WO2019201209A1 - 报文转发 - Google Patents

报文转发 Download PDF

Info

Publication number
WO2019201209A1
WO2019201209A1 PCT/CN2019/082716 CN2019082716W WO2019201209A1 WO 2019201209 A1 WO2019201209 A1 WO 2019201209A1 CN 2019082716 W CN2019082716 W CN 2019082716W WO 2019201209 A1 WO2019201209 A1 WO 2019201209A1
Authority
WO
WIPO (PCT)
Prior art keywords
address
leaf device
host
destination
route
Prior art date
Legal status (The legal status 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 status listed.)
Ceased
Application number
PCT/CN2019/082716
Other languages
English (en)
French (fr)
Inventor
温卫真
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
New H3C Technologies Co Ltd
Original Assignee
New H3C Technologies Co Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by New H3C Technologies Co Ltd filed Critical New H3C Technologies Co Ltd
Publication of WO2019201209A1 publication Critical patent/WO2019201209A1/zh
Anticipated expiration legal-status Critical
Ceased legal-status Critical Current

Links

Images

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L45/00—Routing or path finding of packets in data switching networks
    • H04L45/24—Multipath
    • H04L45/245—Link aggregation, e.g. trunking
    • 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
    • H04L12/4633—Interconnection of networks using encapsulation techniques, e.g. tunneling
    • 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
    • H04L12/4641—Virtual LANs, VLANs, e.g. virtual private networks [VPN]
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L45/00—Routing or path finding of packets in data switching networks
    • H04L45/18—Loop-free operations

Definitions

  • MLAG Multi-Chassis Link Aggregation Group
  • This mechanism improves the link reliability from the board level to the device level by combining two devices to implement cross-device link aggregation to form a MLAG active-active system.
  • a MLAG deployment scheme based on the Spine-Leaf (backbone-leaf) network is provided.
  • Host A is dual-homed to Leaf device 1 and Leaf device 2; host B is connected to Leaf device 3; Leaf device 1 and leaf device are provided. 2 and the leaf device 3 can communicate with the Spine device through the VXLAN tunnel.
  • the destination VTEP VXLAN Tunnel End Point
  • IP Internet Protocol
  • the Spine device receives the packet from the host B and distributes the packet to the Leaf device 1 or the Leaf device 2 based on the load sharing mechanism. If the link between the leaf device 1 and the host A is faulty, the leaf device 1 forwards the packet to the leaf device 2, and the leaf device 2 forwards the packet to the host A.
  • FIG. 1 is a schematic diagram of a MLAG deployment scheme in a Spine-Leaf networking according to an embodiment of the present disclosure.
  • 2A is a schematic diagram of a packet forwarding path before a link failure.
  • 2B is a schematic diagram of a message forwarding path after a link failure.
  • FIG. 3 is a flowchart of a packet forwarding method according to an embodiment of the present disclosure.
  • FIG. 4 is a flow chart showing interaction between devices when a link is faulty according to an embodiment of the present disclosure.
  • FIG. 5 is a schematic diagram of a packet forwarding path after a link failure according to an embodiment of the present disclosure.
  • FIG. 6 is a flow chart of interaction between devices when the link shown in the embodiment of the present disclosure returns to normal.
  • FIG. 7 is a schematic diagram of a hardware structure of a Leaf device according to an embodiment of the present disclosure.
  • FIG. 8 is a schematic structural diagram of a Leaf device according to an embodiment of the present disclosure.
  • first, second, third, etc. may be used to describe various information in the embodiments of the present disclosure, such information should not be limited to these terms. These terms are only used to distinguish the same type of information from each other.
  • first information may also be referred to as the second information without departing from the scope of the embodiments of the present disclosure.
  • second information may also be referred to as the first information.
  • word "if” as used herein may be interpreted as "when” or "when” or "in response to a determination.”
  • FIG. 1 is a schematic diagram of a MLAG deployment scheme in a Spine-Leaf networking according to an embodiment of the present disclosure.
  • the network includes one Spine device 100, three Leaf devices (Leaf device 101, Leaf device 102, Leaf device 103) and two hosts (Host 111 and Host 112).
  • the host 111 is dual-homed to the Leaf device 101 and the Leaf device 102 based on the MLAG mechanism.
  • the Leaf device 101 and the Leaf device 102 are called MLAG members, and have the same virtual IP address (MLAG group address), which is recorded as IP12; and the Leaf device 103 It does not belong to the MLAG to which the Leaf device 101 and the Leaf device 102 belong.
  • Each Leaf device is connected through a VXLAN tunnel.
  • the leaf device 103 When the forwarding device 112 accesses the packet of the host 111, the leaf device 103 performs VXLAN encapsulation on the packet, and the destination VTEP IP address of the encapsulated packet is the virtual IP address IP12.
  • the Spine device 100 receives the packet and distributes the packet to the Leaf device 101 or the Leaf device 102 based on the load sharing mechanism. If the message is distributed to the Leaf device 101, the message is forwarded by the Leaf device 101 to the host 111.
  • 2A is a schematic diagram of a path for forwarding a message by the Leaf device 101, where a dotted line indicates a forwarding path.
  • FIG. 2B it is a schematic diagram of a new packet forwarding path after a link failure. As shown in FIG. 2B, there is a round-trip of the packet forwarding path at the leaf device 101, which results in a low forwarding efficiency of the packet.
  • FIG. 3 is a flowchart of a packet forwarding method according to an embodiment of the present disclosure. The method describes a packet forwarding process from a first Leaf device side in the MLAG.
  • Step 301 The first Leaf device receives the first route revocation message sent by the second Leaf device.
  • the second Leaf device and the first Leaf device belong to the same MLAG and have the same virtual IP address.
  • Step 302 If the destination IP address included in the first route revocation message is the IP address of the local host, and the next hop information is the virtual IP address, the first leaf device sends a first route advertisement message.
  • the next hop information is a virtual IP address shared by the first leaf device and the second leaf device (referred to as a virtual IP address in the following description).
  • the link between the second Leaf device and the local host is faulty and cannot communicate properly.
  • the local host refers to a host that is dual-homed to the first Leaf device and the second Leaf device based on the MLAG mechanism.
  • the first leaf device sends a first route advertisement message, where the destination IP address included in the first route advertisement message is the IP address of the local host, and the next hop information is the real IP address of the first Leaf device (the IP address of the physical device) .
  • the Leaf device 101 and the Leaf device 102 in FIG. 1 have the same virtual IP address IP12 and have their own real IP addresses, which are respectively recorded as IP11 and IP22.
  • Both the route revocation message and the route advertisement message described in the examples of the present disclosure may be in the format of, for example, a BGP message.
  • Any device in the Spine-Leaf network can carry the updated route in a BGP message and send it to another device that has established a BGP neighbor relationship with the device.
  • the Spine-Leaf network further includes a third Leaf device (for example, the Leaf device 103 in FIG. 1) that does not belong to the first Leaf device and the MLAG to which the second Leaf device belongs.
  • the third leaf device When receiving the first route revocation message sent by the second leaf device, the third leaf device deletes the destination IP address and the next hop information included in the first route revocation message, and deletes the destination IP address of the local record as the IP address of the local host.
  • the next hop information is the routing entry of the virtual IP address.
  • the third leaf device When receiving the first route advertisement message sent by the first leaf device, the third leaf device generates the destination IP address as the IP address of the local host according to the destination IP address and the next hop information included in the first route advertisement message, and the next The hop information is the routing entry of the real IP address of the first Leaf device (recorded as the first routing entry). Therefore, the third Leaf device forwards the packet addressed to the local host based on the first routing entry. Specifically, before the third leaf device forwards the packet, the packet is VXLAN encapsulated, and the destination VTEP IP address of the encapsulated packet is the real IP address of the first Leaf device. The message is then forwarded to the first Leaf device.
  • the packet forwarding path is bypassed at the second Leaf device when the link between the second leaf device and the local host is faulty.
  • the packet is forwarded from the second Leaf device to the first Leaf device and then forwarded to the local device. Host. It can be seen that the packet forwarding method provided by the embodiment of the present disclosure can improve the packet forwarding efficiency.
  • the first leaf device when the first leaf device receives the second route advertisement message sent by the second leaf device, if the destination IP address included in the second route advertisement message is the IP address of the local host, the next hop information is The virtual IP address indicates that the link between the second Leaf device and the local host is normal and can communicate. At this time, the first leaf device sends a second route revocation message, where the destination IP address included in the second route revocation message is the IP address of the local host, and the next hop information is the real IP address of the first Leaf device.
  • the third leaf device when the third leaf device receives the second route revocation message sent by the first leaf device, deleting the first routing table of the local record according to the destination IP address and the next hop information included in the second route revocation message.
  • the entry is the IP address of the local host and the next hop information is the routing entry of the real IP address of the first leaf device.
  • the destination IP address is the IP address of the local host, and the destination IP address is generated according to the destination IP address and the next hop information included in the second route advertisement message.
  • the one-hop information is the routing entry of the virtual IP address (recorded as the second routing entry). The third Leaf device forwards the packet addressed to the local host based on the second routing entry.
  • the third leaf device performs VXLAN encapsulation on the packet before forwarding the packet, where the destination VTEP IP address of the encapsulated packet is a virtual IP address. Then, the Spine device redistributes the packet to the second Leaf device based on the load balancing mechanism, and the second Leaf device forwards the packet to the local host. That is, after the link between the second leaf device and the local host returns to normal, the packet forwarding path is restored.
  • the packet is sent to the local host based on the MLAG dual-homing access to the first Leaf device and the second Leaf device.
  • the first leaf device obtains the destination IP address carried in the packet, and the destination IP address is the IP address of the local host.
  • the media access control (address) table determines the port connected to the local host (also referred to as a distributed aggregation port); and determines the communication state of the first leaf device and the local host based on the state of the distributed aggregation port.
  • the communication state between the first leaf device and the local host is normal; if the state of the distributed aggregation interface is DOWN (the chain between the first leaf device and the local host) If the path is faulty, it is determined that the communication state of the first Leaf device and the local host is abnormal.
  • the first leaf device sends a third route revocation message, where the destination IP address included in the third route revocation message is the IP address of the local host, and the next hop information is Virtual IP address.
  • the third route device receives the third route revocation message, the second routing entry of the local record is deleted according to the destination IP address and the next hop information included in the third route revocation message, that is, the destination IP address is the local host.
  • the IP address and the next hop information are the routing entries of the virtual IP address.
  • the communication state of the first leaf device and the local host may be abnormal.
  • the second leaf device sends a third route advertisement message, where the destination IP address included in the third route advertisement message is the IP address of the local host, and the next hop information is the real IP address of the second leaf device.
  • the third leaf device When receiving the third route advertisement message, the third leaf device generates the destination IP address as the local host's IP address and the next hop information as the second, according to the destination IP address and the next hop information included in the third route advertisement message.
  • the routing entry of the real IP address of the leaf device (recorded as the third routing table entry). Therefore, the third Leaf device forwards the packet addressed to the local host based on the third routing entry.
  • the third leaf device performs VXLAN encapsulation on the packet before forwarding the packet, where the destination VTEP IP address of the encapsulated packet is the real IP address of the second Leaf device.
  • the message is then forwarded to the second Leaf device. Therefore, the packet forwarding path is bypassed at the first leaf device when the communication between the first leaf device and the local host is abnormal. For example, the packet is forwarded from the first leaf device to the second leaf device and then forwarded to the local host. It can be seen that the packet forwarding method provided by the embodiment of the present disclosure can improve packet forwarding efficiency.
  • the first leaf device sends a fourth route advertisement message, where the destination IP address included in the fourth route advertisement message is the IP address of the local host, and the next hop information is virtual. IP address.
  • the second leaf device may be configured to detect that the communication state of the first leaf device and the local host is normal, and send a fourth route revocation message, where the destination IP address included in the fourth route revocation message is IP address of the local host.
  • the next hop information is the real IP address of the second Leaf device.
  • the third leaf device When receiving the fourth route revocation message, the third leaf device deletes the third routing entry of the local record according to the destination IP address and the next hop information included in the fourth route revocation message, that is, the destination IP address is the local host.
  • the IP address and the next hop information are the routing entries of the real IP address of the second Leaf device.
  • the third leaf device receives the fourth route advertisement message, the second routing entry is generated according to the destination IP address and the next hop information included in the fourth route advertisement message, that is, the destination IP address is the IP address of the local host.
  • the next hop information is the routing entry of the virtual IP address.
  • the third Leaf device forwards the packet addressed to the local host based on the second routing entry.
  • the third leaf device performs VXLAN encapsulation on the packet before forwarding the packet, where the destination VTEP IP address of the encapsulated packet is a virtual IP address.
  • the Spine device redistributes the packet to the first leaf device based on the load balancing mechanism, and the first leaf device forwards the packet to the local host. In other words, after the communication between the first Leaf device and the local host returns to normal, the forwarding path of the packet is restored accordingly.
  • the packet forwarding method provided by the embodiment of the present disclosure only performs routing update for the host that is currently interacting with data, and the routes of other hosts that do not exchange data under the faulty distributed aggregation interface are not updated, thereby greatly reducing the number of routing updates.
  • the route of the host that is interacting with the data is also updated. Therefore, the route flapping caused by the large-scale routing update can be avoided.
  • the Spine-Leaf networking shown in Figure 1 is used as an example to describe the packet forwarding process in detail.
  • the host 112 accesses the host 111.
  • FIG. 4 it is a flow chart of interaction between devices when the link is faulty.
  • the leaf device 103 forwards the packet sent by the host 112 to the host 111 based on the routing entry R1 (the destination IP address is the IP address of the host 111 and the next hop information is the virtual IP address), and performs VXLAN encapsulation and encapsulation of the packet.
  • the destination VTEP IP address of the packet is IP12 (virtual IP address).
  • step 402 the leaf device 103 sends the encapsulated packet to the Spine device 100 through the outbound interface connected to the Spine device 100.
  • Step 403 After receiving the packet whose destination VTEP IP is IP12, the Spine device 100 distributes the packet to the leaf device 101 based on the load balancing mechanism.
  • Step 404 After receiving the packet whose destination VTEP IP is IP12, the leaf device 101 obtains the IP address of the host 111 carried by the packet, and determines the distributed aggregation interface connected to the host 111. If the distributed aggregation interface is normal (in the UP state) The status of the packet is sent to the host 111 through the distributed aggregation interface. The forwarding path is as shown by the dotted line in FIG. 2A. If the distributed aggregation interface is faulty (down state), go to step 405.
  • step 405 the leaf device 101 sends a route revocation message to the Spine device 100 (the route reflector).
  • the destination IP address included in the route revocation message is the IP address of the host 111, and the next hop information is a virtual IP address.
  • the Spine device 100 reflects the route revocation message to the Leaf device 102 and the Leaf device 103, respectively.
  • Step 407 After receiving the route revocation message, the leaf device 103 deletes the locally recorded routing entry R1 according to the destination IP address and the next hop information included in the route revocation message.
  • Step 408 After receiving the route revocation message, the leaf device 102 sends a route advertisement message to the Spine device 100.
  • the destination IP address included in the route advertisement message is the IP address of the host, and the next hop information is the real IP address of the Leaf device 102. (IP22).
  • step 409 the Spine device 100 reflects the route advertisement message to the Leaf device 103.
  • Step 410 After receiving the route advertisement message, the leaf device 103 generates a routing entry R2 according to the destination IP address and the next hop information included in the route advertisement message.
  • the destination IP address is the IP address of the host 111, and the next hop information. Is the real IP address of the Leaf device 102).
  • the leaf device 103 forwards the packet sent by the host 112 to the host 111 based on the routing entry R2, and performs VXLAN encapsulation on the packet.
  • the destination VTEP IP address of the encapsulated packet is IP22 (the real IP address of the Leaf device 102).
  • step 412 the leaf device 103 sends the encapsulated packet to the Spine device 100 through the outbound interface connected to the Spine device 100.
  • Step 413 After receiving the packet whose destination VTEP IP is IP22, the Spine device 100 forwards the packet directly to the Leaf device 102.
  • step 414 the leaf device 102 sends the packet to the host 111 through the distributed aggregation interface connected to the host 111 on the device.
  • the forwarding path is as shown by the dotted line in FIG.
  • FIG. 6 a flow chart of interaction between devices when the link is restored to normal.
  • Step 601 If the distributed aggregation interface of the leaf device 101 and the host 111 is restored to the UP state, the leaf device 101 sends a route advertisement message to the Spine device 100, where the destination IP address included in the route advertisement message is the IP address of the host 111.
  • the next hop information is a virtual IP address (IP12).
  • the Spine device 100 reflects the route advertisement message to the Leaf device 102 and the Leaf device 103.
  • Step 603 After receiving the route advertisement message, the leaf device 103 generates a routing entry R1 according to the destination IP address and the next hop information included in the route advertisement message.
  • the destination IP address is the IP address of the host 111, and the next hop information. Is a virtual IP address).
  • Step 604 After receiving the route advertisement message, the leaf device 102 sends a route revocation message to the Spine device 100.
  • the destination IP address included in the route revocation message is the IP address of the host 111, and the next hop information is the real IP address of the Leaf device 102. Address (IP22).
  • step 605 the Spine device 100 reflects the route revocation message to the Leaf device 103.
  • Step 606 after receiving the route revocation message, the leaf device 103 deletes the locally recorded routing entry R2 according to the destination IP address and the next hop information included in the route revocation message (the destination IP address is the IP address of the host 111, One hop information is the real IP address of the Leaf device 102).
  • the leaf device 103 forwards the packet sent by the host 112 to the host 111 based on the routing entry R1, and performs VXLAN encapsulation on the packet.
  • the destination VTEP IP address of the encapsulated packet is IP12 (virtual IP address).
  • step 608 the leaf device 103 sends the encapsulated packet to the Spine device 100 through the outbound interface connected to the Spine device 100.
  • Step 609 After receiving the packet with the destination VTEP IP being IP12, the Spine device 100 re-distributes the packet sent by the host 112 to the host 111 to the Leaf device 101 based on the load sharing mechanism.
  • step 610 the leaf device 101 sends the packet to the host 111 through the distributed aggregation port connected to the host 111 on the device, and the forwarding path is restored to the dotted line in FIG. 2A.
  • FIG. 7 is a schematic structural diagram of hardware of a Leaf device according to an embodiment of the present disclosure.
  • the Leaf device can include a processor 701, a machine readable storage medium 702 that stores machine executable instructions.
  • Processor 701 and machine readable storage medium 702 can communicate via system bus 703. And, by reading and executing machine executable instructions in the machine readable storage medium 702 corresponding to the message forwarding logic, the processor 701 can perform the message forwarding method described above.
  • the machine-readable storage medium 702 referred to herein can be any electronic, magnetic, optical, or other physical storage device that can contain or store information such as executable instructions, data, and the like.
  • the machine-readable storage medium may be: RAM (Radom Access Memory), volatile memory, non-volatile memory, flash memory, storage drive (such as a hard drive), solid state drive, any type of storage.
  • RAM Random Access Memory
  • volatile memory volatile memory
  • non-volatile memory such as a hard drive
  • solid state drive any type of storage.
  • a disk such as a compact disc, dvd, etc.
  • a similar storage medium or a combination thereof.
  • the receiving unit 801 is configured to receive a first route revocation message sent by the second Leaf device.
  • the sending unit 802 is configured to: if the destination IP address included in the first route revocation message is an IP address of the local host, and the next hop information is the virtual IP address, send a first route advertisement message, where the first The destination IP address included in the route advertisement message is the IP address of the local host, and the next hop information is the real IP address of the leaf device, so that the third leaf device that does not belong to the MLAG generates the first routing entry.
  • the destination IP address of the first routing entry is the IP address of the local host, and the next hop information is the real IP address of the Leaf device.
  • the receiving unit 801 is further configured to receive a second route advertisement message sent by the second leaf device, where the sending unit 802 is further configured to: if the destination IP address included in the second route advertisement message is included The address is the IP address of the local host, and the next hop information is the virtual IP address, the second route revocation message is sent, and the destination IP address included in the second route revocation message is the IP address of the local host.
  • the next hop information is the real IP address of the leaf device, so that the third leaf device deletes the first routing entry.
  • the device further includes: a determining unit, configured to determine a communication state of the Leaf device and the local host.
  • the sending unit 802 is further configured to: if the communication status of the Leaf device and the local host is abnormal, send a third route revocation message, and the destination IP address included in the third route revocation message
  • the address is the IP address of the local host, and the next hop information is the virtual IP address, so that the third leaf device deletes the second routing entry, and the destination IP address of the second routing entry is the local
  • the IP address of the host, the next hop information is the virtual IP address, and the second leaf device sends a third route advertisement message, where the destination IP address included in the third route advertisement message is the local host
  • the IP address, the next hop information is the real IP address of the second Leaf device.
  • the determining unit is configured to: obtain a destination IP address carried in the packet, where the destination IP address is the local IP address, if the packet with the destination VTEP IP address being the virtual IP address is received The IP address of the host is determined, and the distributed aggregation port connected to the local host is determined based on the destination IP address; and the communication state of the Leaf device and the local host is determined based on the state of the distributed aggregation interface.
  • the sending unit 802 is further configured to: if the communication status of the leaf device and the local host is normal, send a fourth route advertisement message, where the destination IP address included in the fourth route advertisement message is The IP address of the local host, the next hop information is the virtual IP address, so that the third Leaf device generates the second routing entry, and causes the second Leaf device to send a fourth route revocation message.
  • the destination IP address included in the fourth route revocation message is an IP address of the local host, and the next hop information is a real IP address of the second leaf device.
  • Embodiments of the present disclosure also provide a machine readable storage medium including machine executable instructions, such as machine readable storage medium 702 of FIG. 7, which may be processor 701 in a main control board of a network device Executed to implement the packet forwarding method described above.
  • machine executable instructions such as machine readable storage medium 702 of FIG. 7, which may be processor 701 in a main control board of a network device Executed to implement the packet forwarding method described above.

Landscapes

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

Abstract

第一Leaf设备与第二Leaf设备属于同一MLAG，具有相同的虚IP地址。第一Leaf设备接收第二Leaf设备发送的第一路由撤销消息；若第一路由撤销消息中包括的目的IP地址为双归接入第一Leaf设备和第二Leaf设备的主机的IP地址，下一跳信息为虚IP地址，发送第一路由发布消息，第一路由发布消息中包括的目的IP地址为主机的IP地址，下一跳信息为第一Leaf设备的实IP地址，以使不属于MLAG的第三Leaf设备生成第一路由表项，第一路由表项的目的IP地址为主机的IP地址，下一跳信息为第一Leaf设备的实IP地址。

Description

报文转发 背景技术
MLAG(Multi-chassis Link Aggregation Group，跨设备链路集合组)是一种实现跨设备链路聚合的机制。该机制通过将两台设备进行跨设备链路聚合，共同组成一个MLAG双活系统，从而把链路可靠性从单板级提高到了设备级。
提供基于Spine-Leaf(骨干-叶子)组网下的一种MLAG部署方案，其中，主机A双归接入Leaf设备1和Leaf设备2；主机B接入Leaf设备3；Leaf设备1和leaf设备2以及leaf设备3可通过VXLAN隧道与Spine设备通信。当主机B访问主机A时，从主机B发送的报文的目的VTEP(VXLAN Tunnel End Point，VXLAN隧道终端节点)IP(Internet Protocol，网际协议)地址为Leaf设备1和Leaf设备2共用的虚IP地址。Spine设备接收到主机B的报文，基于负载分担机制将该报文分发至Leaf设备1或Leaf设备2。若Leaf设备1与主机A之间的链路故障，则Leaf设备1将报文转发至Leaf设备2，由Leaf设备2再转发给主机A。
附图说明
为了更清楚地说明本公开实施例中的技术方案，下面将对实施例描述中所需要使用的附图作简单地介绍。
图1是本公开实施例示出的一种Spine-Leaf组网下的MLAG部署方案示意图。
图2A是链路故障之前的报文转发路径的示意图。
图2B是链路故障之后的报文转发路径的示意图。
图3是本公开实施例示出的一种报文转发方法的流程图。
图4是本公开实施例示出的链路故障时各设备之间的交互流程图。
图5是本公开实施例示出的链路故障后的报文转发路径的示意图。
图6是本公开实施例示出的链路恢复正常时各设备之间的交互流程图。
图7是本公开实施例示出的一种Leaf设备的硬件结构示意图。
图8是本公开实施例示出的一种Leaf设备的结构示意图。
具体实施方式
这里将详细地对示例性实施例进行说明，其示例表示在附图中。下面的描述涉及附图时，除非另有表示，不同附图中的相同数字表示相同或相似的要素。以下示例性实施例中所描述的实施方式并不代表与本公开实施例相一致的所有实施方式。相反，它们仅是与如所附权利要求书中所详述的、本公开实施例的一些方面相一致的装置和方法的例子。
在本公开实施例使用的术语是仅仅出于描述特定实施例的目的，而非旨在限制本公开实施例。在本公开实施例和所附权利要求书中所使用的单数形式的“一种”、“所述”和“该”也旨在包括多数形式，除非上下文清楚地表示其他含义。还应当理解，本文中使用的术语“和/或”是指并包含一个或多个相关联的列出项目的任何或所有可能组合。
应当理解，尽管在本公开实施例可能采用术语第一、第二、第三等来描述各种信息，但这些信息不应限于这些术语。这些术语仅用来将同一类型的信息彼此区分开。例如，在不脱离本公开实施例范围的情况下，第一信息也可以被称为第二信息，类似地，第二信息也可以被称为第一信息。取决于语境，如在此所使用的词语“如果”可以被解释成为“在……时”或“当……时”或“响应于确定”。
如图1所示，为本公开实施例示出的一种Spine-Leaf组网下的MLAG部署方案示意图。该组网包括一个Spine设备100、三个Leaf设备(Leaf设备101、Leaf设备102、Leaf设备103)、两台主机(主机111和主机112)。其中，主机111基于MLAG机制双归接入Leaf设备101和Leaf设备102，Leaf设备101和Leaf设备102称为MLAG成员，具有相同的虚IP地址(MLAG组地址)，记为IP12；Leaf设备103不属于Leaf设备101和Leaf设备102所属MLAG。各Leaf设备之间通过VXLAN隧道连接。
Leaf设备103在转发主机112访问主机111的报文时，对该报文进行VXLAN封装，封装报文的目的VTEP IP地址为虚IP地址IP12。Spine设备100接收到该报文，基于负载分担机制将该报文分发至Leaf设备101或Leaf设备102。若报文被分发至Leaf设备101，则由Leaf设备101将报文转发至主机111。参见图2A，为由Leaf设备101负责转发报文的路径示意图，其中，虚线所示为转发路径。
若Leaf设备101与主机111之间的物理链路发生故障，则报文转发至Leaf设备101后，由Leaf设备101通过VXLAN隧道将报文转发给Leaf设备102，再由Leaf设备102将报文转发给主机111。参见图2B，为链路故障之后的新的报文转发路径的示意图。从 图2B可以看出，在Leaf设备101处存在报文转发路径的迂回，导致报文的转发效率不高。
参见图3，为本公开实施例示出的一种报文转发方法的流程图，该方法从MLAG中的第一Leaf设备侧对报文转发过程进行描述。
步骤301，第一Leaf设备接收第二Leaf设备发送的第一路由撤销消息。
其中，第二Leaf设备与第一Leaf设备属于同一MLAG，具有相同的虚IP地址。
步骤302，若所述第一路由撤销消息中包括的目的IP地址为本地主机的IP地址，下一跳信息为所述虚IP地址，则第一Leaf设备发送第一路由发布消息。
若第一路由撤销消息中包括的目的IP地址为本地主机的IP地址，下一跳信息为第一Leaf设备和第二Leaf设备共有的虚IP地址(以下描述中简称为虚IP地址)，说明第二Leaf设备与本地主机之间的链路故障，无法正常通讯。以下描述中若无特殊说明，本地主机均指基于MLAG机制双归接入第一Leaf设备和第二Leaf设备的主机。第一Leaf设备发送第一路由发布消息，该第一路由发布消息中包括的目的IP地址为本地主机的IP地址，下一跳信息为第一Leaf设备的实IP地址(物理设备的IP地址)。图1中的Leaf设备101和Leaf设备102除了具有相同的虚IP地址IP12，还具有各自的实IP地址，分别记为IP11和IP22。
在本公开示例中描述的路由撤销消息和路由发布消息均可以是例如BGP消息的格式。Spine-Leaf组网中的任一设备均可将更新的路由携带在BGP消息中发送给与该设备建立了BGP邻居关系的另一设备。
在本公开实施例中，Spine-Leaf组网还包括不属于第一Leaf设备和第二Leaf设备所属MLAG的第三Leaf设备(例如，图1中的Leaf设备103)。第三Leaf设备接收到第二Leaf设备发送的第一路由撤销消息时，根据第一路由撤销消息中包括的目的IP地址和下一跳信息，删除本地记录的目的IP地址为本地主机的IP地址、下一跳信息为虚IP地址的路由表项。第三Leaf设备接收到第一Leaf设备发送的第一路由发布消息时，根据第一路由发布消息中包括的目的IP地址和下一跳信息，生成目的IP地址为本地主机的IP地址、下一跳信息为第一Leaf设备的实IP地址的路由表项(记为第一路由表项)。由此，第三Leaf设备基于第一路由表项转发发往本地主机的报文。具体地，在第三Leaf设备将报文转发出去之前，对报文进行VXLAN封装，其中，封装报文的目的VTEP IP地址为第一Leaf设备的实IP地址。然后，报文被转发至第一Leaf设备。通过这种方式， 避免第二Leaf设备与本地主机之间的链路故障时报文转发路径在第二Leaf设备处迂回，例如，报文从第二Leaf设备转发至第一Leaf设备再转发至本地主机。可见，本公开实施例提供的报文转发方法，可提升报文转发效率。
在一个示例中，当第一Leaf设备接收到第二Leaf设备发送的第二路由发布消息时，若该第二路由发布消息中包括的目的IP地址为本地主机的IP地址，下一跳信息为虚IP地址，说明第二Leaf设备与本地主机之间的链路恢复正常，可以通讯。此时，第一Leaf设备发送第二路由撤销消息，该第二路由撤销消息中包括的目的IP地址为本地主机的IP地址，下一跳信息为第一Leaf设备的实IP地址。
在一个示例中，当第三Leaf设备接收到第一Leaf设备发送的第二路由撤销消息时，根据第二路由撤销消息包括的目的IP地址和下一跳信息，删除本地记录的第一路由表项，即：目的IP地址为本地主机的IP地址、下一跳信息为第一Leaf设备的实IP地址的路由表项。当第三Leaf设备接收到第二Leaf设备发送的第二路由发布消息时，根据第二路由发布消息中包括的目的IP地址和下一跳信息，生成目的IP地址为本地主机的IP地址、下一跳信息为虚IP地址的路由表项(记为第二路由表项)。第三Leaf设备基于该第二路由表项转发发往本地主机的报文。具体地，第三Leaf设备在将报文转发出去之前，对报文进行VXLAN封装，其中，封装报文的目的VTEP IP地址为虚IP地址。然后，Spine设备基于负载分担机制将报文重新分发至第二Leaf设备，由第二Leaf设备转发报文给本地主机。也就是说，第二Leaf设备与本地主机之间的链路恢复正常后，报文转发路径也相应恢复。
此外，若第一Leaf设备接收到的报文的目的VTEP IP地址为虚IP地址，说明该报文要被发往基于MLAG双归接入第一Leaf设备和第二Leaf设备的本地主机。第一Leaf设备获取报文中携带的目的IP地址，该目的IP地址为本地主机的IP地址；基于该目的IP地址，查询ARP(Address Resolution Protocol，地址解析协议)表、MAC(Media Access Control，介质访问控制)地址表，确定连接本地主机的端口(也称为分布式聚合口)；基于该分布式聚合口的状态，确定第一Leaf设备与本地主机的通信状态。例如，若分布式聚合口的状态为UP状态，则确定第一Leaf设备与本地主机的通信状态为正常；若分布式聚合口的状态为DOWN状态(第一Leaf设备与本地主机之间的链路故障)，则确定第一Leaf设备与本地主机的通信状态为异常。
若第一Leaf设备与本地主机的通信状态为异常，则第一Leaf设备发送第三路由撤销消息，该第三路由撤销消息中包括的目的IP地址为本地主机的IP地址，下一跳信息 为虚IP地址。当第三Leaf设备接收到第三路由撤销消息时，根据该第三路由撤销消息包括的目的IP地址和下一跳信息，删除本地记录的第二路由表项，即目的IP地址为本地主机的IP地址、下一跳信息为虚IP地址的路由表项。当第二Leaf设备接收到第三路由撤销消息时，可感知第一Leaf设备与本地主机的通信状态异常。因此，第二Leaf设备发送第三路由发布消息，该第三路由发布消息中包括的目的IP地址为本地主机的IP地址，下一跳信息为第二Leaf设备的实IP地址。第三Leaf设备接收到该第三路由发布消息时，根据该第三路由发布消息包括的目的IP地址和下一跳信息，生成目的IP地址为本地主机的IP地址、下一跳信息为第二Leaf设备的实IP地址的路由表项(记为第三路由表项)。由此，第三Leaf设备基于该第三路由表项转发发往本地主机的报文。具体地，第三Leaf设备在将报文转发出去之前，对报文进行VXLAN封装，其中，封装报文的目的VTEP IP地址为第二Leaf设备的实IP地址。然后，报文被转发至第二Leaf设备。由此，避免第一Leaf设备与本地主机之间的通信异常时报文转发路径在第一Leaf设备处迂回，例如，报文从第一Leaf设备转发至第二Leaf设备再转发至本地主机。可见，本公开实施例提供的报文转发方法可提升报文转发效率。
若第一Leaf设备与本地主机的通信状态恢复正常，第一Leaf设备发送第四路由发布消息，该第四路由发布消息中包括的目的IP地址为本地主机的IP地址，下一跳信息为虚IP地址。第二Leaf设备接收到该第四路由发布消息时，可感知第一Leaf设备与本地主机的通信状态恢复正常，则发送第四路由撤销消息，该第四路由撤销消息中包括的目的IP地址为本地主机的IP地址，下一跳信息为第二Leaf设备的实IP地址。第三Leaf设备接收到该第四路由撤销消息时，根据该第四路由撤销消息包括的目的IP地址和下一跳信息，删除本地记录的第三路由表项，即目的IP地址为本地主机的IP地址、下一跳信息为第二Leaf设备的实IP地址的路由表项。当第三Leaf设备接收到第四路由发布消息时，根据该第四路由发布消息中包括的目的IP地址和下一跳信息，生成第二路由表项，即目的IP地址为本地主机的IP地址、下一跳信息为虚IP地址的路由表项。第三Leaf设备基于该第二路由表项转发发往本地主机的报文。具体地，第三Leaf设备在将报文转发出去之前，对报文进行VXLAN封装，其中，封装报文的目的VTEP IP地址为虚IP地址。Spine设备将报文基于负载分担机制重新分发至第一Leaf设备，由第一Leaf设备转发报文给本地主机。换言之，第一Leaf设备与本地主机之间的通信恢复正常后，报文的转发路径也相应恢复。
由于本公开实施例提供的报文转发方法只针对当前正在交互数据的主机进行路由更新，故障分布式聚合口下的其它未交互数据的主机的路由不更新，因此，可大大降低路 由更新的数量；当故障的分布式聚合口恢复正常时，也只需更新正在交互数据的主机的路由，因此，可避免大批量路由更新导致的路由震荡。
现仍以图1所示Spine-Leaf组网为例，详细介绍报文转发过程。
以主机112访问主机111为例，参见图4，为链路故障时各设备之间的交互流程图。
步骤401，Leaf设备103基于路由表项R1(目的IP地址为主机111的IP地址，下一跳信息为虚IP地址)转发主机112发往主机111的报文，对报文进行VXLAN封装，封装报文的目的VTEP IP地址为IP12(虚IP地址)。
步骤402，Leaf设备103将封装后的报文通过与Spine设备100连接的出接口发送给Spine设备100。
步骤403，Spine设备100接收到目的VTEP IP为IP12的报文后，基于负载分担机制将报文分发至Leaf设备101。
步骤404，Leaf设备101接收到目的VTEP IP为IP12的报文后，获取报文携带的主机111的IP地址，确定与主机111连接的分布式聚合口；若该分布式聚合口正常(处于UP状态)，则通过该分布式聚合口将报文发送至主机111，转发路径如图2A中虚线所示；若该分布式聚合口发生故障(DOWN状态)，转步骤405。
步骤405，Leaf设备101向Spine设备100(路由反射器)发送路由撤销消息，该路由撤销消息中包括的目的IP地址为主机111的IP地址，下一跳信息为虚IP地址。
步骤406，Spine设备100分别向Leaf设备102和Leaf设备103反射路由撤销消息。
步骤407，Leaf设备103接收到路由撤销消息后，根据该路由撤销消息中包括的目的IP地址和下一跳信息，删除本地记录的路由表项R1。
步骤408，Leaf设备102接收到路由撤销消息后，向Spine设备100发送路由发布消息，该路由发布消息中包括的目的IP地址为主机的IP地址，下一跳信息为Leaf设备102的实IP地址(IP22)。
步骤409，Spine设备100向Leaf设备103反射路由发布消息。
步骤410，Leaf设备103接收到路由发布消息后，根据该路由发布消息中包括的目的IP地址和下一跳信息，生成路由表项R2(目的IP地址为主机111的IP地址，下一跳信息为Leaf设备102的实IP地址)。
步骤411，Leaf设备103基于路由表项R2转发后续主机112发往主机111的报文， 对报文进行VXLAN封装，封装报文的目的VTEP IP地址为IP22(Leaf设备102的实IP地址)。
步骤412，Leaf设备103将封装后的报文通过与Spine设备100连接的出接口发送给Spine设备100。
步骤413，Spine设备100接收到目的VTEP IP为IP22的报文后，将报文直接转发给Leaf设备102。
步骤414，Leaf设备102通过本设备上与主机111连接的分布式聚合口，将报文发送至主机111，转发路径如图5中虚线所示。
参见图6，为链路恢复正常时各设备之间的交互流程图。
步骤601，若Leaf设备101连接主机111的分布式聚合口重新恢复为UP状态，则Leaf设备101向Spine设备100发送路由发布消息，该路由发布消息中包括的目的IP地址为主机111的IP地址，下一跳信息为虚IP地址(IP12)。
步骤602，Spine设备100向Leaf设备102和Leaf设备103反射路由发布消息。
步骤603，Leaf设备103接收到路由发布消息后，根据该路由发布消息中包括的目的IP地址和下一跳信息，生成路由表项R1(目的IP地址为主机111的IP地址，下一跳信息为虚IP地址)。
步骤604，Leaf设备102接收到路由发布消息后，向Spine设备100发送路由撤销消息，该路由撤销消息中包括的目的IP地址为主机111的IP地址，下一跳信息为Leaf设备102的实IP地址(IP22)。
步骤605，Spine设备100向Leaf设备103反射路由撤销消息。
步骤606，Leaf设备103接收到路由撤销消息后，根据该路由撤销消息中包括的目的IP地址和下一跳信息，删除本地记录的路由表项R2(目的IP地址为主机111的IP地址，下一跳信息为Leaf设备102的实IP地址)。
步骤607，Leaf设备103基于路由表项R1转发主机112发往主机111的报文，对报文进行VXLAN封装，封装报文的目的VTEP IP地址为IP12(虚IP地址)。
步骤608，Leaf设备103将封装后的报文通过与Spine设备100连接的出接口发送给Spine设备100。
步骤609，Spine设备100接收到目的VTEP IP为IP12的报文后，基于负载分担机 制重新将主机112发往主机111的报文分发至Leaf设备101。
步骤610，Leaf设备101通过本设备上与主机111连接的分布式聚合口，将报文发送至主机111，转发路径恢复到如图2A中虚线所示。
图7为本公开实施例提供的一种Leaf设备的硬件结构示意图。该Leaf设备可包括处理器701、存储有机器可执行指令的机器可读存储介质702。处理器701与机器可读存储介质702可经由系统总线703通信。并且，通过读取并执行机器可读存储介质702中与报文转发逻辑对应的机器可执行指令，处理器701可执行上文描述的报文转发方法。
本文中提到的机器可读存储介质702可以是任何电子、磁性、光学或其它物理存储装置，可以包含或存储信息，如可执行指令、数据，等等。例如，机器可读存储介质可以是：RAM(Radom Access Memory，随机存取存储器)、易失性存储器、非易失性存储器、闪存、存储驱动器(如硬盘驱动器)、固态硬盘、任何类型的存储盘(如光盘、dvd等)，或者类似的存储介质，或者它们的组合。
图8为本公开实施例提供的一种Leaf设备的结构示意图。该Leaf设备可以包括接收单元801和发送单元802。
接收单元801，用于接收所述第二Leaf设备发送的第一路由撤销消息。
发送单元802，用于若所述第一路由撤销消息中包括的目的IP地址为本地主机的IP地址，下一跳信息为所述虚IP地址，则发送第一路由发布消息，所述第一路由发布消息中包括的目的IP地址为所述本地主机的IP地址，下一跳信息为所述Leaf设备的实IP地址，以使不属于所述MLAG的第三Leaf设备生成第一路由表项，所述第一路由表项的目的IP地址为所述本地主机的IP地址，下一跳信息为所述Leaf设备的实IP地址。
可选的，所述接收单元801，还用于接收所述第二Leaf设备发送的第二路由发布消息；所述发送单元802，还用于若所述第二路由发布消息中包括的目的IP地址为所述本地主机的IP地址，下一跳信息为所述虚IP地址，则发送第二路由撤销消息，所述第二路由撤销消息中包括的目的IP地址为所述本地主机的IP地址，下一跳信息为所述Leaf设备的实IP地址，以使所述第三Leaf设备删除所述第一路由表项。
可选的，所述装置还包括：确定单元，用于确定所述Leaf设备与所述本地主机的通信状态。在这种情况下，所述发送单元802，还用于若所述Leaf设备与所述本地主机的通信状态为异常，发送第三路由撤销消息，所述第三路由撤销消息中包括的目的IP地址为所述本地主机的IP地址，下一跳信息为所述虚IP地址，以使所述第三Leaf设备删 除第二路由表项，所述第二路由表项的目的IP为所述本地主机的IP地址，下一跳信息为所述虚IP地址，并使得所述第二Leaf设备发送第三路由发布消息，所述第三路由发布消息中包括的目的IP地址为所述本地主机的IP地址，下一跳信息为所述第二Leaf设备的实IP地址。
可选的，所述确定单元，具体用于若接收到目的VTEP IP地址为所述虚IP地址的报文，获取所述报文中携带的目的IP地址，所述目的IP地址为所述本地主机的IP地址；基于所述目的IP地址，确定连接所述本地主机的分布式聚合口；基于所述分布式聚合口的状态，确定所述Leaf设备与所述本地主机的通信状态。
可选的，所述发送单元802，还用于若所述Leaf设备与所述本地主机的通信状态恢复正常，发送第四路由发布消息，所述第四路由发布消息中包括的目的IP地址为所述本地主机的IP地址，下一跳信息为所述虚IP地址，以使所述第三Leaf设备生成所述第二路由表项，并使得所述第二Leaf设备发送第四路由撤销消息，所述第四路由撤销消息中包括的目的IP地址为所述本地主机的IP地址，下一跳信息为所述第二Leaf设备的实IP地址。
本公开实施例还提供一种包括机器可执行指令的机器可读存储介质，例如图7中的机器可读存储介质702，所述机器可执行指令可由网络设备的主控板中的处理器701执行，以实现以上描述的报文转发方法。
以上所述仅为本公开的较佳实施例而已，并不用以限制本公开，凡在本公开的精神和原则之内，所做的任何修改、等同替换、改进等，均应包含在本公开保护的范围之内。

Claims (16)

  1. 一种报文转发方法，应用于跨设备链路聚合组MLAG中的第一叶子Leaf设备上，其特征在于，所述MLAG还包括第二Leaf设备，所述第一Leaf设备与所述第二Leaf设备具有相同的虚互联网协议IP地址，所述方法包括：
    接收所述第二Leaf设备发送的第一路由撤销消息；
    若所述第一路由撤销消息中包括的目的IP地址为双归接入所述第一Leaf设备和所述第二Leaf设备的第一主机的IP地址，下一跳信息为所述虚IP地址，则发送第一路由发布消息，所述第一路由发布消息中包括的目的IP地址为所述第一主机的IP地址，下一跳信息为所述第一Leaf设备的实IP地址，以使不属于所述MLAG的第三Leaf设备生成第一路由表项，所述第一路由表项的目的IP地址为所述第一主机的IP地址，下一跳信息为所述第一Leaf设备的实IP地址。
  2. 如权利要求1所述的方法，其特征在于，所述方法还包括：
    接收所述第二Leaf设备发送的第二路由发布消息；
    若所述第二路由发布消息中包括的目的IP地址为所述第一主机的IP地址，下一跳信息为所述虚IP地址，则发送第二路由撤销消息，所述第二路由撤销消息中包括的目的IP地址为所述第一主机的IP地址，下一跳信息为所述第一Leaf设备的实IP地址，以使所述第三Leaf设备删除所述第一路由表项。
  3. 如权利要求1所述的方法，其特征在于，所述方法还包括：
    确定所述第一Leaf设备与所述第一主机的通信状态；
    若所述第一Leaf设备与所述第一主机的通信状态为异常，发送第三路由撤销消息，所述第三路由撤销消息中包括的目的IP地址为所述第一主机的IP地址，下一跳信息为所述虚IP地址，以
    使所述第三Leaf设备删除第二路由表项，所述第二路由表项的目的IP地址为所述第一主机的IP地址，下一跳信息为所述虚IP地址，并
    使得所述第二Leaf设备发送第三路由发布消息，所述第三路由发布消息中包括的目的IP地址为所述第一主机的IP地址，下一跳信息为所述第二Leaf设备的实IP地址。
  4. 如权利要求3所述的方法，其特征在于，所述确定所述第一Leaf设备与所述第一主机的通信状态，包括：
    基于接收到的目的VXLAN隧道端点VTEP IP地址为所述虚IP地址的报文中携带的作为目的IP地址的所述第一主机的IP地址，确定连接所述第一主机的分布式聚合口；
    基于所述分布式聚合口的状态，确定所述第一Leaf设备与所述第一主机的通信状态。
  5. 如权利要求3所述的方法，其特征在于，所述方法还包括：
    若所述第一Leaf设备与所述第一主机的通信状态恢复正常，发送第四路由发布消息，所述第四路由发布消息中包括的目的IP地址为所述第一主机的IP地址，下一跳信息为所述虚IP地址，以
    使所述第三Leaf设备生成所述第二路由表项，并
    使得所述第二Leaf设备发送第四路由撤销消息，所述第四路由撤销消息中包括的目的IP地址为所述第一主机的IP地址，下一跳信息为所述第二Leaf设备的实IP地址。
  6. 一种报文转发装置，应用于跨设备链路聚合组MLAG中的第一叶子Leaf设备上，其特征在于，所述MLAG还包括第二Leaf设备，所述第一Leaf设备与所述第二Leaf设备具有相同的虚互联网协议IP地址，所述装置包括：
    接收单元，用于接收所述第二Leaf设备发送的第一路由撤销消息；
    发送单元，用于若所述第一路由撤销消息中包括的目的IP地址为双归接入所述第一Leaf设备和所述第二Leaf设备的第一主机的IP地址，下一跳信息为所述虚IP地址，则发送第一路由发布消息，所述第一路由发布消息中包括的目的IP地址为所述第一主机的IP地址，下一跳信息为所述第一Leaf设备的实IP地址，以使不属于所述MLAG的第三Leaf设备生成第一路由表项，所述第一路由表项的目的IP地址为所述第一主机的IP地址，下一跳信息为所述第一Leaf设备的实IP地址。
  7. 如权利要求6所述的装置，其特征在于：
    所述接收单元，还用于接收所述第二Leaf设备发送的第二路由发布消息；
    所述发送单元，还用于若所述第二路由发布消息中包括的目的IP地址为所述第一主机的IP地址，下一跳信息为所述虚IP地址，则发送第二路由撤销消息，所述第二路由撤销消息中包括的目的IP地址为所述第一主机的IP地址，下一跳信息为所述第一Leaf设备的实IP地址，以使所述第三Leaf设备删除所述第一路由表项。
  8. 如权利要求6所述的装置，其特征在于，所述装置还包括：
    确定单元，用于确定所述第一Leaf设备与所述第一主机的通信状态；
    所述发送单元，还用于若所述第一Leaf设备与所述第一主机的通信状态为异常，发送第三路由撤销消息，所述第三路由撤销消息中包括的目的IP地址为所述第一主机的IP地址，下一跳信息为所述虚IP地址，以使所述第三Leaf设备删除第二路由表项， 所述第二路由表项的目的IP地址为所述第一主机的IP地址，下一跳信息为所述虚IP地址，并使得所述第二Leaf设备发送第三路由发布消息，所述第三路由发布消息中包括的目的IP地址为所述第一主机的IP地址，下一跳信息为所述第二Leaf设备的实IP地址。
  9. 如权利要求8所述的装置，其特征在于：所述确定单元，具体用于
    基于接收到目的VXLAN隧道端点VTEP IP地址为所述虚IP地址的报文中携带的作为目的IP地址的所述第一主机的IP地址，确定连接所述第一主机的分布式聚合口；
    基于所述分布式聚合口的状态，确定所述第一Leaf设备与所述第一主机的通信状态。
  10. 如权利要求8所述的装置，其特征在于：所述发送单元，还用于
    若所述第一Leaf设备与所述第一主机的通信状态恢复正常，发送第四路由发布消息，所述第四路由发布消息中包括的目的IP地址为所述第一主机的IP地址，下一跳信息为所述虚IP地址，以使所述第三Leaf设备生成所述第二路由表项，并使得所述第二Leaf设备发送第四路由撤销消息，所述第四路由撤销消息中包括的目的IP地址为所述第一主机的IP地址，下一跳信息为所述第二Leaf设备的实IP地址。
  11. 一种Leaf设备，所述Leaf设备包括处理器和机器可读存储介质，所述机器可读存储介质存储有能够被所述处理器执行的机器可执行指令，所述处理器被所述机器可执行指令促使：
    接收跨设备链路聚合组MLAG中的第二Leaf设备发送的第一路由撤销消息，其中，所述MLAG包括所述Leaf设备以及所述第二Leaf设备，所述leaf设备与所述第二Leaf设备具有相同的虚互联网协议IP地址；
    若所述第一路由撤销消息中包括的目的IP地址为双归接入所述Leaf设备和所述第二Leaf设备的第一主机的IP地址，下一跳信息为所述虚IP地址，则发送第一路由发布消息，所述第一路由发布消息中包括的目的IP地址为所述第一主机的IP地址，下一跳信息为所述Leaf设备的实IP地址，以使不属于所述MLAG的第三Leaf设备生成第一路由表项，所述第一路由表项的目的IP地址为所述第一主机的IP地址，下一跳信息为所述Leaf设备的实IP地址。
  12. 如权利要求11所述的设备，其特征在于，所述机器可执行指令还促使所述处理器：
    接收所述第二Leaf设备发送的第二路由发布消息；
    若所述第二路由发布消息中包括的目的IP地址为所述第一主机的IP地址，下一跳 信息为所述虚IP地址，则发送第二路由撤销消息，所述第二路由撤销消息中包括的目的IP地址为所述第一主机的IP地址，下一跳信息为所述Leaf设备的实IP地址，以使所述第三Leaf设备删除所述第一路由表项。
  13. 如权利要求11所述的设备，其特征在于，所述机器可执行指令还促使所述处理器：
    确定所述Leaf设备与所述第一主机的通信状态；
    若所述Leaf设备与所述第一主机的通信状态为异常，发送第三路由撤销消息，所述第三路由撤销消息中包括的目的IP地址为所述第一主机的IP地址，下一跳信息为所述虚IP地址，以使所述第三Leaf设备删除第二路由表项，所述第二路由表项的目的IP地址为所述第一主机的IP地址，下一跳信息为所述虚IP地址，并使得所述第二Leaf设备发送第三路由发布消息，所述第三路由发布消息中包括的目的IP地址为所述第一主机的IP地址，下一跳信息为所述第二Leaf设备的实IP地址。
  14. 如权利要求13所述的设备，其特征在于，在确定所述Leaf设备与所述第一主机的通信状态时，所述机器可执行指令还促使所述处理器：
    基于接收到的目的VXLAN隧道端点VTEP IP地址为所述虚IP地址的报文中携带的作为目的IP地址的所述第一主机的IP地址，确定连接所述第一主机的分布式聚合口；
    基于所述分布式聚合口的状态，确定所述Leaf设备与所述第一主机的通信状态。
  15. 如权利要求13所述的设备，其特征在于，所述机器可执行指令还促使所述处理器：
    若所述Leaf设备与所述第一主机的通信状态恢复正常，发送第四路由发布消息，所述第四路由发布消息中包括的目的IP地址为所述第一主机的IP地址，下一跳信息为所述虚IP地址，以使所述第三Leaf设备生成所述第二路由表项，并使得所述第二Leaf设备发送第四路由撤销消息，所述第四路由撤销消息中包括的目的IP地址为所述第一主机的IP地址，下一跳信息为所述第二Leaf设备的实IP地址。
  16. 一种机器可读存储介质，其特征在于，所述机器可读存储介质内存储有机器可执行指令，所述机器可执行指令被处理器执行时实现权利要求1-5任一所述的方法步骤。
PCT/CN2019/082716 2018-04-19 2019-04-15 报文转发 Ceased WO2019201209A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201810355459.1A CN108600099B (zh) 2018-04-19 2018-04-19 一种报文转发方法、装置及叶子设备
CN201810355459.1 2018-04-19

Publications (1)

Publication Number Publication Date
WO2019201209A1 true WO2019201209A1 (zh) 2019-10-24

Family

ID=63611233

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2019/082716 Ceased WO2019201209A1 (zh) 2018-04-19 2019-04-15 报文转发

Country Status (2)

Country Link
CN (1) CN108600099B (zh)
WO (1) WO2019201209A1 (zh)

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113765823A (zh) * 2021-09-29 2021-12-07 新华三信息安全技术有限公司 一种报文转发方法及装置
CN116055388A (zh) * 2022-12-27 2023-05-02 新华三信息技术有限公司 一种自动部署M-lag组的方法、装置及设备
EP4358487A4 (en) * 2021-07-07 2024-09-11 ZTE Corporation Cross-device link aggregation routing calculation method, switch, system, and storage medium
CN118784573A (zh) * 2024-06-27 2024-10-15 新华三技术有限公司 一种流量转发方法及装置

Families Citing this family (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108600099B (zh) * 2018-04-19 2020-04-03 新华三技术有限公司 一种报文转发方法、装置及叶子设备
CN109495368B (zh) * 2018-12-19 2021-07-16 锐捷网络股份有限公司 Mac地址转发表的更新方法及网络设备
CN113300981B (zh) * 2020-02-21 2025-02-07 华为技术有限公司 报文传输方法、装置及系统
CN113381929B (zh) * 2020-03-10 2023-03-24 华为技术有限公司 一种路由处理方法、网关设备及计算机存储介质
CN111698154B (zh) * 2020-04-30 2023-04-07 新华三技术有限公司 一种主机路由频繁迁移的抑制方法及装置
CN115037602B (zh) * 2022-04-12 2026-01-02 新华三技术有限公司 一种故障处理方法及装置
CN115941589A (zh) * 2022-11-07 2023-04-07 新华三信息技术有限公司 一种故障处理方法及装置

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20140071810A1 (en) * 2012-09-11 2014-03-13 Hitachi Metals, Ltd. Communication system and processing method therefor
CN103986634A (zh) * 2013-02-13 2014-08-13 日立金属株式会社 通信系统及网络中继装置
CN104486124A (zh) * 2014-12-19 2015-04-01 盛科网络（苏州）有限公司 使用逻辑端口实现多系统链路聚合的装置及方法
CN107733793A (zh) * 2017-11-28 2018-02-23 新华三技术有限公司 一种转发表项维护方法及装置
CN108600099A (zh) * 2018-04-19 2018-09-28 新华三技术有限公司 一种报文转发方法、装置及Leaf设备

Family Cites Families (13)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102724126A (zh) * 2012-06-27 2012-10-10 福建星网锐捷网络有限公司 一种E-tree业务报文转发方法、装置和设备
CN103200089B (zh) * 2013-04-15 2016-10-19 杭州华三通信技术有限公司 虚拟专用局域网业务网络中的报文转发方法及设备
US10257043B2 (en) * 2016-01-11 2019-04-09 Lenovo Enterprise Solutions (Singapore) Pte. Ltd. Balancing utilization of infrastructure in a networked computing environment
CN106209643B (zh) * 2016-07-14 2019-08-06 新华三技术有限公司 报文转发方法及装置
CN106027352A (zh) * 2016-07-14 2016-10-12 瑞斯康达科技发展股份有限公司 一种业务报文转发方法和装置
CN106452992B (zh) * 2016-09-19 2020-02-11 新华三技术有限公司 一种远端多归属组网的实现方法及装置
CN107094110B (zh) * 2017-04-19 2020-01-03 新华三技术有限公司 一种dhcp报文转发方法及装置
CN106888166B (zh) * 2017-04-24 2020-04-24 新华三技术有限公司 一种报文转发方法和装置
CN107171964A (zh) * 2017-04-25 2017-09-15 中国联合网络通信集团有限公司 大规模组网中报文转发方法及装置
CN107547391B (zh) * 2017-06-08 2020-01-03 新华三技术有限公司 一种报文传输方法和装置
CN107547509B (zh) * 2017-06-27 2020-10-13 新华三技术有限公司 一种报文转发方法及装置
CN107547411B (zh) * 2017-06-30 2021-01-08 新华三技术有限公司 一种路由处理方法、装置、电子设备及机器可读存储介质
CN107547373B (zh) * 2017-07-17 2021-06-25 新华三技术有限公司 一种报文转发方法、装置、设备及机器可读存储介质

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20140071810A1 (en) * 2012-09-11 2014-03-13 Hitachi Metals, Ltd. Communication system and processing method therefor
CN103986634A (zh) * 2013-02-13 2014-08-13 日立金属株式会社 通信系统及网络中继装置
CN104486124A (zh) * 2014-12-19 2015-04-01 盛科网络（苏州）有限公司 使用逻辑端口实现多系统链路聚合的装置及方法
CN107733793A (zh) * 2017-11-28 2018-02-23 新华三技术有限公司 一种转发表项维护方法及装置
CN108600099A (zh) * 2018-04-19 2018-09-28 新华三技术有限公司 一种报文转发方法、装置及Leaf设备

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP4358487A4 (en) * 2021-07-07 2024-09-11 ZTE Corporation Cross-device link aggregation routing calculation method, switch, system, and storage medium
CN113765823A (zh) * 2021-09-29 2021-12-07 新华三信息安全技术有限公司 一种报文转发方法及装置
CN116055388A (zh) * 2022-12-27 2023-05-02 新华三信息技术有限公司 一种自动部署M-lag组的方法、装置及设备
CN118784573A (zh) * 2024-06-27 2024-10-15 新华三技术有限公司 一种流量转发方法及装置

Also Published As

Publication number Publication date
CN108600099A (zh) 2018-09-28
CN108600099B (zh) 2020-04-03

Similar Documents

Publication Publication Date Title
WO2019201209A1 (zh) 报文转发
US9019814B1 (en) Fast failover in multi-homed ethernet virtual private networks
US10333836B2 (en) Convergence for EVPN multi-homed networks
CN108259347B (zh) 一种报文传输方法和装置
US10135717B2 (en) Transferring data in a gateway
WO2018233588A1 (zh) 转发表项生成
WO2018099394A1 (zh) 报文传输
WO2019047855A1 (zh) 一种bras转控分离的备份方法和装置
US9716687B2 (en) Distributed gateways for overlay networks
US11165693B2 (en) Packet forwarding
CN103139071A (zh) 报文转发方法、装置和系统
CN104040527A (zh) 通过三层网络连接二层域
US11296979B2 (en) Method and system for symmetric integrated routing and bridging
CN107659484B (zh) 从vlan网络接入vxlan网络的方法、装置及系统
WO2018024187A1 (zh) 报文监控
WO2018090907A1 (zh) 报文转发
CN103118148A (zh) 一种arp缓存更新方法和设备
CN113381929A (zh) 一种路由处理方法、网关设备及计算机存储介质
JP2023544870A (ja) Ｍｌａｇリンク障害時切換え方法および装置
CN103631652A (zh) 虚拟机迁移的实现方法及系统
WO2021082803A1 (zh) 路由信息传输方法及装置、数据中心互联网络
CN115277305A (zh) 一种网络管理方法、装置、设备及机器可读存储介质
JP6887523B2 (ja) Ｍａｃアドレス同期
CN110708275B (zh) 一种协议报文的处理方法和装置
CN112866107B (zh) Ip地址通告方法、流量引导方法及网络设备

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: 19789253

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: 19789253

Country of ref document: EP

Kind code of ref document: A1