NCP是通过选举算法从所述一个或多个候选管理NCP中确定且与所述主管理NCP不同,所述主管理NCP还配置为:
将用于管理所述一个或多个NCF与所述一个或多个NCP的管理数据备份到所述备份管理NCP。
在一种可能的实施方式中,所述管理业务报文包括所述一个或多个NCF与所述一个或多个NCP的配置信息,所述主管理NCP还配置为:
获取所述系统中所述一个或多个NCF与所述一个或多个NCP的配置信息;以及
在确定新的NCP或NCF加入后,将所述配置信息分发到所述新的NCP或NCF。
在一种可能的实施方式中,所述管理业务报文包括所述一个或多个NCF与所述一个或多个NCP的版本文件,所述主管理NCP还配置为:
接收所述版本文件,并将所述版本文件发送给待升级的NCP和/或待升级的NCF,所述待升级的NCP为所述一个或多个NCP中的至少一个,所述待升级的NCF为所述一个或多个NCF中的至少一个。
在一种可能的实施方式中,所述备份管理NCP还被配置为:
接收来自所述主管理NCP的管理数据进行备份;以及
确定所述主管理NCP发生故障时,切换为新主管理NCP。
在一种可能的实施方式中,所述备份管理NCP还配置为:
在确定所述主管理NCP发生故障时,通过与所述备份管理NCP连接的NCF向所述一个或多个候选管理NCP发送重新选举报文,以使所述一个或多个候选管理NCP通过所述选举算法从所述一个或多个候选管理NCP中确定新备份管理NCP。
在一种可能的实施方式中,所述主管理NCP还配置为:
在从所述一个或多个候选管理NCP中被确定为所述主管理NCP后,将所 述主管理NCP对应的IP地址与所述主管理NCP的对应关系更新至IP地址与NCP的映射表中;以及
在所述备份管理NCP被切换为新主管理NCP后,将所述新主管理NCP对应的IP地址与所述新主管理NCP的对应关系更新至所述IP地址与NCP的映射表中。
在一种可能的实施方式中,所述一个或多个NCP还配置为:
分别处理所述一个或多个NCP的控制业务报文,所述控制业务报文包括用于配置协议的报文或控制所述系统中所述一个或多个NCF和所述一个或多个NCP的报文。
在一种可能的实施方式中,所述一个或多个NCP包括第一NCP,所述第一NCP还配置为:
接收外来报文,并根据所述外来报文的目的互联网协议IP地址通过查询IP地址与NCP的映射表确定是否处理所述外来报文;以及
若确定所述外来报文的目的IP地址对应的NCP非所述第一NCP,则将所述外来报文发送给所述目的IP地址对应的所述NCP进行处理。
本申请的第二方面,提供一种去中心的网络构建方法,其中,应用于去中心的网络设备系统的网络云报文转发器NCP,所述方法包括:
接收配置报文,并在根据所述配置报文确定自身属于候选管理NCP时,通过选举算法确定一个或多个候选管理NCP中的主管理NCP;以及,
获取选举结果,并在根据所述选举结果确定自身属于所述主管理NCP时,集中处理所述系统的管理业务报文。
在一种可能的实施方式中,所述方法还包括:
获取所述系统中一个或多个NCF与一个或多个NCP的配置信息;
在确定新的NCP或NCF加入后,将所述配置信息分发到新的NCP或NCF;以及
将用于管理所述系统中所述一个或多个NCF与所述一个或多个NCP的管理数据备份到备份管理NCP,所述备份管理NCP是从所述一个或多个候选管 理NCP中选举得到的。
在一种可能的实施方式中,所述方法还包括:
接收所述一个或多个NCF与所述一个或多个NCP的版本文件,并将所述版本文件发送给对应的NCP或NCF,以使所述一个或多个NCP或一个或多个NCF基于所述版本文件完成升级操作。
在一种可能的实施方式中,所述方法还包括:
根据来自管理服务器的配置报文确定自身属于候选管理NCP时,通过选举算法,以确定多个候选管理NCP中的备份管理NCP;以及,
根据选举结果确定自身属于所述备份管理NCP时,接收来自所述主管理NCP的管理数据进行备份;以及,确定所述主管理NCP发生故障时,切换为新主管理NCP。
在一种可能的实施方式中,所述方法还包括:
在确定所述主管理NCP发生故障时,通过与自身对应的NCF向所述一个或多个候选管理NCP发送重新选举报文,以使所述一个或多个候选管理NCP通过所述选举算法从所述一个或多个候选管理NCP中确定新备份管理NCP。
在一种可能的实施方式中,所述方法还包括:
在从所述一个或多个候选管理NCP中被确定为主管理NCP后,将所述主管理NCP的IP地址与所述主管理NCP的对应关系更新至IP地址与NCP的映射表中;以及
在所述备份管理NCP被切换为新主管理NCP后,将所述新主管理NCP对应的IP地址与所述新主管理NCP的对应关系更新至所述IP地址与NCP的映射表中。
在一种可能的实施方式中,所述方法还包括:
处理所述一个或多个NCP的控制业务报文,所述控制业务报文包括用于配置协议的报文或控制所述系统中所述一个或多个NCF和所述一个或多个NCP的报文。
在一种可能的实施方式中,所述方法还包括:
接收外来报文,并根据所述外来报文的目的IP地址查询IP地址与NCP的映射表确定是否处理所述外来报文;
若确定所述外来报文的目的IP地址对应的NCP非本机,则将所述外来报文发送给所述目的IP地址对应的所述NCP进行处理。
本申请第三方面,提供一种网络云报文转发器NCP,包括:
处理器和存储器;
其中,所述存储器,用于存储所述处理器可执行指令;
所述处理器被配置为执行所述指令以实现如上述第二方面所述的去中心的网络构建方法。
本申请第四方面,提供一种计算机可读存储介质,其中,其存储有可由计算机执行的计算机程序,当所述程序在所述计算机上运行时,使得所述计算机执行如上述第二方面所述的去中心的网络构建方法。
本申请第五方面,提供一种去中心的网络设备系统,所述系统包括:多个网络云交换网板NCF,和多个网络云报文转发器NCP;
每个所述NCF,用于接收以及转发来自NCP的报文,所述报文包括选举结果报文和管理业务报文;
每个所述NCP,用于根据来自管理服务器的配置报文确定自身属于候选管理NCP时,通过选举模块执行选举算法,以确定多个候选管理NCP中的主管理NCP;以及,根据来自选举模块的选举结果确定自身属于所述主管理NCP时,通过协议处理模块集中处理所述系统的管理业务报文。
在一种可能的实施方式中,所述主管理NCP还用于:
获取所述系统中NCF与NCP的配置信息;
在确定新的NCP或NCF加入后,将所述配置信息分发到新的NCP或NCF;
以及,将用于管理所述系统中NCF与NCP的管理数据备份到备份管理NCP,所述备份管理NCP是从所述多个候选管理NCP中选举得到的。
在一种可能的实施方式中,所述主管理NCP还用于:
接收来自所述管理服务器的版本文件,并将所述版本文件发送给对应的NCP或NCF,以使所述NCP或NCF基于所述版本文件完成升级操作。
在一种可能的实施方式中,每个所述NCP还用于:
根据来自所述管理服务器的配置报文确定自身属于候选管理NCP时,通过选举模块执行选举算法,以确定多个候选管理NCP中的备份管理NCP;以及,
根据来自选举模块的选举结果确定自身属于所述备份管理NCP时,接收来自所述主管理NCP的管理数据进行备份;以及,确定所述主管理NCP发生故障时,切换为新主管理NCP。
在一种可能的实施方式中,所述备份管理NCP还用于:
在确定所述主管理NCP发生故障时,通过NCF向除所述主管理NCP和所述备份管理NCP外的剩余候选管理NCP发送重新选举报文,以使所述剩余候选管理NCP通过选举模块执行选举算法,以确定所述剩余候选管理NCP中的新备份管理NCP。
在一种可能的实施方式中,每个所述NCP还用于:
通过协议处理模块处理每个所述NCP的控制业务报文,所述控制业务报文是用于配置协议或控制所述系统中NCF和NCP的报文。
在一种可能的实施方式中,每个所述NCP还用于:
接收外来报文,并通过管理报文中继模块根据所述外来报文的目的IP地址查询IP地址与NCP的映射表;
若确定所述外来报文的目的IP地址对应的NCP是本机,则通过协议处理模块对所述外来报文进行处理;
若确定所述外来报文的目的IP地址对应的NCP非本机,则将所述外来报文发送给所述目的IP地址对应的NCP进行处理。
在一种可能的实施方式中,每个所述NCP还用于:
在确定主管理NCP后,通过选举模块将所述主管理NCP对应的主管理 IP发送到管理报文中继模块,以使所述管理报文中继模块将IP地址与NCP的映射表中主管理IP对应的NCP设置为主管理NCP;
以及,在主备切换后,通过选举模块将新主管理NCP对应的新主管理IP发送到管理报文中继模块,以使所述管理报文中继模块更新IP地址与NCP的映射表中所述主管理IP对应的NCP。
本申请的第六方面,提供一种去中心的网络构建方法,应用于去中心的网络设备系统的网络云报文转发器NCP,所述方法包括:
接收来自管理服务器的配置报文,并在根据所述配置报文确定自身属于候选管理NCP时,通过选举模块执行选举算法,以确定多个候选管理NCP中的主管理NCP;以及,
获取来自选举模块的选举结果,并在根据所述选举结果确定自身属于所述主管理NCP时,通过协议处理模块集中处理所述系统的管理业务报文。
在一种可能的实施方式中,所述方法还包括:
获取所述系统中NCF与NCP的配置信息;
在确定新的NCP或NCF加入后,将所述配置信息分发到新的NCP或NCF;
以及,将用于管理所述系统中NCF与NCP的管理数据备份到备份管理NCP,所述备份管理NCP是从所述多个候选管理NCP中选举得到的。
本申请第七方面,提供一种去中心的网络构建装置,应用于去中心的网络设备系统网络云报文转发器NCP,所述装置包括:
选举模块,被配置为接收来自管理服务器的配置报文,并在根据所述配置报文确定自身属于候选管理NCP时,通过选举模块执行选举算法,以确定多个候选管理NCP中的主管理NCP;以及,
管理模块,被配置为获取来自选举模块的选举结果,并在根据所述选举结果确定自身属于所述主管理NCP时,通过协议处理模块集中处理所述系统的管理业务报文。
本申请第八方面,提供一种电子设备,包括:
处理器和存储器;
所述存储器,用于存储处理器的可执行指令;
所述处理器,用于执行所述可执行指令,以实现如上述第二方面所述的去中心的网络构建方法。
本申请第九方面,提供一种计算机可读存储介质,当所述计算机可读存储介质中的指令由电子设备的处理器执行时,使得所述电子设备能够执行如上述第二方面所述的去中心的网络构建方法。
本申请第十方面,本申请提供一种计算机程序产品,包括计算机程序:
所述计算机程序被处理器执行时实现如上述第二方面所述的去中心的网络构建方法。
本申请的其它特征和优点将在随后的说明书中阐述,并且,部分地从说明书中变得显而易见,或者通过实施本申请而了解。本申请的目的和其他优点可通过在所写的说明书、权利要求书、以及附图中所特别指出的结构来实现和获得。
附图说明
为了更清楚地说明本申请实施例的技术方案,下面将对本申请实施例中所需要使用的附图作简单地介绍,显而易见地,下面所介绍的附图仅仅是本申请的一些实施例,对于本领域普通技术人员来讲,在不付出创造性劳动的前提下,还可以根据这些附图获得其他的附图。
图1为相关技术中DDC系统的框架示意图;
图2为本申请实施例提供的去中心的网络设备系统的框架示意图;
图3为本申请实施例提供的NCP的结构示意图;
图4为本申请实施例提供的NCP执行报文中继步骤的流程示意图;
图5为本申请实施例提供的报文中继过程的示意图;
图6为本申请实施例提供的去中心的网络构建装置600的结构示意图;
图7为本申请实施例提供的电子设备的结构示意图。
具体实施方式
为使本申请实施例的目的、技术方案和优点更加清楚,下面将结合本申请实施例中的附图,对本申请实施例中的技术方案进行清楚、完整地描述。其中,所描述的实施例是本申请一部分实施例,而不是全部的实施例。基于本申请中的实施例,本领域普通技术人员在没有做出创造性劳动前提下所获得的所有其它实施例,都属于本申请保护的范围。
在本申请实施例的描述中,除非另有说明,“/”表示或的意思,例如,A/B可以表示A或B;文本中的“和/或”仅仅是一种描述关联对象的关联关系,表示可以存在三种关系,例如,A和/或B,可以表示:单独存在A,同时存在A和B,单独存在B这三种情况,另外,在本申请实施例的描述中,“多个”是指两个或多于两个。
以下,术语“第一”、“第二”仅用于描述目的,而不能理解为暗示或暗示相对重要性或者隐含指明所指示的技术特征的数量。由此,限定有“第一”、“第二”、的特征可以明示或者隐含地包括一个或者更多个该特征,在本申请实施例的描述中,除非另有说明,“多个”的含义是两个或两个以上。
在使用本申请中描述的“包括”、“具有”、和“包含”的情况下,除非使用了明确的限定用语,例如“仅”、“由……组成”等,否则还可以添加另一部件。除非相反地提及,否则单数形式的术语可以包括复数形式,并不能理解为其数量仅为一个。
DDC是一种基于专用集成电路(Application Specific Integrated Circuit,ASIC)芯片设计的一种分布式分离机架网络设备系统。该系统是在原来机架网络设备(如交换机或路由器,为便于理解,以下均以交换机为例进行说明)基础上,将管理板、线卡、交换矩阵(Fabric)卡做成物理上独立的设备,去掉机架交换机的机框,再通过线缆相连组成一个逻辑上的机架交换机,如图1所示。
其中,传统的DDC上用NCC代替原来的管理板,用NCF来代替原来的交换网板,用网络云报文转发器(Network Cloud Packet Forwarder,NCP)来 代替原来的线卡。用网络管理设备(Management Device,MGD)中继连接NCC与网络云交换网板(Network Cloud Fabric,NCF)或NCP,即NCC和NCP/NCF之间,通过网络管理设备相连。
由此,DDC形成一个可以灵活扩展、灵活部署的机架系统。传统的机架交换机受限于机框槽位数量、机柜供电、机柜大小的限制,支持的线卡数量目前最多为16个,而DDC系统可以支持96个甚至更多的线卡(该线卡在DDC架构里面称为NCP)。
同时,DDC使用ASIC芯片,利用芯片的信任(credit)调度、背板信元交换、VOQ(Virtual Output Queue,虚拟输出队列)等特性,实现稳定时延、高带宽利用率的网络架构。但由于在各个NCP之间转发时,NCP1需知道NCP2具体的出口才能跨NCP转发,这也导致了它们无法完全按单个盒式设备来管理。因此,传统方案通常将DDC系统逻辑上抽象成一台机架式设备来管理。
DDC系统将管理板和线卡/交换网板在物理上分离后,故障点增加。例如,NCC跟NCP/NCF之间网络,由于多个单点故障导致断联的情况,一旦NCC失联,将导致整个DDC系统失控。而DDC系统需要集成100~500个NCP,网络规模数十倍于传统机架设备,故障半径太大,大部分客户场景无法接受。
为了解决NCC失联导致的系统失控问题,演进出一种去中心的DDC网络系统。
去中心的DDC网络系统中,包括以下特点:
1、组网方案上,去掉NCC,只保留NCP和NCF,每台NCP分别对外提供管理配置接口,分布式处理控制协议报文,如地址解析协议(Address Resolution Protocol ARP)报文;
2、原本由NCC汇总下发给各个NCP的全局表项,如媒体访问控制(Media Access Control MAC)、ARP、路由,直接由NCP同步自己生成的表项给其它NCP;以及
3、原本通过管理网络设备+NCC实现的NCP设备间信息同步(如全局表项、NCP设备和端口等信息),改为直接走fabric网卡通道。
以上,去中心(去掉NCC)的DDC网络系统,可以解决NCC失联的问题。但是,由于没有NCC汇总各台设备的信息,要分别登录不同的NCP、NCF进行设备的维护管理,维护管理的复杂度大大增加,其主要体现在以下几个方面:
1、诊断数据分散在不同的设备上;
2、要查看数据或者下发配置,需要逐一登录所有NCP、NCF设备进行操作;
3、所有设备的诊断数据收集后,部分数据还需要汇总计算,才能得出正确的结果(如一个出口队列的统计信息,需要所有入口所在NCP的统计数据加起来才能得到);以及
4、版本升级需要每台NCP、NCF单独操作。
有鉴于此,本申请各示例性实施例提供了一种去中心的网络设备系统、去中心的网络构建方法及设备,用以解决相关技术中去中心的DDC网络系统维护管理较复杂的问题。
本申请一个可能实施例的构思可概括为:将多个网络云交换网板NCF和多个网络云报文转发器NCP组成去中心的网络设备系统;其中,每个NCP包括选举模块和协议处理模块,每个NCF,用于接收以及转发来自NCP的报文;每个NCP,用于根据来自管理服务器(如可以为前述的网络管理设备)的配置报文确定自身属于候选管理NCP时(也就是说,候选管理NCP可以由所述一个或多个NCP根据接收到的配置报文来确定,即该配置报文用于接收到配置报文的NCP确定自身是否属于候选管理NCP),通过选举模块执行选举算法,以确定多个候选管理NCP中的主管理NCP;以及,根据来自选举模块的选举结果确定自身属于主管理NCP时,通过协议处理模块集中处理所述系统的管理业务报文(即,集中管理可以被定义为,主管理NCP配置为集中处理来自所述系统的所述一个或多个NCP和所述一个或多个NCF的所有管理业务报文)。本申请实施例在去中心的DDC系统的基础上,通过选举所得的主管理NCP汇总管理DDC系统中各台设备的管理业务报文,无需登录不同 的NCP、NCF进行设备的维护管理,降低了去中心的DDC系统的维护管理的复杂度,提升系统维护管理的可用性。
在介绍完本申请实施例的主要发明思想之后,下面对本申请实施例的技术方案能够适用的应用场景做一些简单介绍,需要说明的是,以下介绍的应用场景仅用于说明本申请实施例而非限定。在具体实施时,可以根据实际需要灵活地应用本申请实施例提供的技术方案。
为了便于理解本申请实施例提供的去中心的网络设备系统,下面结合附图对此进行进一步说明。
在一种可能的实施方式中,本申请实施例提供一种去中心的网络设备系统,该网络设备系统属于DDC网络。如图2为本申请实施例提供的去中心的网络设备系统的框架示意图,其中,该网络设备系统包括多个网络云交换网板NCF和多个网络云报文转发器NCP,其中,每个NCF,用于接收以及转发来自NCP的报文;每个NCP包括选举模块和协议处理模块,用于根据来自管理服务器的配置报文确定自身属于候选管理NCP时,通过选举模块执行选举算法,以确定多个候选管理NCP中的主管理NCP;以及,根据来自选举模块的选举结果确定自身属于主管理NCP时,通过协议处理模块集中处理系统的管理业务报文。
需要补充的是,候选管理NCP能够直接获取选举模块的选举结果,从而确定自身是否属于主管理NCP;并将选举结果以选举结果报文的形式经过NCF的转发,使得除候选管理NCP外的所有NCP均可知晓选举结果。
在一种可能的实施方式中,如图2中的去中心的网络设备系统,每个NCP还用于:
通过协议处理模块处理每个NCP的控制业务报文。
需要补充的是,本申请实施例中每个NCP的结构示意图如图3所示,NCP包括:命令行界面(Command Line Interface,CLI)模块、选举模块、管理报文中继模块、收发包模块、协议处理模块和ASIC芯片,其中,CLI模块用于配置NCP的IP地址,选举模块用于选举得到主管理NCP,收发包模块用 于在NCP内部接收和转发报文,协议处理模块用于处理系统的管理业务报文,管理报文中继模块用于中继外来报文,外来报文可以是管理服务器发来的报文,ASIC芯片用于接收和转发来自其他NCP和管理服务器报文。
在上述去中心的网络设备系统的NCP中,与相关技术中的NCP的不同之处至少在于,本申请实施例通过在NCP增加选举模块,选举出主管理NCP,并通过选举所得的主管理NCP汇总管理DDC系统中各台设备的管理业务报文,无需登录不同的NCP、NCF进行设备的维护管理,降低了去中心的DDC系统的维护管理的复杂度。
本申请通过管控分离的方式,解决了去中心的DDC网络系统可靠性和可用性之间的矛盾,即把受集中控制点故障影响较小的管理业务(如配置下发、故障诊断、升级)集中处理,对应上述通过主管理NCP集中处理系统的管理业务报文的步骤,管理业务报文是用于配置下发、故障诊断、系统升级的报文或者管理该系统中NCF和NCP的报文;把受集中控制点故障影响较大的控制面业务的报文(如用于地址解析协议ARP和配置边界网关协议BGP的报文)进行分布式处理。对应上述通过各个NCP独立处理系统的控制业务报文的步骤,控制业务报文是用于配置协议的报文和/或控制该系统中NCF和NCP的报文。
在一种可能的实施方式中,主管理NCP集中处理系统的管理业务报文,包括以下内容:
获取系统中NCF与NCP的配置信息;配置信息是使设备可以正常工作的信息,如全局表项信息。
以及,在确定新的NCP或NCF加入后,将配置信息分发到新的NCP或NCF;
以及,将用于管理系统中NCF与NCP的管理数据备份到备份管理NCP;
以及,接收来自管理服务器的版本文件,并将版本文件发送给对应的NCP或NCF,以使NCP或NCF基于版本文件完成升级操作。
需要说明的是,备份管理NCP是从多个候选管理NCP中选举得到的。在 登录主管理NCP设备并执行升级的指令时,本申请实施例提供三种灵活的升级方式,包括指定单台NCP/NCF设备进行升级、指定多台设备进行升级、整网一键升级所有设备三种功能选项,满足不同场景下的升级需求。
在一种可能的实施方式中,每个NCP从多个候选管理NCP中选举得到备份管理NCP的具体过程,包括:
每个NCP根据来自管理服务器的报文确定自身属于候选管理NCP时,通过选举模块执行选举算法,以确定多个候选管理NCP中的备份管理NCP;
以及,根据来自选举模块的选举结果确定自身属于备份管理NCP时,接收来自主管理NCP的管理数据进行备份;以及,确定主管理NCP发生故障时,切换为新主管理NCP。
需要补充的是,基于相同的候选管理NCP确定自身是否属于主管理NCP的原理,候选管理NCP同样能够直接获取选举模块的选举结果,从而确定自身是否属于备份管理NCP;并将选举结果以选举结果报文的形式经过NCF的转发,使得除候选管理NCP外的所有NCP均可知晓选举结果。
在上述每个NCP执行选举主管理NCP和备份管理NCP之前,本申请实施例需要先确定候选管理NCP,确定候选管理NCP的过程如下:首先确定参与选举的候选管理NCP,比如在100台NCP中选择3台NCP配置为候选管理NCP。为了避免系统版本升级全网中断的问题,需要分批次升级NCP。因此候选管理NCP需要分布到不同的升级批次,本申请实施例可以从NCP设备编号中位数往大于该中位数和小于该中位数的两边自动轮询选择(或随机选择)出候选管理NCP,或者由用户通过配置指定候选管理NCP。自动轮询选择举例如下:假设100台NCP,编号从1~100,则从中位数编号为50的NCP开始,先往前选一台编号49的NCP,再往后选一台编号51的NCP,再往前选一台编号48的NCP,再往后选一台编号52的NCP,以此类推,直到候选管理NCP数量达到预期,如果预期数量为3台,则选择编号49、编号50、编号51这三台NCP。
确定候选管理NCP后,候选管理NCP的选举模块自动在候选管理NCP 中确定主管理NCP和备份管理NCP分别由哪台NCP承担,剩下的NCP则继续作为候选管理NCP。具体选举规则算法业界已有多种成熟方案,比如bully algorithm和ring algorithm,其他算法本实施例不在此赘述。
由于参与选举的NCP越多,复杂度越高,耗时越长,本申请实施例通过上述指定少数候选管理NCP的方式,解决所有NCP参与选举决策困难的问题。
在一种可能的实施方式中,备份管理NCP还用于:
在确定主管理NCP发生故障时,通过NCF向除主管理NCP和备份管理NCP外的剩余候选管理NCP发送重新选举报文,以使剩余候选管理NCP通过选举模块执行选举算法,以确定剩余候选管理NCP中的新备份管理NCP。
当主管理NCP故障时,备份管理NCP切换为新的主管理NCP,并由选举模块在其余候选管理NCP中重新选举一台NCP作为备份管理NCP,避免了主管理NCP故障导致系统瘫痪的问题,提高了系统的可用性。
当需要通过带内管理(in-band management)DDC系统时,可能出现管理服务器接入NCP2,但主管理NCP是NCP1的情况,此时,管理服务器跟主管理NCP通信的报文会先送到NCP2的CPU。本申请实施例在每个NCP增加了管理报文中继模块,如上图3所示,通过报文中继的方式,能够解决跨NCP带内远程管理的问题,即通过NCP2的管理报文中继模块把报文中继给NCP1。本领域技术人员可以理解的,带内管理是一种网络管理方式,其中网络的管理控制信息与用户网络的承载业务信息通过同一个逻辑信道传送。
在一种可能的实施方式中,每个NCP执行报文中继步骤,其流程如图4所示,包括以下步骤。
在步骤401中,接收外来报文,并通过管理报文中继模块根据外来报文的目的IP地址查询IP地址与NCP的映射表。其中,外来报文可以是管理报文。
在步骤402中,若确定外来报文的目的IP地址对应的NCP是本机,则通过协议处理模块对外来报文进行处理。
在步骤403中,若确定外来报文的目的IP地址对应的NCP非本机,则将 外来报文发送给目的IP地址对应的NCP进行处理。
例如,报文中继过程的示意图如图5所示,NCP2收到管理服务器请求访问NCP1的报文,将该请求报文发送NCP2的CPU(CPU中包含了选举模块、管理报文中继模块等模块),NCP2发现访问的目的IP地址是其它NCP的IP地址,管理报文中继模块根据请求报文的目的IP地址查询IP地址与NCP的映射表,确定请求报文的目的IP地址对应的NCP为NCP1的IP地址,将报文中继给NCP1的CPU;NCP1收到请求报文后,发现访问的IP地址是本机的管理地址,通过协议处理模块对该请求报文进行处理。
在另一种可能的实施方式中,每个NCP还用于:
在确定主管理NCP后,通过选举模块将主管理NCP对应的主管理IP发送到管理报文中继模块,以使管理报文中继模块将IP地址与NCP的映射表中主管理IP对应的NCP设置为主管理NCP;
以及,在主备切换后,通过选举模块将新主管理NCP对应的新主管理IP发送到管理报文中继模块,以使管理报文中继模块更新IP地址与NCP的映射表中主管理IP对应的NCP。
需要说明的是,选举模块执行选举步骤之前,主管理IP对应的NCP默认为空,在选举模块选出主管理NCP后,选举模块通告管理报文中继模块更新主管理IP对应的NCP为主管理NCP,主备切换后,选举模块通告管理报文中继模块将IP地址与NCP的映射表中的新主管理IP对应的NCP更新为新的主管理NCP。
根据上述去中心的网络设备系统,本申请实施例还提供一种去中心的网络构建方法,应用于上述去中心的网络设备系统的网络云报文转发器NCP,包括以下内容:
接收来自管理服务器的配置报文,并在根据配置报文确定自身属于候选管理NCP时,通过选举模块执行选举算法,以确定多个候选管理NCP中的主管理NCP;以及,
获取来自选举模块的选举结果,并在根据选举结果确定自身属于主管理 NCP时,通过协议处理模块集中处理系统的管理业务报文。
在一种可能的实施方式中,所述方法还包括:
获取所述系统中NCF与NCP的配置信息;
在确定新的NCP或NCF加入后,将所述配置信息分发到新的NCP或新的NCF;
以及,将用于管理所述系统中NCF与NCP的管理数据备份到备份管理NCP,所述备份管理NCP是从所述多个候选管理NCP中选举得到的。
基于相同的发明构思,本申请实施例还提供一种去中心的网络构建装置,应用于上述去中心的网络设备系统的网络云报文转发器NCP,如图6所示,所述装置600包括:
选举模块601,被配置为接收来自管理服务器的配置报文,并在根据所述配置报文确定自身属于候选管理NCP时,通过选举模块执行选举算法,以确定多个候选管理NCP中的主管理NCP;以及,
管理模块602,被配置为获取来自选举模块的选举结果,并在根据所述选举结果确定自身属于所述主管理NCP时,通过协议处理模块集中处理所述系统的管理业务报文。
综上所述,本申请一个可选的实施例,在去中心的DDC系统的基础上,通过选举所得的主管理NCP汇总管理DDC系统中各台设备的管理业务报文,无需登录不同的NCP、NCF进行设备的维护管理,降低了去中心的DDC系统的维护管理的复杂度,提升系统维护管理的可用性。
下面参照图7来描述根据本申请实施方式的电子设备130。图7显示的电子设备130仅仅是一个示例,不应对本申请实施例的功能和使用范围带来任何限制。
如图7所示,电子设备130以通用电子设备的形式表现。电子设备130的组件可以包括但不限于:上述至少一个处理器131、上述至少一个存储器132、连接不同系统组件(包括存储器132和处理器131)的总线133。
总线133表示几类总线结构中的一种或多种,包括存储器总线或者存储 器控制器、外围总线、处理器或者使用多种总线结构中的任意总线结构的局域总线。
存储器132可以包括非易失性存储介质或易失性存储器形式的可读介质,例如随机存取存储器(RAM)1321和/或高速缓存存储器1322,还可以进一步包括只读存储器(ROM)1323。
存储器132还可以包括具有一组(至少一个)程序模块1324的程序/实用工具1325,这样的程序模块1324包括但不限于:操作系统、一个或者多个应用程序、其它程序模块以及程序数据,这些示例中的每一个或某种组合中可能包括网络环境的实现。
电子设备130也可以与一个或多个外部设备134(例如键盘、指向设备等)通信,还可与一个或者多个使得用户能与电子设备130交互的设备通信,和/或与使得该电子设备130能与一个或多个其它电子设备进行通信的任何设备(例如路由器、调制解调器等等)通信。这种通信可以通过输入/输出(I/O)接口135进行。并且,电子设备130还可以通过网络适配器136与一个或者多个网络(例如局域网(LAN),广域网(WAN)和/或公共网络,例如因特网)通信。如图所示,网络适配器136通过总线133与用于电子设备130的其它模块通信。应当理解,尽管图中未示出,可以结合电子设备130使用其它硬件和/或软件模块,包括但不限于:微代码、设备驱动器、冗余处理器、外部磁盘驱动阵列、RAID系统、磁带驱动器以及数据备份存储系统等。
在示例性实施例中,本申请还提供了一种包括指令的计算机可读存储介质,例如包括指令的存储器132,上述指令可由电子设备130的处理器131执行以完成上述去中心的网络构建方法。可选地,计算机可读存储介质可以是非临时性计算机可读存储介质,例如,所述非临时性计算机可读存储介质可以是ROM、随机存取存储器(RAM)、CD-ROM、磁带、软盘和光数据存储设备等。
在示例性实施例中,还提供一种计算机程序产品,包括计算机程序,所述计算机程序被处理器131执行时实现如本申请提供的去中心的网络构建方 法。
本领域内的技术人员应明白,本申请的实施例可提供为方法、系统、或计算机程序产品。因此,本申请可采用完全硬件实施例、完全软件实施例、或结合软件和硬件方面的实施例的形式。而且,本申请可采用在一个或多个其中包含有计算机可用程序代码的计算机可用存储介质(包括但不限于磁盘存储器、CD-ROM、光学存储器等)上实施的计算机程序产品的形式。
本申请是参照根据本申请的方法、设备(系统)、和计算机程序产品的流程图和/或方框图来描述的。应理解可由计算机程序指令实现流程图和/或方框图中的每一流程和/或方框、以及流程图和/或方框图中的流程和/或方框的结合。可提供这些计算机程序指令到通用计算机、专用计算机、嵌入式处理机或其他可编程数据处理设备的处理器以产生一个机器,使得通过计算机或其他可编程数据处理设备的处理器执行的指令产生用于实现在流程图一个流程或多个流程和/或方框图一个方框或多个方框中指定的功能的装置。
这些计算机程序指令也可存储在能引导计算机或其他可编程数据处理设备以特定方式工作的计算机可读存储器中,使得存储在该计算机可读存储器中的指令产生包括指令装置的制造品,该指令装置实现在流程图一个流程或多个流程和/或方框图一个方框或多个方框中指定的功能。
这些计算机程序指令也可装载到计算机或其他可编程数据处理设备上,使得在计算机或其他可编程设备上执行一系列操作步骤以产生计算机实现的处理,从而在计算机或其他可编程设备上执行的指令提供用于实现在流程图一个流程或多个流程和/或方框图一个方框或多个方框中指定的功能的步骤。
显然,本领域的技术人员可以对本申请进行各种改动和变型而不脱离本申请的精神和范围。这样,倘若本申请的这些修改和变型属于本申请权利要求及其等同技术的范围之内,则本申请也意图包含这些改动和变型在内。

Claims (19)

  1. 一种去中心的网络设备系统,包括:
    一个或多个网络云报文转发器NCP,包括一个或多个候选管理NCP及主管理NCP,其中,所述一个或多个候选管理NCP由所述一个或多个NCP根据接收到的配置报文来确定,所述主管理NCP是通过选举算法从所述一个或多个候选管理NCP中确定,以及所述主管理NCP配置为集中处理所述系统的管理业务报文;以及
    一个或多个网络云交换网板NCF,配置为接收以及转发来自所述一个或多个NCP的报文,所述报文包括通知所述选举算法的结果的报文和管理业务报文。
  2. 根据权利要求1所述的系统,其中,所述一个或多个NCP还包括备份管理NCP,所述备份管理NCP是通过选举算法从所述一个或多个候选管理NCP中确定且与所述主管理NCP不同,所述主管理NCP还配置为:
    将用于管理所述一个或多个NCF与所述一个或多个NCP的管理数据备份到所述备份管理NCP。
  3. 根据权利要求1所述的系统,其中,所述管理业务报文包括所述一个或多个NCF与所述一个或多个NCP的配置信息,所述主管理NCP还配置为:
    获取所述系统中所述一个或多个NCF与所述一个或多个NCP的配置信息;以及
    在确定新的NCP或NCF加入后,将所述配置信息分发到所述新的NCP或NCF。
  4. 根据权利要求1所述的系统,其中,所述管理业务报文包括所述一个或多个NCF与所述一个或多个NCP的版本文件,所述主管理NCP还配置为:
    接收所述版本文件,并将所述版本文件发送给待升级的NCP和/或待升级的NCF,所述待升级的NCP为所述一个或多个NCP中的至少一个,所述待升级的NCF为所述一个或多个NCF中的至少一个。
  5. 根据权利要求2所述的系统,其中,所述备份管理NCP还被配置为:
    接收来自所述主管理NCP的管理数据进行备份;以及
    确定所述主管理NCP发生故障时,切换为新主管理NCP。
  6. 根据权利要求5所述的系统,其中,所述备份管理NCP还配置为:
    在确定所述主管理NCP发生故障时,通过与所述备份管理NCP连接的NCF向所述一个或多个候选管理NCP发送重新选举报文,以使所述一个或多个候选管理NCP通过所述选举算法从所述一个或多个候选管理NCP中确定新备份管理NCP。
  7. 根据权利要求5所述的系统,其中,所述主管理NCP还配置为:
    在从所述一个或多个候选管理NCP中被确定为所述主管理NCP后,将所述主管理NCP对应的IP地址与所述主管理NCP的对应关系更新至IP地址与NCP的映射表中;以及
    在所述备份管理NCP被切换为新主管理NCP后,将所述新主管理NCP对应的IP地址与所述新主管理NCP的对应关系更新至所述IP地址与NCP的映射表中。
  8. 根据权利要求1所述的系统,其中,所述一个或多个NCP还配置为:
    分别处理所述一个或多个NCP的控制业务报文,所述控制业务报文包括用于配置协议的报文或控制所述系统中所述一个或多个NCF和所述一个或多个NCP的报文。
  9. 根据权利要求1所述的系统,其中,所述一个或多个NCP包括第一NCP,所述第一NCP还配置为:
    接收外来报文,并根据所述外来报文的目的互联网协议IP地址通过查询IP地址与NCP的映射表确定是否处理所述外来报文;以及
    若确定所述外来报文的目的IP地址对应的NCP非所述第一NCP,则将所述外来报文发送给所述目的IP地址对应的所述NCP进行处理。
  10. 一种去中心的网络构建方法,其中,应用于去中心的网络设备系统 的网络云报文转发器NCP,所述方法包括:
    接收配置报文,并在根据所述配置报文确定自身属于候选管理NCP时,通过选举算法确定一个或多个候选管理NCP中的主管理NCP;以及,
    获取选举结果,并在根据所述选举结果确定自身属于所述主管理NCP时,集中处理所述系统的管理业务报文。
  11. 根据权利要求10所述的方法,其中,所述方法还包括:
    获取所述系统中一个或多个NCF与一个或多个NCP的配置信息;
    在确定新的NCP或NCF加入后,将所述配置信息分发到新的NCP或NCF;以及
    将用于管理所述系统中所述一个或多个NCF与所述一个或多个NCP的管理数据备份到备份管理NCP,所述备份管理NCP是从所述一个或多个候选管理NCP中选举得到的。
  12. 根据权利要求10所述的方法,其中,所述方法还包括:
    接收所述一个或多个NCF与所述一个或多个NCP的版本文件,并将所述版本文件发送给对应的NCP或NCF,以使所述一个或多个NCP或一个或多个NCF基于所述版本文件完成升级操作。
  13. 根据权利要求10所述的方法,其中,所述方法还包括:
    根据来自管理服务器的配置报文确定自身属于候选管理NCP时,通过选举算法,以确定多个候选管理NCP中的备份管理NCP;以及,
    根据选举结果确定自身属于所述备份管理NCP时,接收来自所述主管理NCP的管理数据进行备份;以及,确定所述主管理NCP发生故障时,切换为新主管理NCP。
  14. 根据权利要求13所述的方法,其中,所述方法还包括:
    在确定所述主管理NCP发生故障时,通过与自身对应的NCF向所述一个或多个候选管理NCP发送重新选举报文,以使所述一个或多个候选管理NCP通过所述选举算法从所述一个或多个候选管理NCP中确定新备份管理NCP。
  15. 根据权利要求11或12所述的方法,其中,所述方法还包括:
    在从所述一个或多个候选管理NCP中被确定为主管理NCP后,将所述主管理NCP的IP地址与所述主管理NCP的对应关系更新至IP地址与NCP的映射表中;以及
    在所述备份管理NCP被切换为新主管理NCP后,将所述新主管理NCP对应的IP地址与所述新主管理NCP的对应关系更新至所述IP地址与NCP的映射表中。
  16. 根据权利要求10所述的方法,其中,所述方法还包括:
    处理所述一个或多个NCP的控制业务报文,所述控制业务报文包括用于配置协议的报文或控制所述系统中所述一个或多个NCF和所述一个或多个NCP的报文。
  17. 根据权利要求10所述的方法,其中,所述方法还包括:
    接收外来报文,并根据所述外来报文的目的IP地址查询IP地址与NCP的映射表确定是否处理所述外来报文;
    若确定所述外来报文的目的IP地址对应的NCP非本机,则将所述外来报文发送给所述目的IP地址对应的所述NCP进行处理。
  18. 一种网络云报文转发器NCP,包括:
    处理器和存储器;
    其中,所述存储器,用于存储所述处理器可执行指令;
    所述处理器被配置为执行所述指令以实现如权利要求10至17任一项所述的去中心的网络构建方法。
  19. 一种计算机可读存储介质,其中,其存储有可由计算机执行的计算机程序,当所述程序在所述计算机上运行时,使得所述计算机执行如权利要求10至17任一所述的方法。
PCT/CN2024/089374 2023-05-08 2024-04-23 去中心的网络设备系统、方法、网络云报文转发器及可读存储介质 WO2024230474A1 (zh)

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202310511065.1 2023-05-08

Publications (1)

Publication Number Publication Date
WO2024230474A1 true WO2024230474A1 (zh) 2024-11-14

Family

ID=

Similar Documents

Publication Publication Date Title
US9148389B2 (en) System and method for a virtual chassis system
US9148391B2 (en) System and method for a pass thru mode in a virtual chassis system
US9148390B2 (en) System and method for virtual chassis split prevention
US9071508B2 (en) Distributed fabric management protocol
EP3316555B1 (en) Mac address synchronization method, device and system
CN112615778A (zh) 报文转发方法、装置、路由器、存储介质及系统
WO2020114017A1 (zh) 数据中心流量互通方法、装置、设备及存储介质
US20210286747A1 (en) Systems and methods for supporting inter-chassis manageability of nvme over fabrics based systems
JP6109954B2 (ja) 仮想シャーシシステムにおけるパススルーモードのためのシステムおよび方法
EP2658172A1 (en) Multicast method and multicast device
CN113839862A (zh) Mclag邻居之间同步arp信息的方法、系统、终端及存储介质
US20120063299A1 (en) Highly available virtual packet network device
WO2007139542A1 (en) Uninterrupted network control message generation during local node outages
CN100508468C (zh) 分布式节点动态管理数据自适应交换方法
US11418382B2 (en) Method of cooperative active-standby failover between logical routers based on health of attached services
WO2024230474A1 (zh) 去中心的网络设备系统、方法、网络云报文转发器及可读存储介质
JP2016501463A (ja) 管理アクションが仮想シャーシの分割をトリガーするという警告を発行するかどうかが決定される、ネットワークノード、および仮想シャーシシステム内で動作可能であるノードにおける方法
EP1712067B1 (en) A method, apparatus and system of organizing servers
EP4184822A1 (en) Method and apparatus for keeping user terminal alive
CN114124803A (zh) 设备管理方法、装置、电子设备及存储介质
JP2000244526A (ja) 多重化したネットワーク接続装置システム
CN118921268A (zh) 去中心的网络设备系统、去中心的网络构建方法及设备
US11979286B1 (en) In-service software upgrade in a virtual switching stack
JP2000004231A (ja) 多重化された端末経路を制御する伝送装置
CN117978649A (zh) 一种通信方法及相关设备