WO2016095750A1 - 虚拟交换集群中通信的方法及装置 - Google Patents

虚拟交换集群中通信的方法及装置 Download PDF

Info

Publication number
WO2016095750A1
WO2016095750A1 PCT/CN2015/097034 CN2015097034W WO2016095750A1 WO 2016095750 A1 WO2016095750 A1 WO 2016095750A1 CN 2015097034 W CN2015097034 W CN 2015097034W WO 2016095750 A1 WO2016095750 A1 WO 2016095750A1
Authority
WO
WIPO (PCT)
Prior art keywords
stack link
control
plane stack
packet
control packet
Prior art date
Application number
PCT/CN2015/097034
Other languages
English (en)
French (fr)
Inventor
韩冲
Original Assignee
中兴通讯股份有限公司
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 中兴通讯股份有限公司 filed Critical 中兴通讯股份有限公司
Publication of WO2016095750A1 publication Critical patent/WO2016095750A1/zh

Links

Images

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L12/00—Data switching networks
    • H04L12/28—Data switching networks characterised by path configuration, e.g. LAN [Local Area Networks] or WAN [Wide Area Networks]
    • H04L12/46—Interconnection of networks

Definitions

  • the present invention relates to the field of communications technologies, and in particular, to a method and apparatus for communication in a virtual switching cluster.
  • VSC Virtual Switch Cluster
  • a Virtual Switch Cluster is a network virtualization technology application of a communication device. It connects multiple independent physical devices to each other through a control plane stack link and a forwarding plane stack link. Virtual cluster device. The physical devices discover each other through the topology protocol, and select the master device and the slave device according to the master/slave election policy to perform unified management.
  • VSC members need to have independent physical links to ensure stack link communication between the control plane stack link and the forwarding plane.
  • the forwarding plane stacking link between VSC members is used to forward data service packets. There are usually multiple links to ensure high bandwidth and data protection for data forwarding. It is separate from the control plane stacking link. .
  • the control plane stack link is used to forward control packets between devices, such as configuration management information, topology discovery information, and data synchronization information. Because the traffic of the packets that need to be forwarded on the control plane stack link is not too large, there is no excessive physical link. Usually only one or two physical links are connected. In this way, when communication interruption occurs in the control plane stack link, the VSC system cannot be guaranteed to operate normally, the system performance is insufficient, and the stability and reliability are poor.
  • the main purpose of the present invention is to provide a method and apparatus for communication in a virtual switching cluster, which aims to solve the technical problem of poor stability and reliability of a control plane stack link when communication is interrupted.
  • the present invention provides a method for communication in a virtual switching cluster, and the method for communicating in the virtual switching cluster includes the following steps:
  • control packet on the control plane stack link is switched to the forwarding plane stack link for transmission.
  • the step of detecting a communication state of a control plane stack link between a local end member and a peer end member in the virtual switching cluster includes:
  • the local end member sends a detection packet to the peer member.
  • the step of switching the control packet on the control plane stack link to the forwarding plane stack link for sending includes:
  • the proxy CPU is obtained, and the control packet is captured by the proxy CPU, and the captured control packet is switched to the forwarding plane stack link for transmission.
  • the acquiring proxy CPU, the step of the proxy CPU capturing the control packet, and switching the captured control packet to the forwarding plane stack link for sending includes:
  • proxy CPU captures the control packet and parses the control packet
  • the captured control packet is added to the proxy CPU flag and switched to the forwarding plane stack link for transmission.
  • the method further comprises:
  • control packet When it is detected that the control plane stack link communication is restored, the control packet is switched to the forwarding plane stack link for transmission.
  • the present invention further provides an apparatus for communicating in a virtual switching cluster, where the apparatus for communicating in the virtual switching cluster includes:
  • the detecting module is configured to detect a communication state of a control plane stack link between the local end member and the peer member in the virtual switching cluster;
  • the switching module is configured to switch the control packet on the control plane stack link to the forwarding plane stack link for sending when the control plane stack link communication interruption is detected.
  • the detecting module comprises:
  • a sending unit configured to send, by the local end member, a detection packet to the peer member
  • the determining unit is configured to determine whether the local end member receives the response message corresponding to the detection packet returned by the peer member within a predetermined time.
  • the switching module comprises:
  • the processing unit is configured to: when the control plane stack link communication interruption is detected, set the spanning tree status of the local end member and the opposite end member to congestion, and clear the MAC address learned by the respective physical interface;
  • a switching unit configured to acquire a proxy CPU, and the proxy CPU captures the control packet, and the captured control The packet is switched to the forwarding plane stack link for transmission.
  • the switching unit comprises:
  • the parsing subunit is configured to obtain a proxy CPU, and the proxy CPU captures the control packet and parses the control packet;
  • a determining subunit configured to determine, by using the MAC address of the control packet obtained by the parsing, whether the control packet is to be forwarded;
  • the method further includes: a recovery module configured to switch the control packet to the forwarding plane stack link for transmission when detecting that the control plane stack link communication is restored.
  • a recovery module configured to switch the control packet to the forwarding plane stack link for transmission when detecting that the control plane stack link communication is restored.
  • the method and the device for communicating in the virtual switching cluster of the present invention when detecting that the control plane stack link is interrupted or fails, the control packet of the control plane stack link is switched to the forwarding plane stack link for transmission, and is forwarded.
  • the plane stacking link is used to implement the control packet communication of the control plane stack link, ensure the continuity of the control plane stack link communication, and improve the stability and reliability of the virtual switch cluster.
  • FIG. 1 is a schematic flowchart of a first embodiment of a method for communication in a virtual switching cluster according to the present invention
  • FIG. 2 is a schematic diagram of a refinement process of step S101 in FIG. 1;
  • step S102 in FIG. 1 is a schematic flowchart of the refinement of step S102 in FIG. 1;
  • FIG. 4 is a schematic flowchart of a second embodiment of a method for communication in a virtual switching cluster according to the present invention.
  • FIG. 5 is a schematic diagram of functional modules of a first embodiment of a device for communicating in a virtual switching cluster according to the present invention
  • FIG. 6 is a schematic diagram of a refinement function module of the detection module of FIG. 5;
  • FIG. 7 is a schematic diagram of a refinement function module of the switching module in FIG. 5;
  • FIG. 8 is a schematic diagram of functional modules of a second embodiment of a device for communication in a virtual switching cluster according to the present invention.
  • the present invention provides a method for communication in a virtual switching cluster.
  • the method includes:
  • Step S101 detecting communication of a control plane stack link between the local end member and the peer member in the virtual switching cluster status
  • the control plane stacking link between the local end member and the peer member sends a control packet between the two through the control plane stack link.
  • the member sends a detection packet to the peer member to detect the communication status between the two through the control plane stack link, and then receives the feedback from the peer member.
  • the feedback packet is used to detect whether the communication status of the control plane stack link is detected. normal.
  • the control plane stack link is disconnected, and the control plane stack between the local end member and the peer end member is considered. Link communication is interrupted.
  • Step S102 When it is detected that the control plane stack link communication is interrupted, the control packet on the control plane stack link is switched to the forwarding plane stack link for transmission.
  • the communication between the members of the virtual switch cluster is controlled by the soft forwarding plane stack link to ensure the control of the members of the virtual switch cluster. Packets can be forwarded efficiently, enabling normal communication between members of the virtual switching cluster.
  • control packet of the control plane stack link when detecting that the control plane stack link is interrupted or fails, the control packet of the control plane stack link is quickly switched to the forwarding plane stack link for transmission, and the forwarding plane stacking link is used.
  • the communication of the control packets of the control plane stack link is continued to ensure the continuity of the control plane stack link communication, and the stability and reliability of the virtual switch cluster can be improved.
  • the step S101 includes:
  • Step S1011 The local end member sends a detection packet to the peer member.
  • step S1012 it is determined whether the local end member receives the response packet corresponding to the detection packet returned by the peer member within a predetermined time.
  • the local end member and the peer end member send detection packets to each other, and receive the packets fed back by the peer, and the feedback packets are used to detect the communication state of the control plane stack link.
  • the link detection packet of the local end member and the peer end member is enabled to be sent and received, and the local end member and the peer end member send detection packets to the other party from the control plane stack link every second, and listen to the feedback of the other party. Feedback message.
  • the detection packet defines a private feature: the Ethernet protocol number of the detection packet is 0Xf002, and the destination MAC address is 0x0102-0304-0506; within a predetermined time (for example, 3 seconds, it may be modified according to the actual value to a reasonable value. If the feedback message fed back by the other party is not received, the control plane stack link communication with the other party is considered to be interrupted.
  • the step S102 includes:
  • Step S1021 When detecting that the control plane stack link communication is interrupted, the local end member and the opposite end member are respectively The spanning tree status is set to congestion, and the MAC address learned by the respective physical interface is cleared;
  • step S1022 the proxy CPU is obtained, and the control packet is captured by the proxy CPU, and the captured control packet is switched to the forwarding plane stack link for transmission.
  • the state of the spanning tree of the local end member and the peer end member is set to be congested to prohibit forwarding of data, and only the detection packet can be forwarded to prevent the control plane from being stacked.
  • a loop problem occurs when the link is restored.
  • the MAC address learned by the physical interface of the switch switch between the two members and the peer member in the control plane stack link is cleared.
  • the known unicast control packet becomes unknown unicast. Control messages.
  • the proxy CPU is selected, and the two members respectively enable the capture of the Ethernet port of the proxy CPU, and the control packet is captured as a proxy and sent by the CPU with the forwarding plane stack link connection. Controls the packet to the port of the switch that is acting on the Ethernet port of the proxy CPU.
  • the proxy CPU captures the control packet and parses the MAC address of the control packet obtained by the parsing to determine whether it is a control packet to be forwarded; if yes, the captured control packet is added to the proxy CPU. The flag bit is switched to the forwarding plane stack link for transmission.
  • the local end resolves the control packet that needs to be forwarded to the other party.
  • the MAC address assigned to each proxy CPU is The unique MAC address of the member is also in the range of the area.
  • the MAC address of the control packet can be sent to the peer.
  • the packets sent to the peer are: The MAC address is a broadcast or multicast packet; the MAC address is unicast and is within the range of addresses to which the peer member belongs.
  • the proxy CPU flag is added to the control packet, so that the peer member can easily distinguish after receiving the control packet.
  • the control packet of the control plane stack link that is captured by the local end member and the peer member is transmitted through the forwarding plane stack link.
  • the control packet is sent to the peer end by forwarding the transmission of the plane stack link.
  • the peer member determines whether the control packet is sent by the control plane stack link according to whether there is a proxy CPU flag in the control packet.
  • the two members respectively send control messages of the control plane stack link sent by the other party to their own control plane communication system.
  • the control packet of the control plane stack link needs to be sent to its own control plane communication system through the packet processing of the CPU network port.
  • the forwarding plane stack link can be sent to send the control packet on the control plane stack link, and finally sent to the corresponding control plane communication system. To ensure continuity of control plane stack link communication.
  • the method further includes:
  • Step 103 When it is detected that the control plane stack link communication is restored, the control packet is switched to the forwarding plane stack link for transmission.
  • control plane stack link when the control plane stack link receives the feedback packet of the detected packet, it is determined that the control plane stack link is restored and connected, and the local end member and the peer member are respectively enabled. Control the packet capture task of the control plane stack link, delete the MAC address of the port on which the switch switch is connected to the proxy CPU network port, and set the physical interface spanning tree status of the two member control plane stack links to forward (FORWARD). Status, restore the control plane stack link to forward control messages, and keep communication going.
  • FORWARD forward
  • the present invention also provides an apparatus for communicating in a virtual switching cluster.
  • the apparatus for communicating in the virtual switching cluster includes:
  • the detecting module 101 is configured to detect a communication state of a control plane stack link between the local end member and the peer member in the virtual switching cluster;
  • the control plane stacking link between the local end member and the peer member sends a control packet between the two through the control plane stack link.
  • the member sends a detection packet to the peer member to detect the communication status between the two through the control plane stack link, and then receives the feedback from the peer member.
  • the feedback packet is used to detect whether the communication status of the control plane stack link is detected. normal.
  • the control plane stack link is disconnected, and the control plane stack between the local end member and the peer end member is considered. Link communication is interrupted.
  • the switching module 102 is configured to switch the control packet on the control plane stack link to the forwarding plane stack link for sending when the control plane stack link communication interruption is detected.
  • the communication between the members of the virtual switch cluster is controlled by the soft forwarding plane stack link to ensure the control of the members of the virtual switch cluster. Packets can be forwarded efficiently, enabling normal communication between members of the virtual switching cluster.
  • control packet of the control plane stack link when detecting that the control plane stack link is interrupted or fails, the control packet of the control plane stack link is quickly switched to the forwarding plane stack link for transmission, and the forwarding plane stacking link is used.
  • the communication of the control packets of the control plane stack link is continued to ensure the continuity of the control plane stack link communication, and the stability and reliability of the virtual switch cluster can be improved.
  • the detecting module 101 includes:
  • the sending unit 1011 is configured to send, by the local end member, a detection packet to the peer member;
  • the determining unit 1012 is configured to determine whether the local end member receives the response message corresponding to the detection packet returned by the peer member within a predetermined time.
  • the local end member and the peer end member send detection packets to each other, and receive the packets fed back by the other party.
  • the feedback message is used to detect the communication status of the control plane stack link.
  • the link detection packet of the local end member and the peer end member is enabled to be sent and received, and the local end member and the peer end member send detection packets to the other party from the control plane stack link every second, and listen to the feedback of the other party. Feedback message.
  • the detection packet defines a private feature: the Ethernet protocol number of the detection packet is 0Xf002, and the destination MAC address is 0x0102-0304-0506; within a predetermined time (for example, 3 seconds, it may be modified according to the actual value to a reasonable value. If the feedback message fed back by the other party is not received, the control plane stack link communication with the other party is considered to be interrupted.
  • the switching module 102 includes:
  • the processing unit 1021 is configured to: when the control plane stack link communication interruption is detected, set the spanning tree status of the local end member and the opposite end member to congestion, and clear the MAC address learned by the respective physical interface. ;
  • the switching unit 1022 is configured to acquire the proxy CPU, and the proxy CPU captures the control packet, and the captured control packet is switched to the forwarding plane stack link for transmission.
  • the state of the spanning tree of the local end member and the peer end member is set to be congested to prohibit forwarding of data, and only the detection packet can be forwarded to prevent the control plane from being stacked.
  • a loop problem occurs when the link is restored.
  • the MAC address learned by the physical interface of the switch switch between the two members and the peer member in the control plane stack link is cleared.
  • the known unicast control packet becomes unknown unicast. Control messages.
  • the proxy CPU is selected, and the two members respectively enable the capture of the Ethernet port of the proxy CPU, and the control packet is captured as a proxy and sent by the CPU with the forwarding plane stack link connection. Controls the packet to the port of the switch that is acting on the Ethernet port of the proxy CPU.
  • the proxy CPU captures the control packet and parses the MAC address of the control packet obtained by the parsing to determine whether it is a control packet to be forwarded; if yes, the captured control packet is added to the proxy CPU. The flag bit is switched to the forwarding plane stack link for transmission.
  • the local end resolves the control packet that needs to be forwarded to the other party.
  • the MAC address assigned to each proxy CPU is The unique MAC address of the member is also in the range of the area.
  • the MAC address of the control packet can be sent to the peer.
  • the packets sent to the peer are: The MAC address is a broadcast or multicast packet; the MAC address is unicast and is within the range of addresses to which the peer member belongs.
  • the proxy CPU flag is added to the control packet, so that the peer member can easily distinguish after receiving the control packet.
  • the control packet of the control plane stack link that is captured by the local end member and the peer member is transmitted through the forwarding plane stack link.
  • the control packet is sent to the peer end by forwarding the transmission of the plane stack link.
  • the peer member determines whether the control packet is sent by the control plane stack link according to whether there is a proxy CPU flag in the control packet.
  • the two members respectively send control messages of the control plane stack link sent by the other party to their own control plane communication system.
  • the control packet of the control plane stack link needs to be sent to its own control plane communication system through the packet processing of the CPU network port.
  • the forwarding plane stack link can be sent to send the control packet on the control plane stack link, and finally sent to the corresponding control plane communication system. To ensure continuity of control plane stack link communication.
  • a recovery module 103 configured to: when detecting the control plane stack link communication recovery, The control packet is switched to the forwarding plane stack link for transmission.
  • control plane stack link when the control plane stack link receives the feedback packet of the detected packet, it is determined that the control plane stack link is restored and connected, and the local end member and the peer member are respectively enabled. Control the packet capture task of the control plane stack link, delete the MAC address of the port on which the switch switch is connected to the proxy CPU network port, and set the physical interface spanning tree status of the two member control plane stack links to forward (FORWARD). Status, restore the control plane stack link to forward control messages, and keep communication going.
  • FORWARD forward
  • the method and apparatus for communication in a virtual switching cluster provided by the embodiments of the present invention have the following beneficial effects: when detecting that the control plane stack link is interrupted or fails, the control packet of the control plane stack link is quickly configured. Switching to the forwarding plane stacking link to transmit, and forwarding the control packet of the control plane stack link by the forwarding plane stacking link, ensuring the continuity of the control plane stack link communication, and improving the stability of the virtual switching cluster Sex and reliability.

Landscapes

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

Abstract

本发明公开了一种虚拟交换集群中通信的方法及装置，所述虚拟交换集群中通信的方法包括以下步骤：检测虚拟交换集群中的本端成员与对端成员之间的控制平面堆叠链路的通信状态；当检测到所述控制平面堆叠链路通信中断时，将所述控制平面堆叠链路上的控制报文切换至转发平面堆叠链路进行发送。本发明能够在控制平面堆叠链路通信中断时保证通信的连续性，提高系统的稳定性及可靠性。

Description

虚拟交换集群中通信的方法及装置 技术领域
本发明涉及通信技术领域，尤其涉及一种虚拟交换集群中通信的方法及装置。
背景技术
虚拟集群交换机系统(Virtual Switch Cluster，简称为VSC)是通信设备的一种网络虚拟化技术的应用，其将多台独立运行的物理设备通过控制平面堆叠链路和转发平面堆叠链路连接成一台虚拟集群设备。各物理设备之间通过拓扑协议发现对方，并根据主备选举策略选出主设备(Master)和从设备(Slave)，进行统一管理。
由于多台物理设备虚拟化为一台虚拟集群设备，VSC成员间需要有独立的物理链路来保证控制平面堆叠链路和转发平面的堆叠链路通信。VSC成员间的转发平面堆叠链路用来转发数据业务报文的，通常会有多条链路来保证数据转发的高带宽需要和冗余保护的需要，其与控制平面堆叠链路是分开的。控制平面堆叠链路用来转发一些设备间的控制报文，如配置管理信息，拓扑发现信息，数据同步信息等等。由于控制平面堆叠链路需要转发的报文流量不会太大，所以不会投入过多的物理链路，通常只有一两条物理链路来连接就可以了。这样在控制平面堆叠链路都出现通信中断的时候，就无法保证VSC系统正常运行，系统性能存在不足，稳定性及可靠性都较差。
上述内容仅用于辅助理解本发明的技术方案，并不代表承认上述内容是现有技术。
发明内容
本发明的主要目的在于提供一种虚拟交换集群中通信的方法及装置，旨在解决控制平面堆叠链路出现通信中断时系统的稳定性及可靠性差的技术问题。
为实现上述目的，本发明提供一种虚拟交换集群中通信的方法，所述虚拟交换集群中通信的方法包括以下步骤：
检测虚拟交换集群中的本端成员与对端成员之间的控制平面堆叠链路的通信状态；
当检测到所述控制平面堆叠链路通信中断时，将所述控制平面堆叠链路上的控制报文切换至转发平面堆叠链路进行发送。
优选地，所述检测虚拟交换集群中的本端成员与对端成员之间的控制平面堆叠链路的通信状态的步骤包括：
由所述本端成员向对端成员发送检测报文；
判断所述本端成员是否在预定时间内接收到所述对端成员返回的所述检测报文对应的响 应报文。
优选地，当检测到所述控制平面堆叠链路通信中断时，将所述控制平面堆叠链路上的控制报文切换至转发平面堆叠链路进行发送的步骤包括：
当检测到所述控制平面堆叠链路通信中断时，将所述本端成员与对端成员各自的生成树状态设为拥塞，并清除各自的物理接口学习到的MAC地址；
获取代理CPU，由所述代理CPU抓取所述控制报文，将所抓取的控制报文切换至转发平面堆叠链路进行发送。
优选地，所述获取代理CPU，由所述代理CPU抓取所述控制报文，将所抓取的控制报文切换至转发平面堆叠链路进行发送的步骤包括：
获取代理CPU，由所述代理CPU抓取所述控制报文并进行解析；
通过解析得到的所述控制报文的MAC地址判断是否为所要转发的控制报文；
若是，则将所抓取的控制报文添加所述代理CPU的标志位并切换至转发平面堆叠链路进行发送。
优选地，还包括：
当检测到所述控制平面堆叠链路通信恢复时，将所述控制报文切换至转发平面堆叠链路进行发送。
此外，为实现上述目的，本发明还提供一种虚拟交换集群中通信的装置，所述虚拟交换集群中通信的装置包括：
检测模块，设置为检测虚拟交换集群中的本端成员与对端成员之间的控制平面堆叠链路的通信状态；
切换模块，设置为当检测到所述控制平面堆叠链路通信中断时，将所述控制平面堆叠链路上的控制报文切换至转发平面堆叠链路进行发送。
优选地，所述检测模块包括：
发送单元，设置为由所述本端成员向对端成员发送检测报文；
判断单元，设置为判断所述本端成员是否在预定时间内接收到所述对端成员返回的所述检测报文对应的响应报文。
优选地，所述切换模块包括：
处理单元，设置为当检测到所述控制平面堆叠链路通信中断时，将所述本端成员与对端成员各自的生成树状态设为拥塞，并清除各自的物理接口学习到的MAC地址；
切换单元，设置为获取代理CPU，由所述代理CPU抓取所述控制报文，将所抓取的控制 报文切换至转发平面堆叠链路进行发送。
优选地，所述切换单元包括：
解析子单元，设置为获取代理CPU，由所述代理CPU抓取所述控制报文并进行解析；
判断子单元，设置为通过解析得到的所述控制报文的MAC地址判断是否为所要转发的控制报文；
切换子单元，设置为若是，则将所抓取的控制报文添加所述代理CPU的标志位并切换至转发平面堆叠链路进行发送
优选地，还包括恢复模块，设置为当检测到所述控制平面堆叠链路通信恢复时，将所述控制报文切换至转发平面堆叠链路进行发送。
本发明一种虚拟交换集群中通信的方法及装置，在检测到控制平面堆叠链路中断或失效时，快速将控制平面堆叠链路的控制报文切换到转发平面堆叠链路来发送，由转发平面堆叠链路来继续实现控制平面堆叠链路的控制报文的通信，保证控制平面堆叠链路通信的连续性，并可提高虚拟交换集群的稳定性和可靠性。
附图说明
图1为本发明虚拟交换集群中通信的方法第一实施例的流程示意图；
图2为图1中步骤S101的细化流程示意图；
图3为图1中步骤S102的细化流程示意图；
图4为本发明虚拟交换集群中通信的方法第二实施例的流程示意图；
图5为本发明虚拟交换集群中通信的装置第一实施例的功能模块示意图；
图6为图5中检测模块的细化功能模块示意图；
图7为图5中切换模块的细化功能模块示意图；
图8为本发明虚拟交换集群中通信的装置第二实施例的功能模块示意图。
本发明目的的实现、功能特点及优点将结合实施例，参照附图做进一步说明。
具体实施方式
应当理解，此处所描述的具体实施例仅仅用以解释本发明，并不用于限定本发明。
本发明提供一种虚拟交换集群中通信的方法，参照图1，在一实施例中，该方法包括：
步骤S101，检测虚拟交换集群中的本端成员与对端成员之间的控制平面堆叠链路的通信 状态；
本实施例中，在虚拟交换集群中，本端成员与对端成员之间的控制平面堆叠链路在通信正常时，通过控制平面堆叠链路发送两者间的控制报文，同时，本端成员通过控制平面堆叠链路向对端成员发送用来检测两者间的通信状态的检测报文，然后接收对端成员反馈，通过所反馈的报文来检测控制平面堆叠链路的通信状态是否正常。
本实施例中，如果本端成员或对端成员在预定的时间内未收到对方反馈的报文，则认为控制平面堆叠链路断开，本端成员与对端成员之间的控制平面堆叠链路通信中断。
步骤S102，当检测到所述控制平面堆叠链路通信中断时，将所述控制平面堆叠链路上的控制报文切换至转发平面堆叠链路进行发送。
本实施例中，当检测到控制平面堆叠链路通信中断时，通过借助软转发平面堆叠链路实现虚拟交换集群各个成员之间控制平面堆叠链路的通信，保证虚拟交换集群的成员间的控制报文能够有效转发，使虚拟交换集群成员间能够正常通信。
与现有技术相比，本实施例在检测到控制平面堆叠链路中断或失效时，快速将控制平面堆叠链路的控制报文切换到转发平面堆叠链路来发送，由转发平面堆叠链路来继续实现控制平面堆叠链路的控制报文的通信，保证控制平面堆叠链路通信的连续性，并可提高虚拟交换集群的稳定性和可靠性。
在一优选的实施例中，如图2所示，在上述图1的实施例的基础上，所述步骤S101包括：
步骤S1011，由所述本端成员向对端成员发送检测报文；
步骤S1012，判断所述本端成员是否在预定时间内接收到所述对端成员返回的所述检测报文对应的响应报文。
本实施例中，本端成员与对端成员之间相互发送检测报文，并接收对方反馈的报文，通过反馈的报文来检测控制平面堆叠链路的通信状态。
本实施例中，启用本端成员与对端成员的链路检测报文发送接收任务，本端成员与对端成员每秒从控制平面堆叠链路发送检测报文到对方，并监听对方反馈的反馈报文。
本实施例中，检测报文定义私有特征：检测报文的以太网协议号为0Xf002，目的MAC地址为0x0102-0304-0506；在预定时间内(例如3秒，也可根据实际修改为合理值)未收到对方反馈的反馈报文，就认为与对方间的控制平面堆叠链路通信中断。
在一优选的实施例中，如图3所示，在上述图1的实施例的基础上，所述步骤S102包括：
步骤S1021，当检测到所述控制平面堆叠链路通信中断时，将所述本端成员与对端成员各 自的生成树状态设为拥塞，并清除各自的物理接口学习到的MAC地址；
步骤S1022，获取代理CPU，由所述代理CPU抓取所述控制报文，将所抓取的控制报文切换至转发平面堆叠链路进行发送。
本实施例中，当检测到控制平面堆叠链路通信中断时，将本端成员与对端成员各自的生成树状态设为拥塞，以禁止转发数据，只能转发检测报文，防止控制平面堆叠链路恢复时出现环路问题；清除两成员及对端成员两者在控制平面堆叠链路间的交换机switch的物理接口学习到的MAC地址，让已知单播控制报文变成未知单播控制报文。
本实施例中，选择出代理CPU，在两成员各自启用代理CPU的以太网口抓包任务，抓取控制报文，以与对端成员有转发平面堆叠链路连接的CPU做为代理，发送控制报文到代理CPU以太网口的交换机的端口上。
可选地，由代理CPU抓取控制报文并进行解析，通过解析得到的控制报文的MAC地址判断是否为所要转发的控制报文；若是，则将所抓取的控制报文添加代理CPU的标志位并切换至转发平面堆叠链路进行发送。
本实施例中，本端成员与对端成员抓取到控制报文后，各自解析需要转发到对方的控制报文；具体的，在虚拟交换集群系统中，为各代理CPU分配的MAC地址是唯一的，不同的成员占用的MAC地址也在一个区域范围内，这样可以根据抓取的控制报文的MAC地址来确定是否需要发送到对端成员，需要发送到对端成员的报文有：MAC地址是广播或者组播的报文；MAC地址是单播，并且在对端成员所属的地址范围内。
本实施例中，如果是需要发送到对端成员的控制报文，则在控制报文中添加代理CPU标志位，以使对端成员收到控制报文后便于区分。本实施例中，本端成员与对端成员之间各自通过转发平面堆叠链路来发送抓取的控制平面堆叠链路的控制报文。通过转发平面堆叠链路的发送，将控制报文发送到对端。对端成员接收到控制报文后，根据控制报文中是否有代理CPU标志位来确定是不是控制平面堆叠链路发送的控制报文。
本实施例中，两成员各自把对方发送过来的控制平面堆叠链路的控制报文，发送到自身的控制平面通信系统。对于控制平面堆叠链路的控制报文，需要通过CPU网口的发包处理，发送到自身的控制平面通信系统。
本实施例通过以上方式，当控制平面堆叠链路通信发生中断时，可以实现转发平面堆叠链路来发送控制平面堆叠链路上的控制报文的发送，并最终发送至对应的控制平面通信系统，保证了控制平面堆叠链路通信的连续性。
在一优选的实施例中，如图4所示，在上述图1的实施例的基础上，该方法还包括：
步骤103：当检测到所述控制平面堆叠链路通信恢复时，将所述控制报文切换至转发平面堆叠链路进行发送。
本实施例中，当控制平面堆叠链路接收到所发送的检测报文的反馈报文时，即判断为检测到控制平面堆叠链路恢复连通时，关闭本端成员及对端成员各自启用的控制平面堆叠链路的控制报文的抓包任务，删除交换机switch连接到代理CPU网口的端口的MAC地址，设置连接两成员控制平面堆叠链路各自的物理接口生成树状态为转发(FORWARD)状态，恢复控制平面堆叠链路来转发控制报文，保持通信继续。
本发明还提供一种虚拟交换集群中通信的装置，如图5所示，在一实施例中，所述虚拟交换集群中通信的装置包括：
检测模块101，设置为检测虚拟交换集群中的本端成员与对端成员之间的控制平面堆叠链路的通信状态；
本实施例中，在虚拟交换集群中，本端成员与对端成员之间的控制平面堆叠链路在通信正常时，通过控制平面堆叠链路发送两者间的控制报文，同时，本端成员通过控制平面堆叠链路向对端成员发送用来检测两者间的通信状态的检测报文，然后接收对端成员反馈，通过所反馈的报文来检测控制平面堆叠链路的通信状态是否正常。
本实施例中，如果本端成员或对端成员在预定的时间内未收到对方反馈的报文，则认为控制平面堆叠链路断开，本端成员与对端成员之间的控制平面堆叠链路通信中断。
切换模块102，设置为当检测到所述控制平面堆叠链路通信中断时，将所述控制平面堆叠链路上的控制报文切换至转发平面堆叠链路进行发送。
本实施例中，当检测到控制平面堆叠链路通信中断时，通过借助软转发平面堆叠链路实现虚拟交换集群各个成员之间控制平面堆叠链路的通信，保证虚拟交换集群的成员间的控制报文能够有效转发，使虚拟交换集群成员间能够正常通信。
与现有技术相比，本实施例在检测到控制平面堆叠链路中断或失效时，快速将控制平面堆叠链路的控制报文切换到转发平面堆叠链路来发送，由转发平面堆叠链路来继续实现控制平面堆叠链路的控制报文的通信，保证控制平面堆叠链路通信的连续性，并可提高虚拟交换集群的稳定性和可靠性。
在一优选的实施例中，如图6所示，在上述图5的实施例的基础上，所述检测模块101包括：
发送单元1011，设置为由所述本端成员向对端成员发送检测报文；
判断单元1012，设置为判断所述本端成员是否在预定时间内接收到所述对端成员返回的所述检测报文对应的响应报文。
本实施例中，本端成员与对端成员之间相互发送检测报文，并接收对方反馈的报文，通 过反馈的报文来检测控制平面堆叠链路的通信状态。
本实施例中，启用本端成员与对端成员的链路检测报文发送接收任务，本端成员与对端成员每秒从控制平面堆叠链路发送检测报文到对方，并监听对方反馈的反馈报文。
本实施例中，检测报文定义私有特征：检测报文的以太网协议号为0Xf002，目的MAC地址为0x0102-0304-0506；在预定时间内(例如3秒，也可根据实际修改为合理值)未收到对方反馈的反馈报文，就认为与对方间的控制平面堆叠链路通信中断。
在一优选的实施例中，如图7所示，在上述图5的实施例的基础上，所述切换模块102包括：
处理单元1021，设置为当检测到所述控制平面堆叠链路通信中断时，将所述本端成员与对端成员各自的生成树状态设为拥塞，并清除各自的物理接口学习到的MAC地址；
切换单元1022，设置为获取代理CPU，由所述代理CPU抓取所述控制报文，将所抓取的控制报文切换至转发平面堆叠链路进行发送。
本实施例中，当检测到控制平面堆叠链路通信中断时，将本端成员与对端成员各自的生成树状态设为拥塞，以禁止转发数据，只能转发检测报文，防止控制平面堆叠链路恢复时出现环路问题；清除两成员及对端成员两者在控制平面堆叠链路间的交换机switch的物理接口学习到的MAC地址，让已知单播控制报文变成未知单播控制报文。
本实施例中，选择出代理CPU，在两成员各自启用代理CPU的以太网口抓包任务，抓取控制报文，以与对端成员有转发平面堆叠链路连接的CPU做为代理，发送控制报文到代理CPU以太网口的交换机的端口上。
可选地，由代理CPU抓取控制报文并进行解析，通过解析得到的控制报文的MAC地址判断是否为所要转发的控制报文；若是，则将所抓取的控制报文添加代理CPU的标志位并切换至转发平面堆叠链路进行发送。
本实施例中，本端成员与对端成员抓取到控制报文后，各自解析需要转发到对方的控制报文；具体的，在虚拟交换集群系统中，为各代理CPU分配的MAC地址是唯一的，不同的成员占用的MAC地址也在一个区域范围内，这样可以根据抓取的控制报文的MAC地址来确定是否需要发送到对端成员，需要发送到对端成员的报文有：MAC地址是广播或者组播的报文；MAC地址是单播，并且在对端成员所属的地址范围内。
本实施例中，如果是需要发送到对端成员的控制报文，则在控制报文中添加代理CPU标志位，以使对端成员收到控制报文后便于区分。本实施例中，本端成员与对端成员之间各自通过转发平面堆叠链路来发送抓取的控制平面堆叠链路的控制报文。通过转发平面堆叠链路的发送，将控制报文发送到对端。对端成员接收到控制报文后，根据控制报文中是否有代理CPU标志位来确定是不是控制平面堆叠链路发送的控制报文。
本实施例中，两成员各自把对方发送过来的控制平面堆叠链路的控制报文，发送到自身的控制平面通信系统。对于控制平面堆叠链路的控制报文，需要通过CPU网口的发包处理，发送到自身的控制平面通信系统。
本实施例通过以上方式，当控制平面堆叠链路通信发生中断时，可以实现转发平面堆叠链路来发送控制平面堆叠链路上的控制报文的发送，并最终发送至对应的控制平面通信系统，保证了控制平面堆叠链路通信的连续性。
在一优选的实施例中，如图8所示，在上述图5的实施例的基础上，还包括恢复模块103，设置为当检测到所述控制平面堆叠链路通信恢复时，将所述控制报文切换至转发平面堆叠链路进行发送。
本实施例中，当控制平面堆叠链路接收到所发送的检测报文的反馈报文时，即判断为检测到控制平面堆叠链路恢复连通时，关闭本端成员及对端成员各自启用的控制平面堆叠链路的控制报文的抓包任务，删除交换机switch连接到代理CPU网口的端口的MAC地址，设置连接两成员控制平面堆叠链路各自的物理接口生成树状态为转发(FORWARD)状态，恢复控制平面堆叠链路来转发控制报文，保持通信继续。
以上仅为本发明的优选实施例，并非因此限制本发明的专利范围，凡是利用本发明说明书及附图内容所作的等效结构或等效流程变换，或直接或间接运用在其他相关的技术领域，均同理包括在本发明的专利保护范围内。
工业实用性
如上所述，本发明实施例提供的一种虚拟交换集群中通信的方法及装置具有以下有益效果：在检测到控制平面堆叠链路中断或失效时，快速将控制平面堆叠链路的控制报文切换到转发平面堆叠链路来发送，由转发平面堆叠链路来继续实现控制平面堆叠链路的控制报文的通信，保证控制平面堆叠链路通信的连续性，并可提高虚拟交换集群的稳定性和可靠性。

Claims (10)

  1. 一种虚拟交换集群中通信的方法，所述虚拟交换集群中通信的方法包括以下步骤：
    检测虚拟交换集群中的本端成员与对端成员之间的控制平面堆叠链路的通信状态；
    当检测到所述控制平面堆叠链路通信中断时，将所述控制平面堆叠链路上的控制报文切换至转发平面堆叠链路进行发送。
  2. 如权利要求1所述的虚拟交换集群中通信的方法，其中，所述检测虚拟交换集群中的本端成员与对端成员之间的控制平面堆叠链路的通信状态的步骤包括：
    由所述本端成员向对端成员发送检测报文；
    判断所述本端成员是否在预定时间内接收到所述对端成员返回的所述检测报文对应的响应报文。
  3. 如权利要求1所述的虚拟交换集群中通信的方法，其中，当检测到所述控制平面堆叠链路通信中断时，将所述控制平面堆叠链路上的控制报文切换至转发平面堆叠链路进行发送的步骤包括：
    当检测到所述控制平面堆叠链路通信中断时，将所述本端成员与对端成员各自的生成树状态设为拥塞，并清除各自的物理接口学习到的MAC地址；
    获取代理CPU，由所述代理CPU抓取所述控制报文，将所抓取的控制报文切换至转发平面堆叠链路进行发送。
  4. 如权利要求3所述的虚拟交换集群中通信的方法，其中，所述获取代理CPU，由所述代理CPU抓取所述控制报文，将所抓取的控制报文切换至转发平面堆叠链路进行发送的步骤包括：
    获取代理CPU，由所述代理CPU抓取所述控制报文并进行解析；
    通过解析得到的所述控制报文的MAC地址判断是否为所要转发的控制报文；
    若是，则将所抓取的控制报文添加所述代理CPU的标志位并切换至转发平面堆叠链路进行发送。
  5. 如权利要求1至4任一项所述的虚拟交换集群中通信的方法，其中，还包括：
    当检测到所述控制平面堆叠链路通信恢复时，将所述控制报文切换至转发平面堆叠链路进行发送。
  6. 一种虚拟交换集群中通信的装置，所述虚拟交换集群中通信的装置包括：
    检测模块，设置为检测虚拟交换集群中的本端成员与对端成员之间的控制平面堆叠链路的通信状态；
    切换模块，设置为当检测到所述控制平面堆叠链路通信中断时，将所述控制平面堆 叠链路上的控制报文切换至转发平面堆叠链路进行发送。
  7. 如权利要求6所述的虚拟交换集群中通信的装置，其中，所述检测模块包括：
    发送单元，设置为由所述本端成员向对端成员发送检测报文；
    判断单元，设置为判断所述本端成员是否在预定时间内接收到所述对端成员返回的所述检测报文对应的响应报文。
  8. 如权利要求6所述的虚拟交换集群中通信的装置，其中，所述切换模块包括：
    处理单元，设置为当检测到所述控制平面堆叠链路通信中断时，将所述本端成员与对端成员各自的生成树状态设为拥塞，并清除各自的物理接口学习到的MAC地址；
    切换单元，设置为获取代理CPU，由所述代理CPU抓取所述控制报文，将所抓取的控制报文切换至转发平面堆叠链路进行发送。
  9. 如权利要求8所述的虚拟交换集群中通信的装置，其中，所述切换单元包括：
    解析子单元，设置为获取代理CPU，由所述代理CPU抓取所述控制报文并进行解析；
    判断子单元，设置为通过解析得到的所述控制报文的MAC地址判断是否为所要转发的控制报文；
    切换子单元，设置为若是，则将所抓取的控制报文添加所述代理CPU的标志位并切换至转发平面堆叠链路进行发送
  10. 如权利要求6-9任一项所述的虚拟交换集群中通信的装置，其中，还包括恢复模块，设置为当检测到所述控制平面堆叠链路通信恢复时，将所述控制报文切换至转发平面堆叠链路进行发送。
PCT/CN2015/097034 2014-12-16 2015-12-10 虚拟交换集群中通信的方法及装置 WO2016095750A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201410782626.2A CN105763412A (zh) 2014-12-16 2014-12-16 虚拟交换集群中通信的方法及装置
CN201410782626.2 2014-12-16

Publications (1)

Publication Number Publication Date
WO2016095750A1 true WO2016095750A1 (zh) 2016-06-23

Family

ID=56125913

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2015/097034 WO2016095750A1 (zh) 2014-12-16 2015-12-10 虚拟交换集群中通信的方法及装置

Country Status (2)

Country Link
CN (1) CN105763412A (zh)
WO (1) WO2016095750A1 (zh)

Families Citing this family (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108234321B (zh) * 2016-12-14 2021-02-23 中兴通讯股份有限公司 一种报文转发的方法及系统
CN110601972A (zh) * 2019-09-18 2019-12-20 新华三信息安全技术有限公司 一种报文传输方法、装置及智能弹性架构系统

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101277195A (zh) * 2007-03-30 2008-10-01 杭州华三通信技术有限公司 一种交换网通信系统、实现方法及交换装置
CN102136936A (zh) * 2010-11-24 2011-07-27 华为技术有限公司 避免控制平面故障影响转发平面运行的方法、节点及系统
US8379676B1 (en) * 2006-06-01 2013-02-19 World Wide Packets, Inc. Injecting in-band control messages without impacting a data rate
CN103036724A (zh) * 2012-12-14 2013-04-10 北京华为数字技术有限公司 状态信息传输方法、网络设备及组合设备

Family Cites Families (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN100450039C (zh) * 2007-01-12 2009-01-07 华为技术有限公司 快速收敛端到端业务的方法和装置
CN102130817B (zh) * 2010-01-20 2013-09-11 杭州华三通信技术有限公司 堆叠系统中保持流量不中断的方法和机框交换机

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US8379676B1 (en) * 2006-06-01 2013-02-19 World Wide Packets, Inc. Injecting in-band control messages without impacting a data rate
CN101277195A (zh) * 2007-03-30 2008-10-01 杭州华三通信技术有限公司 一种交换网通信系统、实现方法及交换装置
CN102136936A (zh) * 2010-11-24 2011-07-27 华为技术有限公司 避免控制平面故障影响转发平面运行的方法、节点及系统
CN103036724A (zh) * 2012-12-14 2013-04-10 北京华为数字技术有限公司 状态信息传输方法、网络设备及组合设备

Also Published As

Publication number Publication date
CN105763412A (zh) 2016-07-13

Similar Documents

Publication Publication Date Title
EP3041179B1 (en) A method and apparatus for use in network management
US9350680B2 (en) Protection switching over a virtual link aggregation
JP4840236B2 (ja) ネットワークシステム及びノード装置
US9876706B2 (en) Relay system and switching device for a layer 3 network
CN110324165B (zh) 网络设备的管理方法、装置及系统
US9100203B2 (en) IP multicast over multi-chassis trunk
EP2555476A1 (en) Method, system and device for protecting multicast in communication network
WO2016165463A1 (zh) 一种基于软件定义网络的链路故障上报方法和转发设备
US9660898B2 (en) Enhanced protocol independent multicast source registration over a reliable transport
US9363094B2 (en) Relay system and switching device
WO2015096409A1 (zh) 软件定义网络中的链路发现方法、装置及系统
WO2012106915A1 (zh) 故障通告方法、检测装置、转发装置、系统及数据结构
CN112822103B (zh) 一种信息上报方法和信息处理方法及设备
JPWO2014133015A1 (ja) 制御装置、通信システム、スイッチ制御方法及びプログラム
WO2016172926A1 (zh) 通信系统中的通信方法和设备及系统
WO2016037443A1 (zh) 一种临时通道的建立方法、设备、系统和计算机存储介质
US20140269746A1 (en) Load balancing of logical connections over multi-chassis trunk
WO2016095750A1 (zh) 虚拟交换集群中通信的方法及装置
WO2014008651A1 (zh) Bfd会话建立的方法及路由设备
US9985926B2 (en) Address acquiring method and network virtualization edge device
US9083542B2 (en) Method for using a computer network
WO2016086693A1 (zh) 报文传输的方法、骨干交换机和接入交换机
JP5954793B2 (ja) 通信接続装置、通信制御装置およびそのプログラム並びに通信制御方法
US10205652B2 (en) Path checking method, sink node device, and communications system
CN106992878B (zh) 一种组播检测的方法及装置

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 15869258

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 15869258

Country of ref document: EP

Kind code of ref document: A1