WO2018090907A1 - 报文转发 - Google Patents

报文转发 Download PDF

Info

Publication number
WO2018090907A1
WO2018090907A1 PCT/CN2017/110883 CN2017110883W WO2018090907A1 WO 2018090907 A1 WO2018090907 A1 WO 2018090907A1 CN 2017110883 W CN2017110883 W CN 2017110883W WO 2018090907 A1 WO2018090907 A1 WO 2018090907A1
Authority
WO
WIPO (PCT)
Prior art keywords
vtep
packet
address
port
routing entry
Prior art date
Application number
PCT/CN2017/110883
Other languages
English (en)
French (fr)
Inventor
付志华
Original Assignee
新华三技术有限公司
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 新华三技术有限公司 filed Critical 新华三技术有限公司
Priority to EP17872007.4A priority Critical patent/EP3528441B1/en
Priority to US16/461,563 priority patent/US11159341B2/en
Priority to JP2019526244A priority patent/JP6682045B2/ja
Publication of WO2018090907A1 publication Critical patent/WO2018090907A1/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
    • 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
    • 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
    • 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
    • H04L45/00—Routing or path finding of packets in data switching networks
    • H04L45/22—Alternate routing
    • 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
    • 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
    • H04L45/28—Routing or path finding of packets in data switching networks using route fault recovery
    • 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
    • H04L45/74—Address processing for routing
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L47/00—Traffic control in data switching networks
    • H04L47/10—Flow control; Congestion control
    • H04L47/12—Avoiding congestion; Recovering from congestion
    • H04L47/125—Avoiding congestion; Recovering from congestion by balancing the load, e.g. traffic engineering

Definitions

  • Multihoming networking is described in the Ethernet Virtual Private Network (EVPN) protocol.
  • EVPN Ethernet Virtual Private Network
  • the so-called multi-homing means that the same client belongs to more than two service provider edge devices, such as a virtual extended network (VXLAN) tunnel end point (VTEP: VXLAN Tunnel End Point).
  • VXLAN virtual extended network tunnel end point
  • Figure 1 is a flow chart of a method provided by the present disclosure
  • FIG. 2 is a schematic diagram of an application networking according to an embodiment of the present disclosure
  • FIG. 3 is a schematic diagram of an application networking according to another embodiment of the present disclosure.
  • FIG. 4 is a schematic diagram of an application networking according to another embodiment of the present disclosure.
  • FIG. 5 is a schematic diagram of an application networking according to another embodiment of the present disclosure.
  • Figure 6 is a structural view of a device provided by the present disclosure.
  • FIG. 7 is a structural diagram of hardware of a device provided by the present disclosure.
  • links of the same client accessing each VTEP can use the same ES (Ethernet Segment) identifier.
  • Each VTEP to which the client belongs is configured with the above ES identifier and advertised to the remote VTEP.
  • the remote VTEP associates the received ES identity with the VTEP that sent the ES identity.
  • the remote VTEP determines that the next hop of the route is the VTEP associated with the ES identifier, and then distributes the packet sent to the client to each of the clients belonging to the client according to the load balancing mode.
  • VTEP the next hop of the route is the VTEP associated with the ES identifier
  • the VTEP sends an ES Revocation message (the ES identity used by the failed link) to the remote VTEP.
  • the remote VTEP finds the route that meets the condition from all the routes associated with the ES identifier carried in the ES revocation message, and updates the next hop of the found route to other VTEPs to which the client belongs.
  • the condition that is satisfied is that the next hop of the route is the VTEP that sends the above ES revocation message.
  • the remote VTEP needs to process the ES revocation message sent by the VTEP connected to the faulty link, and on the other hand, the other information received is processed, which will increase The load of the remote VTEP causes the network to converge slowly.
  • each board gateway can be considered a VTEP.
  • the multi-homing networking mechanism described above cannot be applied to multiple board networking because the concept of ES is not involved in the multi-board networking.
  • the flow may include the following steps.
  • Step 101 Obtain a route of the specified role, the next hop of the route is a group address of the multi-homing group, and the route is advertised to the remote VTEP, so that the remote VTEP sends the report to the specified role according to the route.
  • the text is distributed to each VTEP in the multi-homing group.
  • the route can be advertised according to the routing protocol.
  • the application is applied to the EVPN.
  • the routing protocol may be an MP-BGP protocol, or an Intermediate System-to-Intermediate System (IS-IS) IS-IS protocol.
  • IS-IS Intermediate System-to-Intermediate System
  • the disclosure is not specifically limited.
  • the route here may be a Layer 2 route or a Layer 3 route.
  • all VTEPs to which a given role belongs are grouped into a multi-homing group in advance, after that,
  • the VTEP uses the group address of the multi-homing group as the next hop of the above route and advertises the route.
  • the remote VTEP determines that the next hop of the route is all the VTEPs included in the multi-homing group to which the specified role belongs, and then distributes the traffic sent to the specified role to each VTEP included in the multi-homing group according to the load balancing mode. Load sharing of links.
  • the so-called multi-board networking refers to the networking of multiple board gateways connected to the external network in the EVPN distributed gateway network.
  • the traffic destined for the external network is load-sharing between multiple homeboard gateways.
  • the multi-homed network is different from the multi-homed network.
  • the multi-homing network uses the ES identifier.
  • the multi-board network does not have the ES concept.
  • the multi-homing networking mechanism cannot be applied to multiple board networking.
  • Step 102 Receive a packet sent to the specified role from the remote VTEP, and send the packet to the multiple when receiving a link failure between the VTEP of the packet and the specified role.
  • Other VTEPs in the group are assigned to cause other VTEPs to send the message to the specified role.
  • all VTEPs to which a given role belongs are grouped into one multi-homing group.
  • the VTEP in the multi-homing group advertises the route of the specified role to the remote VTEP
  • the group address of the multi-homing group is advertised to the remote VTEP as the next hop of the route.
  • All VTEPs that are assigned to the specified role in the multi-homing network are no longer authorized to advertise the common ES. This allows the present disclosure to be applied not only to multi-homing networking but also to multi-Board networking, extending the application of the present disclosure.
  • the VTEP when a link between a VTEP and a specified role that belongs to a specified role is faulty, the VTEP sends the packet to other VTEPs in the multi-homing group to forward the packet to the specified role by other VTEPs.
  • the present disclosure can achieve fast network convergence regardless of whether it is applied to a multi-homed network or a multi-board network.
  • Each VTEP in a multi-homing group to which a specified role belongs is usually independent of each other, and there is no connection relationship between VTEPs.
  • the VTEP may send a message to other VTEPs in the multi-homing group when a link failure between the VTEP to which the specified role belongs and the specified role is broken.
  • an end-to-end link (Peer link) can be established serially between VTEPs to which a given role belongs.
  • a multi-homing group includes four VTEPs, namely VTEP1, VTEP2, VTEP3, and VTEP4, a Peer link can be established between VTEP1 and VTEP2, a Peer link can be established between VTEP2 and VTEP3, and a Peer link can be established between VTPE3 and VTEP4. .
  • VTEP1 and VTEP2 a Peer link can be established between VTEP2 and VTEP3
  • a Peer link can be established between VTPE3 and VTEP4.
  • the peer link supports all VLANs in the network.
  • sending the packet to the other VTEP in the multi-homing group includes: determining a VXLAN identifier (VNI) carried in the packet (for example, carried in the VXLAN encapsulation), and performing the packet on the packet.
  • VNI VXLAN identifier
  • the VXLAN is decapsulated, and the VLAN ID corresponding to the VNI is carried in the decapsulated packet and sent through the established Peer link.
  • FIG. 2 is a schematic diagram of an application networking according to an embodiment of the present disclosure.
  • the client is a customer-side edge device (CE: Customer Edge), which is denoted as CE 20.
  • CE Customer Edge
  • the VLAN to which the CE 20 belongs is recorded as VLAN1_1.
  • the CE 20 is attributed to the VTEP 201 and the VTEP 202.
  • VTEP 201 and VTEP 202 belong to the multi-homed group A, and the group address of the multi-homing group A is recorded as the group address 1_1.
  • a Peer link is established between VTEP 201 and VTEP 202.
  • Both VTEP 201 and VTEP 202 establish a BGP neighbor with the remote VTEP 203.
  • the VTEP 201 and the VTEP 202 both establish a BGP neighbor by using a respective loopback interface address and a remote VTEP 203.
  • the CE 20 is attributed to the VTEP 201 and the VTEP 202, only one of the VTEP 201 and the VTEP 202 may be based on the CE 20 due to the link aggregation characteristics of the respective links of the CE 20 to the VTEP 201 and the VTEP 202.
  • the sent packet learns the route of CE 20.
  • the CE 20 can send the message to VTEP 201 or VTEP 202 through HASH.
  • the route of the CE 20 learned by the VTEP 201 based on the packet sent by the CE 20 is taken as an example.
  • the route of the CE 20 learned by the VTEP 201 is recorded as the route 1_1.
  • the VTEP 201 generates a routing entry corresponding to the route 1_1 (denoted as the routing entry 1_1) and records it to the routing table.
  • the routing entry 1_1 can include a destination address, a VNI, and an egress port.
  • the destination address can be the MAC address or IP address of the CE 20, and the VNI corresponds to the VLAN to which the CE 20 belongs (ie, VLAN 1_1) and is recorded as VNI1_1.
  • the egress port is the port on the VTEP 201 that receives the packet sent by the CE 20. 2 Take Port1_1 as an example).
  • Table 1 shows the routing entry 1_1:
  • the VTEP 201 carries the next hop of the route 1_1 and the route 1_1 in the synchronization message and sends it to the VTEP 202 in the same home group.
  • the next hop of route 1_1 is group address 1_1.
  • the synchronization message here is a standard EVPN message.
  • the VTEP 201 advertises the route 1_1 to the remote VTEP 203, and the next hop of the advertised route 1_1 is the group address 1_1.
  • Routing entry 2_1 contains the destination address, VNI, and egress port.
  • the destination address may be the MAC address or the IP address of the CE 20, and the VNI is the VLAN identifier of the synchronization message, that is, the VNI corresponding to the VLAN 1_1 (that is, the VNI1_1), and the egress port is the VLAN ID of the VTEP 202 and the VLAN tag carried in the synchronization message, that is, VLAN1_1.
  • Corresponding port ( Figure 2 takes Port2_1 as an example). Table 2 shows routing entry 2_1:
  • the VTEP 202 advertises the route 1_1 to the remote VTEP 203, and the next hop of the advertised route 1_1 is the group address 1_1.
  • VTEP 202 can advertise route 1_1 through a BGP message.
  • the VTEP 203 When the VTEP 203 receives the route 1_1 advertised by the VTEP 201 and the VTEP 202 respectively, it is found that the next hop of the route 1_1 advertised by the VTEP 201 and the VTEP 202 is the group address of the multi-homing group A, and the next hop of the route 1_1 is determined to be Group address of multi-homed group A (including VTEP 201 and VTEP 202).
  • the VTEP 203 When the VTEP 203 receives the message destined for the CE 20, the VTEP 203 selects whether to send the message to the VTEP 201 or the VTEP 202 of the multi-homing group A according to the load sharing mode.
  • the VTEP 201 determines the VXLAN identifier VNI in the VXLAN package of the message, and performs VXLAN decapsulation on the packet, based on the VNI and The destination address of the packet is searched for the corresponding routing entry in the routing table, that is, the routing entry 1_1.
  • the outgoing packet of the routing entry 1_1, that is, Port1_1, is sent to the CE 20 by the VXLAN decapsulated packet.
  • the VTEP 202 determines the VXLAN identifier VNI in the VXLAN encapsulation of the packet, and performs VXLAN decapsulation on the packet, based on the VNI and The destination address of the packet is searched in the routing table for the corresponding routing entry, that is, the routing entry 2_1, and the outbound port of the routing entry 2_1, that is, Port2_1, is sent to the CE 20 by the VXLAN decapsulated packet.
  • the link failure between the VTEP 201 and the CE 20 is taken as an example. Then, the VTEP 201 first finds a matching routing entry in the routing table. The routing entry 1_1 deletes the outgoing port of the found routing entry 1_1 from Port1_1 to the peer port 1a of the device connected to the VTEP 202.
  • the VTEP 201 determines the corresponding routing entry in the routing table based on the VLAN identifier corresponding to the VXLAN identifier VNI in the VXLAN encapsulation of the packet and the destination address of the packet.
  • the routing entry 1_1 finds that the outgoing port of the routing entry 1_1 is the peer port 1a.
  • the VLAN ID corresponding to the VNI is carried in the decapsulated packet and sent through the established peer link.
  • the VNI is determined according to the VLAN identifier carried in the packet, and the routing table is determined in the routing table based on the destination address of the VNI and the packet.
  • the entry is the routing entry 2_1, and the packet is sent to the CE 20 through the outgoing port of the routing entry 2_1, that is, Port2_1.
  • the present embodiment has been described above by taking a multi-homing network as an example.
  • the remote VTEP 203 does not affect the packets sent to the CE 20 to the VTEP 201 and the VTEP 202 according to the load sharing mode.
  • the VTEP connected to the faulty link in the network is still involved in load balancing, which greatly improves the network link usage.
  • the VTEP 201 when the link between the VTEP 201 and the CE 20 fails, the VTEP 201 does not send an ES revocation message to the remote VTEP 203, and the remote VTEP 203 does not need to process the ES revocation message, which greatly improves the network convergence speed.
  • the related routing entries are modified only in the VTEP 201 to further improve the network convergence speed.
  • FIG. 3 is a schematic diagram of an application networking according to another embodiment of the present disclosure.
  • the external network belongs to the board gateway 401 and the board gateway 402.
  • the board gateway 401 and the board gateway 402 belong to the multi-homing group 4_1, and the group address of the multi-homing group 4_1 is recorded as the group address 4_1.
  • a Peer link is established between the Board Gateway 401 and the Board Gateway 402.
  • the board gateway 401 and the board gateway 402 both establish BGP neighbors with the remote board gateway 403.
  • the board gateway 401 and the board gateway 402 both establish a BGP neighbor by using the respective loopback address and the remote board gateway 403.
  • both the Board Gateway 401 and the Board Gateway 402 are configured with default routes.
  • the default routing entry configured by the board gateway 401 is recorded as the default routing entry 4_0
  • the default routing entry configured by the board gateway 402 is recorded as the default routing entry 0_4.
  • the default routing entry 4_0 is used by the board gateway 401 to send a message to the board gateway 402 through the Peer link with the board gateway 402 when the link between the board and the external network is faulty.
  • the default routing entry 4_0 contains the destination address, VNI, and egress port.
  • the destination address can be the MAC address or IP address of the client in the external network.
  • the VNI corresponds to the VLAN of the external network and is recorded as VNI4_1, and the egress port is the peer port 401a.
  • Table 3 shows the default routing entry 4_0:
  • the default routing entry 0_4 is used by the board gateway 402 to send a message to the board gateway 401 through the Peer link with the board gateway 401 when the link between the board and the external network is faulty.
  • the default routing entry 0_4 contains the destination address, VNI, and egress port.
  • the destination address can be the MAC address or IP address of the client in the external network.
  • the VNI corresponds to the VLAN of the external network and is recorded as VNI4_1, and the egress port is the peer port 402a.
  • Table 4 shows the default routing entry 0_4:
  • both the board gateway 401 and the board gateway 402 establish a connection with the external network based on the routing protocol. Therefore, the board gateway 401 and the board gateway 402 learn the route of the external network, and it is not necessary to learn in the multi-homed network. The route is sent to other board gateways in the same multi-homing group through the synchronization message.
  • Routing entry 4_1 contains the destination address, VNI, and egress port. Wherein, the destination address can be in an external network
  • the MAC address or IP address of the client, the VNI corresponds to the VLAN to which the external network belongs and is recorded as VNI4_1, and the outgoing port is the routing interface for the board gateway 401 to access the external network (Figure 4 takes Port4_1 as an example).
  • Table 5 shows routing entry 4_1:
  • the board gateway 401 advertises the route of the learned external network to the remote board gateway 403, and the next hop of the advertised route is the group address 4_1.
  • the board gateway 402 When learning the route of the external network, the board gateway 402 generates a corresponding routing entry (referred to as a routing table entry 4_2) and records it to the routing table.
  • the routing entry 4_2 contains the destination address, VNI, and egress port.
  • the destination address may be the MAC address or IP address of the client in the external network.
  • the VNI corresponds to the VLAN of the external network and is recorded as VNI4_1, and the egress port is the routing interface of the board gateway 402 to the external network ( Figure 5 uses Port4_2 as an example).
  • Table 6 shows the routing entry 4_2:
  • the board gateway 402 advertises the route of the learned external network to the remote board gateway 403, and the next hop of the advertised route is the group address 4_1.
  • the board gateway 403 receives the route of the external network advertised by the board gateway 401 and the board gateway 402 respectively, it is found that the next hop of the route advertised by the board gateway 401 and the board gateway 402 is the group address group address 4_1 of the multi-homing group 4_1, Then, the next hop of the route is determined to be the group address group address 4_1 of the multi-homing group 4_1.
  • the board gateway 403 When the board gateway 403 receives the packet destined for the external network, the board gateway 403 selects whether to send the packet to the board gateway 401 of the multi-homing group 4_1 or the board gateway 402 according to the load sharing manner.
  • the board gateway 403 sends the message to the board gateway 401
  • the board gateway 401 determines the VNI in the VXLAN package of the message, and performs VXLAN decapsulation on the packet.
  • the destination address of the VNI and the packet is used to find the corresponding routing entry in the routing table, that is, the routing entry 4_1, and the outbound port of the routing entry 4_1, that is, the port 4_1, sends the VXLAN decapsulated packet to the external network. .
  • the board gateway 403 sends the message to the board gateway 402
  • the board gateway 402 determines the VNI in the VXLAN package of the packet, and performs VXLAN decapsulation on the packet.
  • the destination address of the VNI and the packet is searched for the corresponding routing entry in the routing table, that is, the routing entry 4_2, and the VXLAN decapsulated packet is sent to the external network through the outbound port of the discovered routing entry 4_2, that is, Port4_2. .
  • the board gateway 401 finds the above route in the routing table. Entry 4_1 is deleted.
  • the board gateway 403 sends the message to the board gateway 401
  • the board gateway 401 determines the corresponding routing table item, that is, the above default routing table, in the routing table based on the VNI in the VXLAN encapsulation of the packet and the destination address of the packet. If the outbound port of the routing entry 4_0 is the peer port 401a, the VLAN ID corresponding to the VNI is carried in the decapsulated packet and sent through the established peer link.
  • the board gateway 402 When the board gateway 402 receives the packet through the established Peer link with the board gateway 401, the VNI is determined according to the VLAN identifier carried in the packet, and the routing entry is determined in the routing table based on the destination address of the VNI and the packet.
  • the routing entry 4_2 sends a packet to the external network through the outbound port of the routing entry 4_2, that is, Port4_2.
  • the present embodiment is described by taking a multi-board network as an example.
  • the VTEP In order to implement a link failure between the VTEP and the specified role to which the specified role belongs, the VTEP sends the message to other VTEPs in the multi-homing group.
  • the VTEPs in the multi-homing group can be serially established. Peer link.
  • a VXLAN tunnel can be established serially between the various VTEPs to which the assigned role belongs.
  • sending the message to other VTEPs in the multi-homing group includes: modifying an outer IP header of the packet (for example, an IP header in a VXLAN encapsulation) and passing the established VXLAN.
  • the tunnel sends a packet, where the modification of the IP header includes: modifying the source IP address in the IP header to the loopback address of the VTEP, and modifying the destination IP address in the IP header to be a VXLAN tunnel established in the multi-homed group with the VTEP.
  • the loopback address of other VTEPs includes: modifying an outer IP header of the packet (for example, an IP header in a VXLAN encapsulation) and passing the established VXLAN.
  • the tunnel sends a packet, where the modification of the IP header includes: modifying the source IP address in the IP header to the loopback address of the VTEP, and modifying the destination IP address in the IP header to be a VXLAN tunnel established in the multi-homed group with the VTEP.
  • FIG. 4 is a schematic diagram of an application networking of an embodiment provided by the present disclosure.
  • the client is CE5_1.
  • the VLAN to which CE5_1 belongs is recorded as VLAN5_1.
  • VTEP5_1 and VTEP5_2 belong to the multi-homing group 5_1, and the group address of the multi-homing group 5_1 is recorded as the group address 5_1.
  • a VXLAN tunnel is established between VTEP5_1 and VTEP5_2.
  • VTEP5_1 and VTEP5_2 both establish BGP neighbors with the remote VTEP5_3.
  • VTEP5_1, VTEP5_2 are The BGP neighbors are established using the respective loopback addresses (loopback address 51 and loopback address 52) and the remote VTEP5_3.
  • CE5_1 belongs to VTEP5_1 and VTEP5_2, only one of VTEP5_1 and VTEP5_2 learns the route of CE5_1 based on the packet sent by CE5_1.
  • CE5_1 can send packets to VTEP5_1 or VTEP5_2 through HASH.
  • the route of CE5_1 learned by VTEP5_1 based on CE5_1 is taken as an example.
  • the route of CE5_1 learned by VTEP5_1 is recorded as route 5_1.
  • VTEP5_1 generates a routing entry corresponding to route 5_1 (recorded as routing entry 5_1) and records it in the routing table.
  • Routing entry 5_1 contains the destination address, VNI, and egress port.
  • the destination address can be the MAC address or IP address of CE5_1.
  • the VNI corresponds to the VLAN to which CE5_1 belongs (that is, VLAN5_1) and is recorded as VNI5_1.
  • the egress port is the port on VTEP5_1 that receives the packet sent by CE5_1 ( Figure 4 uses Port5_1 as the port). example).
  • Table 7 shows the routing entry 5_1:
  • VTEP5_1 sends a synchronization message to VTEP5_2 in the same home group.
  • the synchronization message carries the next hop of route 5_1 and route 5_1, and the next hop of route 5_1 is group address 5_1.
  • VTEP5_1 advertises route 5_1 to the remote VTEP5_3.
  • the next hop of the advertised route 5_1 is the group address 5_1.
  • Routing entry 5_2 contains the destination address, VNI, and egress port.
  • the destination address can be the MAC address or the IP address of the CE5_1.
  • the VNI is the VLAN ID of the synchronization message, that is, the VNI corresponding to the VLAN 5_1 (that is, the VNI5_1).
  • the egress port is: VTEP5_2 corresponds to the VLAN ID of the synchronization message, that is, VLAN5_1.
  • Port ( Figure 4 takes Port5_2 as an example). Table 8 shows the routing entry 5_2:
  • VTEP5_2 advertises route 5_1 to the remote VTEP5_3.
  • the next hop of the advertised route 5_1 is the group address 5_1.
  • the VTEP5_3 receives the route 5_1 advertised by the VTEP5_1 and the VTEP5_2 respectively, it is found that the next hop of the route 5_1 advertised by the VTEP5_1 and the VTEP5_2 is the group address group5_1 of the multi-homing group 5_1, and the next hop of the route 5_1 is determined to be the multi-homing group 5_1.
  • the group address group5_1 (including VTEP5_1 and VTEP5_2).
  • VTEP5_3 When VTEP5_3 receives the packet destined for CE5_1, VTEP5_3 selects whether to send the packet to VTEP5_1 or VTEP5_2 of the multi-homing group 5_1 according to the load sharing mode.
  • VTEP5_3 When VTEP5_3 sends a packet to VTEP5_1, if the link between VTEP5_1 and CE5_1 is normal, VTEP5_1 determines the VXLAN identifier VNI in the VXLAN encapsulation of the packet, and looks up the corresponding table in the routing table based on the VNI and the destination address of the packet.
  • the routing entry that is, the routing entry 5_1, sends the VXLAN decapsulated packet to CE5_1 through the outgoing port of the routing entry 5_1, that is, Port5_1.
  • VTEP5_3 When VTEP5_3 sends a packet to VTEP5_2, if the link between VTEP5_2 and CE5_1 is normal, VTEP5_2 determines the VXLAN identifier VNI in the VXLAN encapsulation of the packet, and searches for the corresponding in the routing table based on the VNI and the destination address of the packet.
  • the routing entry that is, the routing entry 5_2, sends the VXLAN decapsulated packet to CE5_1 through the outgoing port of the routing entry 5_2, that is, Port5_2.
  • VTEP5_1 finds the corresponding routing entry in the routing table. 5_1. The outbound port of the found routing entry 5_1 is changed from Port5_1 to the loopback address 52 of VTEP5_2.
  • VTEP5_3 sends the packet to VTEP5_1
  • VTEP5_1 determines the corresponding routing entry in the routing table based on the VNI and the destination address of the packet in the VXLAN encapsulation of the packet, that is, the above routing entry 5_2, and finds the route at this time.
  • the outgoing port of the entry 5_2 is the loopback address 52 of the VTEP5_2.
  • the IP header in the VXLAN encapsulation of the packet is modified and sent through the established VXLAN tunnel between VTEP5_1 and VTEP5_2.
  • the IP header is modified to: change the source IP address in the IP header to the loopback address 51 of the VTEP5_1, and modify the destination IP address in the IP header to the loopback address 52 of the outgoing port of the routing entry 5_2, that is, VTEP5_2.
  • the VTEP5_2 When the VTEP5_2 receives the packet through the established VXLAN tunnel with the VTEP5_1, it determines the VNI carried in the VXLAN encapsulation of the packet, and determines the routing entry in the routing table according to the determined VNI and the destination address of the decapsulated packet. That is, the above routing entry 5_2 sends a packet to CE5_1 through the outbound port of the routing entry 5_2, that is, Port5_2.
  • the present embodiment has been described above by taking a multi-homing network as an example.
  • FIG. 5 is a schematic diagram of an application networking according to another embodiment of the present disclosure.
  • the external network belongs to the Board Gateway 6_1 and the Board Gateway 6_2.
  • the board gateway 6_1 and the board gateway 6_2 belong to the multi-homing group 6_1, and the group address of the multi-homing group 6_1 is recorded as the group address 6_1.
  • a VXLAN tunnel is established between the Board Gateway 6_1 and the Board Gateway 6_2.
  • the board gateway 6_1 and the board gateway 6_2 establish a BGP neighbor relationship with the remote board gateway 6_3.
  • the board gateway 6_1 and the board gateway 6_2 both establish a BGP neighbor by using a respective loopback address (a loopback address 61 and a loopback address 62) and a remote board gateway 6_3.
  • both the Board Gateway 6_1 and the Board Gateway 6_2 are configured with default routes.
  • the default routing entry configured by the board gateway 6_1 is recorded as the default routing entry 6_0
  • the default routing entry configured by the board gateway 6_2 is recorded as the default routing entry 0_6.
  • the default routing entry 6_0 is used by the board gateway 6_1 to send a message to the board gateway 6_2 through a VXLAN tunnel with the board gateway 6_2 when the link between the board and the external network is faulty.
  • the default routing entry 6_0 contains the destination address, VNI, and egress port.
  • the destination address may be the MAC address or the IP address of the client in the external network.
  • the VNI corresponds to the VLAN of the external network and is recorded as VNI6_1, and the outgoing port is the loopback address 62 of the board gateway 6_2.
  • Table 9 shows the default routing entry 6_0:
  • the default routing entry 0_6 is used by the board gateway 6_2 to send a message to the board gateway 6_1 through a VXLAN tunnel with the board gateway 6_1 when the link between the board and the external network is faulty.
  • the default routing entry 0_6 contains the destination address, VNI, and egress port.
  • the destination address may be the MAC address or IP address of the client in the external network.
  • the VNI corresponds to the VLAN to which the external network belongs and is recorded as VNI6_1, and the outgoing port is the loopback address 61 of the board gateway 6_1.
  • Table 10 shows the default routing entry 0_6:
  • the board gateway 6_1, the board gateway 6_2, and the external network establish a routing protocol. Therefore, the board gateway 8_1 and the board gateway 8_2 directly learn the route of the external network, and there is no need to learn the route in the multi-homed network. It is sent to other board gateways in the same multi-homing group through synchronization messages.
  • Routing entry 6_1 contains the destination address, VNI, and egress port.
  • the destination address can be the MAC address or IP address of the client in the external network.
  • the VNI corresponds to the VLAN of the external network and is recorded as VNI6_1.
  • the outgoing port is the routing interface of the board gateway 6_1 to access the external network (Figure 5 takes Port6_1 as an example).
  • Table 11 shows routing entry 6_1:
  • the board gateway 6_1 advertises the route of the learned external network to the remote board gateway 6_3.
  • the next hop of the advertised route is the group address 6_1.
  • Routing entry 6_2 contains the destination address, VNI, and egress port.
  • the destination address can be the MAC address or IP address of the client in the external network.
  • the VNI corresponds to the VLAN of the external network and is recorded as VNI6_1.
  • the outgoing port is the routing interface of the board gateway 6_2 to access the external network (Figure 5 takes Port6_2 as an example).
  • Table 12 shows routing entry 6_2:
  • the Board gateway 6_2 advertises the route of the learned external network to the remote board gateway 6_3.
  • the next hop of the advertised route is the group address 6_1.
  • the board gateway 6_3 receives the route of the external network advertised by the board gateway 6_1 and the board gateway 6_2, it is found that the next hop of the route advertised by the board gateway 6_1 and the board gateway 6_2 is the group address group address 6_1 of the multi-homing group 6_1. Then, it is determined that the next hop of the route is the group address group address 6_1 of the multi-homing group 6_1.
  • the board gateway 6_3 When the board gateway 6_3 receives the packet destined for the external network, the board gateway 6_3 will follow the load sharing side.
  • the mode selects whether the message is sent to the board gateway 6_1 of the multi-homing group 6_1 or the board gateway 6_2.
  • the board gateway 6_3 sends the message to the board gateway 6_1
  • the board gateway 6_1 determines the VNI in the VXLAN package of the message, based on the VNI and the destination address of the message.
  • the routing table entry is searched for the corresponding routing entry, that is, the routing entry 6_1, and the outbound port of the routing entry 6_1, that is, the port 6_1, is sent to the external network by the VXLAN decapsulated packet.
  • the board gateway 6_3 sends the message to the board gateway 6_2
  • the board gateway 6_2 determines the VNI in the VXLAN package of the message, based on the VNI and the destination address of the message.
  • the routing table entry is searched for the corresponding routing entry, that is, the routing entry 6_2, and the VXLAN decapsulated packet is sent to the external network through the outbound port of the routing entry 6_2, that is, Port6_2.
  • the board gateway 6_1 finds the above route in the routing table. Entry 6_1 is deleted.
  • the board gateway 6_3 sends the message to the board gateway 6_1
  • the board gateway 6_1 determines the corresponding routing entry in the routing table based on the VNI in the VXLAN encapsulation of the packet and the destination address of the packet, that is, the above default routing table.
  • the IP header in the VXLAN encapsulation of the packet is modified according to the default routing entry 6_0, and is sent through the established VXLAN tunnel between the board gateway 6_1 and the board gateway 6_2.
  • the IP header in the VXLAN encapsulation of the modified packet is: the source IP address in the IP header is changed to the loopback address 61 of the VTEP6_1, and the destination IP address in the IP header is modified to the outbound port of the default routing entry 6_0.
  • the loopback address of VTEP6_2 is 62.
  • the board gateway 6_2 When the board gateway 6_2 receives the packet through the established VXLAN tunnel with the board gateway 6_1, it determines the VNI in the VXLAN encapsulation of the packet, and determines the routing entry in the routing table according to the determined VNI and the destination address of the packet. That is, the routing entry 6_2 described above sends a packet to the external network through the outgoing port of the routing entry 6_2, that is, Port6_2.
  • FIG. 6 is a structural diagram of a device provided by the present disclosure.
  • the apparatus is applied to any of the VTEPs in the multi-homing group, which includes all of the VTEPs to which the assigned role belongs.
  • the device includes:
  • the obtaining unit 601 is configured to obtain a route of the specified role, where the multi-homing group includes different VTEPs to which the specified role belongs;
  • the advertising unit 602 is configured to advertise the route to the remote VTEP, so that the remote VTEP distributes the packet sent to the specified role to each VTEP in the multi-homing group according to the route, where The next hop of the route advertised to the remote VTEP is the group address of the multi-homing group;
  • the processing unit 603 is configured to receive, from the remote VTEP, a packet sent to the specified role, and send the packet to the multi-homing when a link between the VTEP and the specified role is faulty. Other VTEPs in the group to cause other VTEPs to send the message to the specified role.
  • the end-to-end link Peer link is established serially between the VTEPs in the multi-homing group;
  • the processing unit 603 sends the message to other VTEPs in the multi-homing group, including:
  • the packet is decapsulated by the VXLAN to obtain the decapsulated packet, and the VLAN identifier corresponding to the VNI is carried in the decapsulated packet and sent through the established Peer link.
  • the designated role is a client; the processing unit 703 further determines, in the routing table, a routing entry that satisfies the following conditions when the link between the VTEP and the specified role is faulty: the routing table The egress port of the entry is the port connected to the faulty link on the VTEP, and the egress port in the determined routing entry is changed to the peer port on the VTEP;
  • the processing unit 603 carries the VLAN identifier corresponding to the VNI in the decapsulated packet and sends the packet through the established Peer link, including: determining the routing entry in the routing table based on the destination address of the VNI and the packet, and corresponding to the VNI
  • the VLAN ID is carried in the decapsulated packet and is sent through the outbound port in the determined routing entry.
  • the egress port is a peer port.
  • the VTEP is a board gateway; the processing unit 603 further finds and deletes a routing entry that satisfies the condition in the routing table when the link between the VTEP and the specified role is faulty.
  • the condition is that the outbound port of the routing entry is a port that connects to the faulty link on the VTEP;
  • the processing unit 603 carries the VLAN identifier corresponding to the VNI in the decapsulated packet and sends the message through the established Peer link, including: setting the first default route in the routing table based on the destination address of the VNI and the packet.
  • the VLAN ID corresponding to the VNI is carried in the decapsulated packet and sent through the outbound port of the first default route.
  • the egress port of the first default route is the peer port.
  • a VXLAN tunnel is established serially between VTEPs in the multi-homing group; the processing unit 603 sends the message to other VTEPs in the multi-homing group including:
  • the outer IP header of the modified packet includes: modifying the source IP address in the outer IP header to be the loopback address of the loopback interface of the VTEP.
  • the destination IP address is a loopback address of another VTEP in the multi-homing group that has established a VXLAN tunnel with the VTEP.
  • the processing unit 603 when the link between the VTEP and the specified role is faulty, the processing unit 603 further determines, in the routing table, a routing entry that meets the following conditions: the outgoing port of the routing entry is The port connected to the faulty link on the VTEP, and the outbound port in the determined routing entry is modified to be the loopback address of the other VTEP in the multi-homing group that has established the VXLAN tunnel with the VTEP;
  • the processing unit 603 modifies the outer IP header of the packet and sends the VXLAN tunnel through the established VXLAN tunnel, including: determining the routing entry in the routing table based on the VNI and the destination address of the packet, and determining the routing table. If the egress port of the entry is a loopback address, the source IP address of the outer IP header of the packet is the loopback address of the VTEP. The destination IP address is the egress port and is sent to the egress port.
  • the VTEP is a board gateway; the processing unit 603 further determines, when the link between the VTEP and the specified role is faulty, a routing entry that satisfies the condition in the routing table, and deletes the The condition is: the egress port is the port connecting the faulty link on the VTEP;
  • the processing unit 603 modifies the outer IP header of the packet and sends the VXLAN tunnel through the established VXLAN tunnel, including: determining, according to the destination address of the VNI and the packet, the second default route set in the routing table, and modifying the report.
  • the source IP address of the outer IP header is the loopback address of the VTEP
  • the destination IP address is the egress port of the second default route, and is sent to the egress port of the second default route
  • the egress port of the second default route A loopback address of another VTEP that establishes a VXLAN tunnel with the VTEP in the multi-homing group.
  • FIG. 7 is a schematic structural diagram of a VTEP provided by the present disclosure.
  • the apparatus includes a processor 701, a machine readable storage medium 702 storing machine executable instructions corresponding to message forwarding control logic.
  • Processor 701 and machine readable storage medium 702 can communicate via system bus 703. And, by reading and executing machine executable instructions in machine readable storage medium 702, processor 701 can perform the message forwarding method described above.
  • the structure may also include a power supply component 704 configured to perform power management of the device, a wired or wireless network interface 705 configured to connect the device to the network, and an input/output (I/O) interface. 706.
  • I/O input/output
  • 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 disk drive), solid state drive, any type of storage disk. .
  • the machine readable storage medium 702 is configured to store machine executable instructions corresponding to the message forwarding control logic, where the machine executable instructions include the program instructions run by the obtaining unit 601, the program instructions run by the notification unit 602, and the processing unit 603. Program instructions.
  • processor 701 when processor 701 reads and executes machine executable instructions in machine readable storage medium 702, the processor 701 is caused to perform the following operations:
  • the next hop of the route to the remote VTEP is the group address of the multi-homing group
  • the VLAN identifier corresponding to the VNI is carried in the decapsulated packet and sent through the established Peer link.
  • the processor 701 is prompted by the machine executable instructions:
  • an outgoing port of the routing entry is a port that is connected to the faulty link on the VTEP,
  • the VLAN ID corresponding to the VNI is carried in the decapsulated packet and sent through the outbound port in the determined routing entry.
  • the egress port is a peer port.
  • the VTEP is a border board gateway; when a link between the VTEP and the specified role fails, the processor 701 is prompted by the machine executable instructions:
  • the routing entry that meets the condition is found in the routing table and is deleted.
  • the condition is: the outbound port of the routing entry is the port connected to the faulty link on the VTEP;
  • the VLAN identifier corresponding to the VNI is carried in the decapsulated packet and sent through the established Peer link.
  • the processor is caused by the machine executable instructions:
  • the destination address of the VNI and the packet is determined by using the first default route in the routing table, and the VLAN identifier corresponding to the VNI is carried in the decapsulated packet and sent through the outbound port of the first default route.
  • the outbound port of a default route is the peer port.
  • a VXLAN tunnel is established serially between VTEPs in the multi-homing group
  • the processor 701 Upon transmitting the message to other VTEPs in the multi-homing group, the processor 701 is prompted by the machine executable instructions:
  • Modifying the outer layer protocol IP header of the packet and sending it through the established VXLAN tunnel, the modifying the outer IP header of the packet includes: modifying the source IP address in the outer IP header to be the VTEP Loopback address of the loopback interface.
  • the destination IP address is the loopback address of other VTEPs in the multi-homed group that have established a VXLAN tunnel with the VTEP.
  • the processor 701 is prompted by the machine executable instructions:
  • an outgoing port of the routing entry is a port that is connected to the faulty link on the VTEP,
  • the processor Upon modifying the outer IP header of the message and transmitting over the established VXLAN tunnel, the processor is prompted by the machine executable instructions:
  • the source IP address of the outer IP header of the packet is the loopback address of the VTEP, and the destination IP address is the egress port in the determined routing entry, and is sent to the egress port, where the determined The outbound port in the routing entry is the loopback address.
  • the VTEP is a border board gateway; when a link between the VTEP and the specified role fails, the processor 701 is prompted by the machine executable instructions:
  • the source IP address of the outer IP header of the packet is the loopback address of the VTEP
  • the destination IP address is the egress port of the second default route, and is sent to the egress port of the second default route.
  • the outbound port of the second default route is a loopback address of another VTEP in the multi-homing group that has established a VXLAN tunnel with the VTEP.
  • the processor 701 upon obtaining a route for a specified role, the processor 701 is prompted by the machine executable instructions:
  • the processor 701 is prompted by the machine executable instructions:
  • the routing entry corresponding to the route and recording the routing entry to the routing table, and sending a synchronization message to the other VTEPs in the multi-homing group, where the synchronization message carries the route and the VLAN identifier of the VLAN to which the route belongs; the routing entry The destination address, the VXLAN identifier VNI and the egress port carried in the packet, the destination address is the address of the specified role, the VNI corresponds to the VLAN to which the specified role belongs, and the egress port is the port on the VTEP that receives the packet sent by the specified role. ;
  • the processor 701 Upon obtaining a route for the specified role, the processor 701 is prompted by the machine executable instructions:
  • the processor 701 Upon learning the route from the received synchronization message, the processor 701 is prompted by the machine executable instructions:
  • the routing entry corresponding to the route is generated and recorded in the routing table, and the generated routing entry includes the destination address, the VNI, and the egress port, where the destination address is the address of the specified role, and the VNI corresponds to the VLAN identifier carried in the synchronization message.
  • the egress port is a port corresponding to the VLAN identifier carried in the synchronization message on the VTEP.
  • the device embodiment since it basically corresponds to the method embodiment, reference may be made to the partial description of the method embodiment.
  • the device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separate, and the components displayed as units may or may not be physical units, ie may be located A place, or it can be distributed to multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of the embodiment. Those of ordinary skill in the art can understand and implement without any creative effort.

Landscapes

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

Abstract

本公开提供了报文转发方法和VXLAN隧道端点VTEP。在一个示例中，VTEP获取指定角色的路由；将所述路由通告给远端VTEP，以使远端VTEP根据所述路由将发往所述指定角色的报文分发至多归属组中的各个VTEP，其中，所述多归属组包括所述指定角色归属的不同VTEP，所述路由的下一跳为所述多归属组的组地址。所述VTEP从所述远端VTEP接收发往所述指定角色的报文，当所述VTEP与所述指定角色之间的链路故障时，所述VTEP将所述报文发送至所述多归属组中的其他VTEP，以使其他VTEP将所述报文发送至所述指定角色。

Description

报文转发
相关申请的交叉引用
本专利申请要求于2016年11月17日提交的、申请号为201611033162.0、发明名称为“报文转发方法和装置”的中国专利申请的优先权，该申请的全文以引用的方式并入本文中。
背景技术
以太虚拟私有网络(EVPN：Ethernet Virtual Private Network)协议中描述了多归属(multihoming)组网。所谓多归属，是指同一客户端归属两个以上服务商边缘设备，例如虚拟扩展局域网(VXLAN：Virtual eXtensible LAN)隧道端点(VTEP：VXLAN Tunnel End Point)。
附图说明
此处的附图被并入说明书中并构成本说明书的一部分，示出了符合本公开的实施例，并与说明书一起用于解释本公开的原理。
图1为本公开提供的方法流程图；
图2为本公开提供的一实施例应用组网示意图；
图3为本公开提供的另一实施例的应用组网示意图；
图4为本公开提供的另一实施例的应用组网示意图；
图5为本公开提供的另一实施例的应用组网示意图；
图6为本公开提供的装置结构图；
图7为本公开提供的装置硬件结构图。
具体实施方式
下面将结合本公开实施例中的附图，对本公开实施例中的技术方案进行清楚、完整地描 述，显然，所描述的实施例仅仅是本公开一部分实施例，而不是全部的实施例。基于本公开中的实施例，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他实施例，都属于本申请保护的范围。
在多归属组网中，同一客户端接入各个VTEP的链路可使用同一以太段(ES：Ethernet Segment)标识。客户端归属的各个VTEP会配置上述ES标识，并通告给远端VTEP。远端VTEP将接收的ES标识与发送该ES标识的VTEP关联。在接收到上述其中一个VTEP通告的路由时，远端VTEP确定该路由的下一跳为上述ES标识关联的各个VTEP，之后按照负载分担方式将发送至客户端的报文分担至客户端归属的各个VTEP。但是，当其中一个VTEP与客户端之间的链路故障(down)时，该VTEP发送ES撤销消息(携带故障链路使用的ES标识)给远端VTEP。远端VTEP收到ES撤销消息后从与ES撤销消息携带的ES标识所关联的所有路由中找到满足条件的路由，将找到的路由的下一跳更新为客户端归属的其他VTEP。所满足的条件为：路由的下一跳为发送上述ES撤销消息的VTEP。但是，在VTEP与客户端之间的链路故障时，一方面需要远端VTEP处理故障链路连接的VTEP发送的ES撤销消息，另一方面还要处理接收到的其他信息，这会加大远端VTEP的负荷，并且导致网络收敛速度慢。
在一个例子中，在部署了多个边界(Board)网关的多Board组网中，每个Board网关可被视为一个VTEP。由于多Board组网中并不涉及ES的概念，因此以上描述的多归属组网机制无法应用于多Board组网。
参见图1，图1为本公开提供的方法流程图。该流程应用于多归属组中的任一VTEP。多归属组，其包括指定角色归属的所有VTEP。其中，多归属组通过一个组地址标识来唯一标识多归属组，不同多归属组具有不同的组地址。
如图1所示，该流程可包括以下步骤。
步骤101，获取指定角色的路由，所述路由的下一跳为多归属组的组地址，将路由通告给远端VTEP，以使远端VTEP根据所述路由将发往所述指定角色的报文分发至所述多归属组中的各个VTEP。
本步骤101中可以根据路由协议来通告路由。应用于EVPN，该路由协议可为MP-BGP协议，也可为中间系统到中间系统(IS-IS：Intermediate System-to-Intermediate System)IS-IS协议，本公开并不具体限定。作为一个实施例，这里的路由可为二层路由或三层路由。
可以看出，在本公开中，预先将指定角色归属的所有VTEP组成一个多归属组，之后， 当指定角色归属的任一VTEP学习到指定角色的路由时(下文会描述指定角色归属的VTEP如何学习路由)，VTEP以多归属组的组地址作为上述路由的下一跳、并将该路由通告给远端VTEP。如此，远端VTEP会确定该路由的下一跳为指定角色归属的多归属组中包括的所有VTEP，之后按照负载分担方式将发送至指定角色的流量分担至多归属组中包括的各个VTEP，实现链路的负载分担。
至于指定角色，在本公开中，作为一个实施例，在归属组网中，指定角色可为客户端。作为另一个实例，在多Board组网中，多归属组中的VTEP为Board网关，则指定角色为归属Board网关连接的外部网络。
所谓多Board组网是指在EVPN分布式网关组网里，部署了跟外部网络相连的多个Board网关的组网。发往外部网络的流量在归属的多个Board网关之间进行负载分担。
多Board组网跟多归属组网的区别在于多归属组网使用ES标识，而多Board组网并没有ES概念，多归属组网的机制不能应用于多Board组网。
步骤102，从远端VTEP接收发往所述指定角色的报文，当接收到该报文的本VTEP与所述指定角色之间的链路故障时，将所述报文发送至所述多归属组中的其他VTEP，以使其他VTEP将所述报文发送至所述指定角色。
通过图1所示流程可以看出，在本公开中，将指定角色归属的所有VTEP组成一个多归属组。多归属组中的VTEP在向远端VTEP通告指定角色的路由时，将多归属组的组地址作为路由的下一跳通告给远端VTEP。不再强制多归属组网中指定角色归属的所有VTEP通告共同的ES，这使得本公开不仅应用于多归属组网，也能应用于多Board组网，扩展了本公开的应用。
在一个示例中，当指定角色归属的VTEP与指定角色之间的链路故障时，该VTEP不会触发远端VTEP处理ES撤销消息，远端VTEP依然会向该VTEP发送发往指定角色的报文。即使应用于双归属组网，比如客户端双归属(dual-homed to)VTEP1和VTEP2，则当其中一个VTEP比如VTEP1与客户端之间的链路故障时，远端VTEP不会因为VTEP1与客户端之间的链路故障而不将发往客户端的报文发给VTEP1，依然将发往客户端的报文分发至VTEP1和VTEP2。通过远端VTEP与客户端双归属的两个VTEP之间的链路实现了负载分担。
在一个示例中，当指定角色归属的VTEP与指定角色之间的链路故障时，该VTEP会将报文发送至多归属组中其他VTEP，以由其他VTEP转发报文至指定角色，这能够保证本公开不管是应用于多归属组网还是多Board组网，都能实现网络快速收敛。
下面通过两个实施例对本公开提供的方法进行描述：
指定角色归属的多归属组中的各VTEP之间通常是相互独立的，各VTEP之间没有连接关系。然而，根据本公开的方法，在指定角色归属的VTEP与指定角色之间的链路故障时，该VTEP可将报文发送至多归属组中其他VTEP。鉴于此，在一个实施例中，指定角色归属的各VTEP之间可串行建立端至端链路(Peer link)。例如，多归属组中包括4个VTEP，即VTEP1、VTEP2、VTEP3和VTEP4，则VTEP1与VTEP2之间可建立Peer link，VTEP2与VTEP3之间可建立Peer link，VTPE3与VTEP4之间可建立Peer link。这仅是一个建立peer link的示例，不具有限制性。在本实施例中，peer link支持网络中所有VLAN。
基于此，在本实施例中，上述步骤102中，将报文发送至多归属组中其他VTEP包括：确定报文携带的(例如在VXLAN封装中携带)的VXLAN标识(VNI)，对报文进行VXLAN解封装，将VNI对应的VLAN标识携带在解封装后的报文中并通过已建立的Peer link发送。
下面以应用于双归属组网(多归属组网的一个特例)为例对本实施例进行描述：
参见图2，图2为本公开提供的一实施例的应用组网示意图。在图2中，客户端为客户侧边缘设备(CE：Customer Edge)，记为CE 20。CE 20所属的VLAN记为VLAN1_1。CE 20归属于VTEP 201和VTEP 202。VTEP 201和VTEP 202同属多归属组A，多归属组A的组地址记为group地址1_1。VTEP 201和VTEP 202之间建立了Peer link。
VTEP 201、VTEP 202均和远端VTEP 203建立BGP邻居。其中，VTEP 201、VTEP 202都是使用各自的环回(Loopback)接口地址和远端VTEP 203建立BGP邻居。
在图2中，尽管CE 20归属于VTEP 201和VTEP 202，但是，由于CE 20至VTEP 201和VTEP 202的各个链路的链路聚合特性，可以仅VTEP 201和VTEP 202中的一个基于CE 20发送的报文学习CE 20的路由。CE 20可以通过HASH方式将报文发到VTEP 201或VTEP202。
以VTEP 201基于CE 20发送的报文学习CE 20的路由为例，为便于描述，这里将VTEP201学习的CE 20的路由记为路由1_1。
VTEP 201生成路由1_1对应的路由表项(记为路由表项1_1)并记录至路由表。路由表项1_1可包含目的地址、VNI、出端口。其中，目的地址可为CE 20的MAC地址或者IP地址，VNI对应于CE 20所属的VLAN(即VLAN1_1)并记为VNI1_1，出端口为VTEP 201上接收到CE 20发送的报文的端口(图2以Port1_1为例)。表1示出了路由表项1_1：
Figure PCTCN2017110883-appb-000001
表1
VTEP 201将路由1_1和路由1_1的下一跳携带在同步消息中发送给同一归属组中的VTEP 202。路由1_1的下一跳为group地址1_1。这里的同步消息为标准的EVPN消息。
VTEP 201通告路由1_1给远端VTEP 203，通告的路由1_1的下一跳为group地址1_1。
当VTEP 202接收到VTEP 201发送的同步消息后，从同步消息学习路由1_1，生成路由1_1对应的路由表项(记为路由表项2_1)并记录至路由表。路由表项2_1包含目的地址、VNI、出端口。其中，目的地址可为CE 20的MAC地址或者IP地址，VNI为同步消息携带的VLAN标识即VLAN1_1对应的VNI(即上述的VNI1_1)，出端口为VTEP 202上与同步消息携带的VLAN标识即VLAN1_1对应的端口(图2以Port2_1为例)。表2示出了路由表项2_1：
Figure PCTCN2017110883-appb-000002
表2
VTEP 202通告路由1_1给远端VTEP 203，通告的路由1_1的下一跳为group地址1_1。这里，VTEP 202可通过BGP消息通告路由1_1。
当VTEP 203接收到VTEP 201和VTEP 202分别通告的路由1_1时，发现VTEP 201和VTEP 202通告的路由1_1的下一跳均为多归属组A的group地址，则确定路由1_1的下一跳为多归属组A(包含VTEP 201和VTEP 202)的group地址。
当VTEP 203接收到发往CE 20的报文时，VTEP 203会按照负载分担方式选择将报文发送至多归属组A的VTEP 201还是VTEP 202。
当VTEP 203将报文发送至VTEP 201时，若VTEP 201与CE 20之间的链路正常，则VTEP201确定报文的VXLAN封装中的VXLAN标识VNI，对报文进行VXLAN解封装，基于VNI和报文的目的地址在路由表中查找对应的路由表项即上述的路由表项1_1，通过查找到的路由表项1_1的出端口即Port1_1发送VXLAN解封装后的报文至CE 20。
当VTEP 203将报文发送至VTEP 202时，若VTEP 202与CE 20之间的链路正常，则VTEP202确定报文的VXLAN封装中的VXLAN标识VNI，对报文进行VXLAN解封装，基于VNI和报文的目的地址在路由表中查找对应的路由表项即上述的路由表项2_1，通过查找到的路由表项2_1的出端口即Port2_1发送VXLAN解封装后的报文至CE 20。
当VTEP 201和VTEP 202中一个VTEP与CE 20之间的链路故障时，以VTEP 201与CE20之间的链路故障为例，则，VTEP 201先在路由表中找到匹配的路由表项即上述路由表项1_1，将该找到的路由表项1_1的出端口从Port1_1修改为本设备上的与VTEP 202连接的Peer link端口1a。
之后，当VTEP 203将报文发送至VTEP 201时，VTEP 201基于报文的VXLAN封装中的VXLAN标识VNI对应的VLAN标识和报文的目的地址在路由表中确定对应的路由表项即上述的路由表项1_1，发现此时路由表项1_1的出端口为Peer link端口1a，则将VNI对应的VLAN标识携带在解封装后的报文中并通过已建立的Peer link发送。
当VTEP 202通过与VTEP 201之间已建立的Peer link从Peer link端口1b收到报文时，根据报文携带的VLAN标识确定VNI，基于VNI和报文的目的地址在路由表中确定路由表项即上述的路由表项2_1，通过路由表项2_1的出端口即Port2_1发送报文至CE 20。
以上以多归属组网为例对本实施例进行了描述。
通过上面描述可以看出，即使VTEP 201与CE 20之间的链路故障，但并不影响远端VTEP203按照负载分担方式将发向CE 20的报文分担至VTEP 201和VTEP 202，实现了多归属组网中故障链路连接的VTEP仍参与负载分担，大大提高了网络链路使用率。
需要指出的是，在VTEP 201与CE 20之间的链路故障时，VTEP 201并不向远端VTEP 203发送ES撤销消息，远端VTEP 203无需处理ES撤销消息，大大提高了网络收敛速度。
在VTEP 201与CE 20之间的链路故障时，仅在VTEP 201修改相关路由表项，进一步提高网络收敛速度。
下面再以双Board组网为例对本实施例进行描述：
参见图3，图3为本公开提供的另一实施例的应用组网示意图。在图3中，外部网络归属于Board网关401和Board网关402。Board网关401和Board网关402同属多归属组4_1，多归属组4_1的组地址记为group地址4_1。Board网关401和Board网关402之间建立了Peer link。
Board网关401、Board网关402均和远端Board网关403建立BGP邻居。其中，Board网关401、Board网关402都是使用各自的Loopback地址和远端Board网关403建立BGP邻居。
在图3中，Board网关401和Board网关402均配置了默认路由。为便于描述，将Board网关401配置的默认路由表项记为默认路由表项4_0，将Board网关402配置的默认路由表项记为默认路由表项0_4。
默认路由表项4_0用于Board网关401在与外部网络之间的链路故障时，通过与Board网关402之间的Peer link将报文发送给Board网关402。默认路由表项4_0包含目的地址、VNI、出端口。其中，目的地址可为外部网络中客户端的MAC地址或者IP地址，VNI对应于外部网络所属VLAN并记为VNI4_1，出端口为Peer link端口401a。表3示出了默认路由表项4_0：
Figure PCTCN2017110883-appb-000003
表3
默认路由表项0_4用于Board网关402在与外部网络之间的链路故障时，通过与Board网关401之间的Peer link将报文发送给Board网关401。默认路由表项0_4包含目的地址、VNI、出端口。其中，目的地址可为外部网络中客户端的MAC地址或者IP地址，VNI对应于外部网络所属VLAN并记为VNI4_1，出端口为Peer link端口402a。表4示出了默认路由表项0_4：
Figure PCTCN2017110883-appb-000004
表4
在图3中，Board网关401、Board网关402均和外部网络基于路由协议建立连接，因此，Board网关401、Board网关402上都会学习到外部网络的路由，没有必要像多归属组网中将学习的路由通过同步消息发送给同一多归属组中其他Board网关。
Board网关401学习到外部网络的路由时生成对应的路由表项(记为路由表项4_1)并记录至路由表。路由表项4_1包含目的地址、VNI、出端口。其中，目的地址可为外部网络中 客户端的MAC地址或者IP地址，VNI对应于外部网络所属VLAN并记为VNI4_1，出端口为Board网关401接入外部网络的路由接口(图3以Port4_1为例)。表5示出了路由表项4_1：
Figure PCTCN2017110883-appb-000005
表5
Board网关401通告学习的外部网络的路由给远端Board网关403，通告的路由的下一跳为group地址4_1。
Board网关402在学习到外部网络的路由时生成对应的路由表项(记为路由表项4_2)并记录至路由表。路由表项4_2包含目的地址、VNI、出端口。其中，目的地址可为外部网络中客户端的MAC地址或者IP地址，VNI对应于外部网络所属VLAN并记为VNI4_1，出端口为Board网关402接入外部网络的路由接口(图5以Port4_2为例)。表6示出了路由表项4_2：
Figure PCTCN2017110883-appb-000006
表6
Board网关402通告学习的外部网络的路由给远端Board网关403，通告的路由的下一跳为group地址4_1。
当Board网关403接收到Board网关401和Board网关402分别通告的外部网络的路由时，发现Board网关401和Board网关402通告的路由的下一跳均为多归属组4_1的组地址group地址4_1，则确定该路由的下一跳为多归属组4_1的组地址group地址4_1。
当Board网关403接收到发往外部网络的报文时，Board网关403会按照负载分担方式选择将报文发送至多归属组4_1的Board网关401还是Board网关402。
当Board网关403将报文发送至Board网关401时，若Board网关401与外部网络之间的链路正常，则Board网关401确定报文的VXLAN封装中的VNI，对报文进行VXLAN解封装，基于VNI和报文的目的地址在路由表中查找对应的路由表项即上述的路由表项4_1，通过查找到的路由表项4_1的出端口即Port4_1发送VXLAN解封装后的报文至外部网络。
当Board网关403将报文发送至Board网关402时，若Board网关402与外部网络之间的链路正常，则Board网关402确定报文的VXLAN封装中的VNI，对报文进行VXLAN解封装，基于VNI和报文的目的地址在路由表中查找对应的路由表项即上述的路由表项4_2，通过查找到的路由表项4_2的出端口即Port4_2发送VXLAN解封装后的报文至外部网络。
当Board网关401和Board网关402中一个Board网关与外部网络之间的链路故障时，以Board网关401与外部网络之间的链路故障为例，则Board网关401在路由表中找到上述路由表项4_1并删除。
之后，当Board网关403将报文发送至Board网关401时，Board网关401基于报文的VXLAN封装中的VNI和报文的目的地址在路由表中确定对应的路由表项即上述的默认路由表项4_0，发现此时路由表项4_0的出端口为Peer link端口401a，则将VNI对应的VLAN标识携带在解封装后的报文中并通过已建立的Peer link发送。
当Board网关402通过与Board网关401之间已建立的Peer link收到报文时，根据报文携带的VLAN标识确定VNI，基于该VNI和报文的目的地址在路由表中确定路由表项即上述的路由表项4_2，通过路由表项4_2的出端口即Port4_2发送报文至外部网络。
以上以多Board组网为例对本实施例进行了描述。
为了实现指定角色归属的VTEP与指定角色之间的链路故障时，该VTEP将报文发送至多归属组中其他VTEP，在上述实施例中，在多归属组的各VTEP之间可串行建立Peer link。然而，在另一实施例中，指定角色归属的各VTEP之间可串行建立VXLAN隧道。
基于此，在一实施例中，上述步骤102中，将报文发送至多归属组中其他VTEP包括：修改报文的外层IP头(例如是VXLAN封装中的IP头)并通过已建立的VXLAN隧道发送报文，其中，修改IP头包括：将IP头中的源IP地址修改为本VTEP的Loopback地址，将IP头中的目的IP地址修改为多归属组中与本VTEP建立了VXLAN隧道的其他VTEP的Loopback地址。
下面以应用于双归属组网为例对本实施例进行描述：
参见图4，图4为本公开提供的实施例的应用组网示意图。在图4中，客户端为CE5_1。CE5_1所属的VLAN记为VLAN5_1。CE5_1归属于VTEP5_1和VTEP5_2。VTEP5_1和VTEP5_2同属多归属组5_1，多归属组5_1的组地址记为group地址5_1。VTEP5_1和VTEP5_2之间建立了VXLAN隧道。
VTEP5_1、VTEP5_2均和远端VTEP5_3建立BGP邻居。其中，VTEP5_1、VTEP5_2都 是使用各自的Loopback地址（loopback地址51和Loopback地址52）和远端VTEP5_3建立BGP邻居。
在图4中，尽管CE5_1归属于VTEP5_1和VTEP5_2，但是，VTEP5_1和VTEP5_2中只有一个会基于CE5_1发送的报文学习CE5_1的路由。CE5_1可以通过HASH方式将报文发到VTEP5_1或VTEP5_2。
以VTEP5_1基于CE5_1发送的报文学习CE5_1的路由为例，为便于描述，这里将VTEP5_1学习的CE5_1的路由记为路由5_1。
VTEP5_1生成路由5_1对应的路由表项（记为路由表项5_1）并记录至路由表。路由表项5_1包含目的地址、VNI、出端口。其中，目的地址可为CE5_1的MAC地址或者IP地址，VNI对应于CE5_1所属的VLAN（即VLAN5_1）并记为VNI5_1，出端口为VTEP5_1上接收到CE5_1发送的报文的端口（图4以Port5_1为例）。表7示出了路由表项5_1：
Figure PCTCN2017110883-appb-000007
表7
VTEP5_1发送同步消息给同一归属组中的VTEP5_2，同步消息中携带路由5_1和路由5_1的下一跳，路由5_1的下一跳为group地址5_1。
VTEP5_1通告路由5_1给远端VTEP5_3。这里，通告的路由5_1的下一跳为group地址5_1。
当VTEP5_2接收到VTEP5_1发送的同步消息后，从同步消息学习路由5_1，生成路由5_1对应的路由表项（记为路由表项5_2）并记录至路由表。路由表项5_2包含目的地址、VNI、出端口。其中，目的地址可为CE5_1的MAC地址或者IP地址，VNI为同步消息携带的VLAN标识即VLAN5_1对应的VNI（即上述的VNI5_1），出端口为：VTEP5_2上与同步消息携带的VLAN标识即VLAN5_1对应的端口（图4以Port5_2为例）。表8示出了路由表项5_2：
Figure PCTCN2017110883-appb-000008
表8
VTEP5_2通告路由5_1给远端VTEP5_3。这里，通告的路由5_1的下一跳为group地址5_1。
当VTEP5_3接收到VTEP5_1和VTEP5_2分别通告的路由5_1时，发现VTEP5_1和VTEP5_2通告的路由5_1的下一跳均为多归属组5_1的group地址group5_1，则确定路由5_1的下一跳为多归属组5_1(包含VTEP5_1和VTEP5_2)的group地址group5_1。
当VTEP5_3接收到发往CE5_1的报文时，VTEP5_3会按照负载分担方式选择将报文发送至多归属组5_1的VTEP5_1还是VTEP5_2。
当VTEP5_3将报文发送至VTEP5_1时，若VTEP5_1与CE5_1之间的链路正常，则VTEP5_1确定报文的VXLAN封装中的VXLAN标识VNI，基于VNI和报文的目的地址在路由表中查找对应的路由表项即上述的路由表项5_1，通过查找到的路由表项5_1的出端口即Port5_1发送VXLAN解封装后的报文至CE5_1。
当VTEP5_3将报文发送至VTEP5_2时，若VTEP5_2与CE5_1之间的链路正常，则VTEP5_2确定报文的VXLAN封装中的VXLAN标识VNI，基于VNI和报文的目的地址在路由表中查找对应的路由表项即上述的路由表项5_2，通过查找到的路由表项5_2的出端口即Port5_2发送VXLAN解封装后的报文至CE5_1。
当VTEP5_1和VTEP5_2中一个VTEP与CE5_1之间的链路故障时，以VTEP5_1与CE5_1之间的链路故障为例，则，VTEP5_1先在在路由表中找到对应的路由表项即上述路由表项5_1，将该找到的路由表项5_1的出端口从Port5_1修改为VTEP5_2的Loopback地址52。
之后，当VTEP5_3将报文发送至VTEP5_1时，VTEP5_1基于报文的VXLAN封装中的VNI和报文的目的地址在路由表中确定对应的路由表项即上述的路由表项5_2，发现此时路由表项5_2的出端口为VTEP5_2的Loopback地址52，则修改报文的VXLAN封装中的IP头并通过VTEP5_1与VTEP5_2之间已建立的VXLAN隧道发送。其中，修改IP头包括：将IP头中的源IP地址修改为VTEP5_1的Loopback地址51，将IP头中的目的IP地址修改为路由表项5_2的出端口即VTEP5_2的Loopback地址52。
当VTEP5_2通过与VTEP5_1之间已建立的VXLAN隧道收到报文时，确定报文的VXLAN封装携带的VNI，根据确定出的VNI和解封装后的报文的目的地址在路由表中确定路由表项即上述的路由表项5_2，通过路由表项5_2的出端口即Port5_2发送报文至CE5_1。
以上以多归属组网为例对本实施例进行了描述。
下面再以应用于多Board组网为例对本实施例进行描述：
参见图5，图5为本公开提供的另一实施例的应用组网示意图。在图5中，外部网络归属于Board网关6_1和Board网关6_2。Board网关6_1和Board网关6_2同属多归属组6_1，多归属组6_1的组地址记为group地址6_1。Board网关6_1和Board网关6_2之间建立了VXLAN隧道。
Board网关6_1、Board网关6_2均和远端Board网关6_3建立BGP邻居。其中，Board网关6_1、Board网关6_2都是使用各自的Loopback地址（Loopback地址61和Loopback地址62）和远端Board网关6_3建立BGP邻居。
在图5中，Board网关6_1和Board网关6_2均配置了默认路由。为便于描述，将Board网关6_1配置的默认路由表项记为默认路由表项6_0，将Board网关6_2配置的默认路由表项记为默认路由表项0_6。
默认路由表项6_0用于Board网关6_1在与外部网络之间的链路故障时，通过与Board网关6_2之间的VXLAN隧道将报文发送给Board网关6_2。默认路由表项6_0包含目的地址、VNI、出端口。其中，目的地址可为外部网络中客户端的MAC地址或者IP地址，VNI对应于外部网络所属VLAN并记为VNI6_1，出端口为Board网关6_2的Loopback地址62。表9示出了默认路由表项6_0：
Figure PCTCN2017110883-appb-000009
表9
默认路由表项0_6用于Board网关6_2在与外部网络之间的链路故障时，通过与Board网关6_1之间的VXLAN隧道将报文发送给Board网关6_1。默认路由表项0_6包含目的地址、VNI、出端口。其中，目的地址可为外部网络中客户端的MAC地址或者IP地址，VNI对应于外部网络所属VLAN并记为VNI6_1，出端口为Board网关6_1的Loopback地址61。表10示出了默认路由表项0_6：
Figure PCTCN2017110883-appb-000010
表10
在图5中，Board网关6_1、Board网关6_2和外部网络建立路由协议，因此，Board网关8_1、Board网关8_2上都会直接学习到外部网络的路由，没有必要像多归属组网中将学习的路由通过同步消息发送给同一多归属组中其他Board网关。
Board网关6_1在学习到外部网络的路由时生成对应的路由表项(记为路由表项6_1)并记录至路由表。路由表项6_1包含目的地址、VNI、出端口。其中，目的地址可为外部网络中客户端的MAC地址或者IP地址，VNI对应于外部网络所属VLAN并记为VNI6_1，出端口为Board网关6_1接入外部网络的路由接口(图5以Port6_1为例)。表11示出了路由表项6_1：
Figure PCTCN2017110883-appb-000011
表11
Board网关6_1通告学习的外部网络的路由给远端Board网关6_3。这里，通告的路由的下一跳为group地址6_1。
Board网关6_2在学习到外部网络的路由时生成对应的路由表项(记为路由表项6_2)并记录至路由表。路由表项6_2包含目的地址、VNI、出端口。其中，目的地址可为外部网络中客户端的MAC地址或者IP地址，VNI对应于外部网络所属VLAN并记为VNI6_1，出端口为Board网关6_2接入外部网络的路由接口(图5以Port6_2为例)。表12示出了路由表项6_2：
Figure PCTCN2017110883-appb-000012
表12
Board网关6_2通告学习的外部网络的路由给远端Board网关6_3。这里，通告的路由的下一跳为group地址6_1。
当Board网关6_3接收到Board网关6_1和Board网关6_2分别通告的外部网络的路由时，发现Board网关6_1和Board网关6_2通告的路由的下一跳均为多归属组6_1的组地址group地址6_1，则确定该路由的下一跳为多归属组6_1的组地址group地址6_1。
当Board网关6_3接收到发往外部网络的报文时，Board网关6_3会按照负载分担方 式选择将报文发送至多归属组6_1的Board网关6_1还是Board网关6_2。
当Board网关6_3将报文发送至Board网关6_1时，若Board网关6_1与外部网络之间的链路正常，则Board网关6_1确定报文的VXLAN封装中的VNI，基于VNI和报文的目的地址在路由表中查找对应的路由表项即上述的路由表项6_1，通过查找到的路由表项6_1的出端口即Port6_1发送VXLAN解封装后的报文至外部网络。
当Board网关6_3将报文发送至Board网关6_2时，若Board网关6_2与外部网络之间的链路正常，则Board网关6_2确定报文的VXLAN封装中的VNI，基于VNI和报文的目的地址在路由表中查找对应的路由表项即上述的路由表项6_2，通过查找到的路由表项6_2的出端口即Port6_2发送VXLAN解封装后的报文至外部网络。
当Board网关6_1和Board网关6_2中一个Board网关与外部网络之间的链路故障时，以Board网关6_1与外部网络之间的链路故障为例，则Board网关6_1在路由表中找到上述路由表项6_1并删除。
之后，当Board网关6_3将报文发送至Board网关6_1时，Board网关6_1基于报文的VXLAN封装中的VNI和报文的目的地址在路由表中确定对应的路由表项即上述的默认路由表项6_0，根据默认路由表项6_0修改报文的VXLAN封装中的IP头，并通过Board网关6_1与Board网关6_2之间已建立的VXLAN隧道发送。其中，修改报文的VXLAN封装中的IP头是指：将IP头中的源IP地址修改为VTEP6_1的Loopback地址61，将IP头中的目的IP地址修改为默认路由表项6_0的出端口即VTEP6_2的Loopback地址62。
当Board网关6_2通过与Board网关6_1之间已建立的VXLAN隧道收到报文时，确定报文的VXLAN封装中的VNI，根据确定的VNI和报文的目的地址在路由表中确定路由表项即上述的路由表项6_2，通过路由表项6_2的出端口即Port6_2发送报文至外部网络。
以上对本公开提供的方法进行了描述。下面对本公开提供的装置进行描述：
参见图6，图6为本公开提供的装置结构图。该装置应用于多归属组中的任一VTEP，所述多归属组包括指定角色归属的所有的VTEP。如图6所示，该装置包括：
获取单元601，用于获取所述指定角色的路由，其中，所述多归属组包括所述指定角色归属的不同VTEP；
通告单元602，用于将所述路由通告给远端VTEP，以使远端VTEP根据所述路由将发往所述指定角色的报文分发至所述多归属组中的各个VTEP，其中，所述通告给远端VTEP的所述路由的下一跳为所述多归属组的组地址；
处理单元603，用于从远端VTEP接收到发往所述指定角色的报文，当所述VTEP与所述指定角色之间的链路故障时，将所述报文发送至所述多归属组中的其他VTEP，以使其他VTEP将所述报文发送至所述指定角色。
在一个示例中，所述多归属组中的VTEP之间串行建立端至端链路Peer link；
所述处理单元603将报文发送至多归属组中的其他VTEP包括：
确定报文携带的VXLAN标识VNI；
对报文进行VXLAN解封装得到解封装后的报文，将所述VNI对应的VLAN标识携带在所述解封装后的报文中并通过已建立的Peer link发送。
在一个示例中，所述指定角色为客户端；所述处理单元703在所述VTEP与指定角色之间的链路故障时，进一步在路由表中确定满足以下条件的路由表项：该路由表项的出端口为所述VTEP上连接故障链路的端口，将确定的路由表项中的出端口修改为所述VTEP上的Peer link端口；
所述处理单元603将VNI对应的VLAN标识携带在解封装后的报文中并通过已建立的Peer link发送包括：基于VNI和报文的目的地址在路由表中确定路由表项，将VNI对应的VLAN标识携带在解封装后的报文中并通过确定出的路由表项中的出端口发送，所述出端口为Peer link端口。
在一个示例中，所述VTEP为Board网关；所述处理单元603在所述VTEP与所述指定角色之间的链路故障时，进一步在路由表中找到满足条件的路由表项并删除，所述条件为：所述路由表项的出端口为所述VTEP上连接故障链路的端口；
所述处理单元603将VNI对应的VLAN标识携带在解封装后的报文中并通过已建立的Peer link发送包括：基于VNI和报文的目的地址在路由表中确定设定的第一默认路由，将VNI对应的VLAN标识携带在解封装后的报文中并通过第一默认路由的出端口发送，第一默认路由的出端口为Peer link端口。
在一个示例中，所述多归属组中的VTEP之间串行建立VXLAN隧道；所述处理单元603将报文发送至多归属组中其他VTEP包括：
修改报文的外层IP头并通过已建立的VXLAN隧道发送，所述修改报文的外层IP头包括：修改外层IP头中的源IP地址为所述VTEP的环回接口Loopback地址，目的IP地址为所述多归属组中与所述VTEP建立了VXLAN隧道的其他VTEP的Loopback地址。
在一个示例中，所述处理单元603在所述VTEP与所述指定角色之间的链路故障时，进一步在路由表中确定满足以下条件的路由表项：该路由表项的出端口为所述VTEP上连接故障链路的端口，将确定的路由表项中的出端口修改为所述多归属组中与所述VTEP已建立了VXLAN隧道的其他VTEP的Loopback地址；
所述处理单元603修改报文的外层IP头并通过已建立的VXLAN隧道发送包括：基于报文携带的VNI和报文的目的地址在路由表中确定路由表项，当确定出的路由表项的出端口为Loopback地址时，修改报文的外层IP头的源IP地址为所述VTEP的Loopback地址，目的IP地址为所述出端口，并向所述出端口发送。
在一个示例中，所述VTEP为Board网关；所述处理单元603进一步当所述VTEP与所述指定角色之间的链路故障时，在路由表中确定满足条件的路由表项并删除，所述条件为：出端口为所述VTEP上连接故障链路的端口；
所述处理单元603修改报文的外层IP头并通过已建立的VXLAN隧道发送包括：基于报文携带的VNI和报文的目的地址在路由表中确定设定的第二默认路由，修改报文的外层IP头的源IP地址为所述VTEP的Loopback地址，目的IP地址为第二默认路由的出端口，并向所述第二默认路由的出端口发送；第二默认路由的出端口为所述多归属组中与所述VTEP建立了VXLAN隧道的其他VTEP的Loopback地址。
参见图7，图7为本公开提供的VTEP的结构示意图。该设备包含：处理器701、存储有对应于报文转发控制逻辑的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。如图7所示，该结构还可以包括一个电源组件704被配置为执行设备的电源管理，一个有线或无线网络接口705被配置为将设备连接到网络，和一个输入输出(I/O)接口706。
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。
机器可读存储介质702，用于存放报文转发控制逻辑对应的机器可执行指令，所述机器可执行指令包括获取单元601运行的程序指令、通告单元602运行的程序指令、处理单元603运行的程序指令。
例如，当处理器701读取并执行机器可读存储介质702中的机器可执行指令时，该处理器701被促使执行以下操作：
获取指定角色的路由；
将所述路由通告给远端VTEP，以使远端VTEP根据所述路由将发往所述指定角色的报文分发至多归属组中的各个VTEP，其中，所述多归属组包括所述指定角色归属的不同VTEP，所述通告给远端VTEP的所述路由的下一跳为所述多归属组的组地址；
从远端VTEP接收到发往所述指定角色的报文，当所述VTEP与所述指定角色之间的链路故障时，将所述报文发送至所述多归属组中的其他VTEP，以使其他VTEP将所述报文发送至所述指定角色。
在一个示例中，所述多归属组中的VTEP之间串行建立端至端链路Peer link，在将所述报文发送至所述多归属组中的其他VTEP时，所述处理器701被所述机器可执行指令促使：
确定所述报文携带的VXLAN标识VNI；
对所述报文进行VXLAN解封装，得到解封后的报文；
将所述VNI对应的VLAN标识携带在解封装后的报文中并通过所述已建立的Peer link发送。
在一个示例中，当所述VTEP与所述指定角色之间的链路故障时，所述处理器701被所述机器可执行指令促使：
在路由表中确定满足以下条件的路由表项：该路由表项的出端口为所述VTEP上连接故障链路的端口，
将确定的路由表项中的出端口修改为所述VTEP上的Peer link端口；
当将所述VNI对应的VLAN标识携带在解封装后的报文中并通过已建立的Peer link发送时，所述处理器被所述机器可执行指令促使：
基于VNI和所述报文的目的地址在路由表中确定路由表项，
将所述VNI对应的VLAN标识携带在解封装后的报文中并通过确定的路由表项中的出端口发送，所述出端口为Peer link端口。
在一个示例中，所述VTEP为边界Board网关；当所述VTEP与所述指定角色之间的链路故障时，所述处理器701被所述机器可执行指令促使：
在路由表中找到满足条件的路由表项并删除，所述条件为：该路由表项的出端口为所述VTEP上连接故障链路的端口；
当将所述VNI对应的VLAN标识携带在解封装后的报文中并通过已建立的Peer link发送 时，所述处理器被所述机器可执行指令促使：
基于VNI和报文的目的地址在路由表中确定设定的第一默认路由，将所述VNI对应的VLAN标识携带在解封装后的报文中并通过第一默认路由的出端口发送，第一默认路由的出端口为Peer link端口。
在一个示例中，所述多归属组中的VTEP之间串行建立VXLAN隧道；
在将所述报文发送至所述多归属组中其他VTEP时，所述处理器701被所述机器可执行指令促使：
修改所述报文的外层互联网协议IP头并通过已建立的VXLAN隧道发送，所述修改所述报文的外层IP头包括：修改外层IP头中的源IP地址为所述VTEP的环回接口Loopback地址，目的IP地址为所述多归属组中与所述VTEP建立了VXLAN隧道的其他VTEP的Loopback地址。
在一个示例中，当所述VTEP与所述指定角色之间的链路故障时，所述处理器701被所述机器可执行指令促使：
在路由表中确定满足以下条件的路由表项：该路由表项的出端口为所述VTEP上连接故障链路的端口，
将确定的路由表项中的出端口修改为所述多归属组中与所述VTEP已建立了VXLAN隧道的其他VTEP的Loopback地址；
在修改所述报文的外层IP头并通过已建立的VXLAN隧道发送时，所述处理器被所述机器可执行指令促使：
基于所述报文携带的所述VNI和所述报文的目的地址在本地路由表中匹配路由表项，
修改所述报文的外层IP头的源IP地址为所述VTEP的Loopback地址，目的IP地址为所确定的路由表项中的出端口，并向所述出端口发送，其中，所确定的路由表项中的出端口为Loopback地址。
在一个示例中，所述VTEP为边界Board网关；当所述VTEP与所述指定角色之间的链路故障时，所述处理器701被所述机器可执行指令促使：
在路由表中确定满足条件的路由表项并删除，所述条件为：该路由表项的出端口为所述VTEP上连接故障链路的端口；
当修改所述报文的外层IP头并通过已建立的VXLAN隧道发送时，所述处理器被所述机器可执行指令促使：
基于所述报文携带的所述VNI和所述报文的目的地址在路由表中确定设定的第二默认路 由，修改所述报文的外层IP头的源IP地址为所述VTEP的Loopback地址，目的IP地址为第二默认路由的出端口，并向所述第二默认路由的出端口发送；第二默认路由的出端口为所述多归属组中与所述VTEP建立了VXLAN隧道的其他VTEP的Loopback地址。
在一个示例中，在获取指定角色的路由时，所述处理器701被所述机器可执行指令促使：
通过所述指定角色发送的报文学习所述路由；
当获取所述路由后，所述处理器701被所述机器可执行指令促使：
生成所述路由对应的路由表项并记录至路由表，发送同步消息给所述多归属组中其他VTEP，同步消息携带所述路由、以及所述路由所属VLAN的VLAN标识；所述路由表项包含目的地址、所述报文携带的VXLAN标识VNI和出端口，目的地址为指定角色的地址，VNI对应于指定角色所属VLAN，出端口为所述VTEP上接收到指定角色发送的报文的端口；
或者，
在获取所述指定角色的路由时，所述处理器701被所述机器可执行指令促使：
接收所述多归属组中其他VTEP发送的同步消息，从接收的同步消息学习该同步消息携带的所述路由；
当从接收的同步消息学习所述路由后，所述处理器701被所述机器可执行指令促使：
生成所述路由对应的路由表项并记录至路由表，生成的路由表项包含目的地址、VNI和出端口，其中，目的地址为指定角色的地址，VNI对应于与同步消息携带的VLAN标识，出端口为所述VTEP上与同步消息携带的VLAN标识对应的端口。
对于装置实施例而言，由于其基本对应于方法实施例，所以相关之处参见方法实施例的部分说明即可。以上所描述的装置实施例仅仅是示意性的，其中所述作为分离部件说明的单元可以是或者也可以不是物理上分开的，作为单元显示的部件可以是或者也可以不是物理单元，即可以位于一个地方，或者也可以分布到多个网络单元上。可以根据实际的需要选择其中的部分或者全部模块来实现本实施例方案的目的。本领域普通技术人员在不付出创造性劳动的情况下，即可以理解并实施。
需要说明的是，在本文中，诸如第一和第二等之类的关系术语仅仅用来将一个实体或者操作与另一个实体或操作区分开来，而不一定要求或者暗示这些实体或操作之间存在任何这种实际的关系或者顺序。术语“包括”、“包含”或者其任何其他变体意在涵盖非排他性的包含，从而使得包括一系列要素的过程、方法、物品或者设备不仅包括那些要素，而且还包括没有明确列出的其他要素，或者是还包括为这种过程、方法、物品或者设备所固有的要素。 在没有更多限制的情况下，由语句“包括一个……”限定的要素，并不排除在包括所述要素的过程、方法、物品或者设备中还存在另外的相同要素。
以上对本公开实施例所提供的方法和装置进行了详细介绍，本文中应用了具体个例对本公开的原理及实施方式进行了阐述，以上实施例的说明只是用于帮助理解本公开的方法及其核心思想；同时，对于本领域的一般技术人员，依据本公开的思想，在具体实施方式及应用范围上均会有改变之处，综上所述，本说明书内容不应理解为对本公开的限制。

Claims (14)

  1. 一种报文转发方法，包括：
    多归属组中的任一VXLAN隧道端点VTEP获取指定角色的路由，其中，所述多归属组包括所述指定角色归属的所有VTEP；
    所述VTEP将所述路由通告给远端VTEP，以使所述远端VTEP根据所述路由将发往所述指定角色的报文分发至所述多归属组中的各个VTEP，所述通告给远端VTEP的所述路由的下一跳为所述多归属组的组地址；
    所述VTEP从所述远端VTEP接收到发往所述指定角色的报文；
    当所述VTEP与所述指定角色之间的链路故障时，所述VTEP将所述报文发送至所述多归属组中的其他VTEP，以使其他VTEP将所述报文发送至所述指定角色。
  2. 根据权利要求1所述的方法，其中，所述多归属组中的VTEP之间串行建立端至端链路Peer link；
    将所述报文发送至所述多归属组中的其他VTEP包括：
    所述VTEP确定所述报文携带的VXLAN标识VNI，并对所述报文进行VXLAN解封装，得到解封后的报文；
    所述VTEP将所述VNI对应的VLAN标识携带在所述解封装后的报文中并通过已建立的Peer link发送。
  3. 根据权利要求2所述的方法，其中，所述方法还包括：
    当所述VTEP与所述指定角色之间的链路故障时，
    所述VTEP在路由表中确定满足以下条件的路由表项：该路由表项的出端口为所述VTEP上连接故障链路的端口；
    所述VTEP将确定的路由表项中的出端口修改为所述VTEP上的Peer link端口；
    将所述VNI对应的VLAN标识携带在解封装后的报文中并通过已建立的Peer link发送包括：
    所述VTEP基于VNI和所述报文的目的地址在路由表中确定路由表项，
    所述VTEP将所述VNI对应的VLAN标识携带在解封装后的报文中并通过确定的路由表项中的出端口发送，所述出端口为Peer link端口。
  4. 根据权利要求2所述的方法，其中，所述VTEP为边界Board网关；
    所述方法还包括：
    当所述VTEP与所述指定角色之间的链路故障时，在路由表确定满足条件的路由表项并删除，所述条件为：该路由表项的出端口为所述VTEP上连接故障链路的端口；
    将所述VNI对应的VLAN标识携带在解封装后的报文中并通过已建立的Peer link发送包括：
    所述VTEP基于VNI和报文的目的地址在路由表中确定设定的第一默认路由，将所述VNI对应的VLAN标识携带在解封装后的报文中并通过第一默认路由的出端口发送，第一默认路由的出端口为Peer link端口。
  5. 根据权利要求1所述的方法，其中，所述多归属组中的VTEP之间串行建立VXLAN隧道；
    将所述报文发送至所述多归属组中其他VTEP包括：
    所述VTEP修改所述报文的外层互联网协议IP头并通过已建立的VXLAN隧道发送，所述修改所述报文的外层IP头包括：修改外层IP头中的源IP地址为所述VTEP的环回接口Loopback地址，目的IP地址为所述多归属组中与所述VTEP建立了VXLAN隧道的其他VTEP的Loopback地址。
  6. 根据权利要求5所述的方法，其中，所述方法还包括：
    当所述VTEP与所述指定角色之间的链路故障时，
    所述VTEP在路由表中确定满足以下条件的路由表项：该路由表项的出端口为所述VTEP上连接故障链路的端口；
    所述VTEP将确定的路由表项中的出端口修改为所述多归属组中与所述VTEP已建立了VXLAN隧道的其他VTEP的Loopback地址；
    所述修改所述报文的外层IP头并通过已建立的VXLAN隧道发送包括：
    所述VTEP基于所述报文携带的所述VNI和所述报文的目的地址在路由表中确定路由表项；
    所述VTEP修改所述报文的外层IP头的源IP地址为所述VTEP的Loopback地址，目的IP地址为所确定的路由表项中的出端口，并向所述出端口发送，其中，所确定的路由表项中的出端口为Loopback地址。
  7. 根据权利要求5所述的方法，其中，所述VTEP为边界Board网关；
    所述方法还包括：
    当所述VTEP与所述指定角色之间的链路故障时，在路由表中确定满足条件的路由表项并删除，所述条件为：该路由表项的出端口为所述VTEP上连接故障链路的端口；
    所述修改所述报文的外层IP头并通过已建立的VXLAN隧道发送包括：
    所述VTEP基于所述报文携带的所述VNI和所述报文的目的地址在路由表中确定设定的第二默认路由，修改所述报文的外层IP头的源IP地址为所述VTEP的Loopback地址，目的 IP地址为第二默认路由的出端口，并向所述第二默认路由的出端口发送；第二默认路由的出端口为所述多归属组中与所述VTEP建立了VXLAN隧道的其他VTEP的Loopback地址。
  8. 一种VXLAN隧道端点VTEP，包括：
    处理器；以及
    存储有机器可执行指令的机器可读存储介质，其中，通过读取并执行所述机器可执行指令，所述处理器被使得：
    获取指定角色的路由，
    将所述路由通告给远端VTEP，以使所述远端VTEP根据所述路由将发往所述指定角色的报文分发至多归属组中的各个VTEP，其中，所述多归属组包括所述指定角色归属的不同VTEP，所述通告给远端VTEP的所述路由的下一跳为所述多归属组的组地址；
    从所述远端VTEP接收到发往所述指定角色的报文，当所述VTEP与所述指定角色之间的链路故障时，将所述报文发送至所述多归属组中的其他VTEP，以使其他VTEP将所述报文发送至所述指定角色。
  9. 根据权利要求8所述的VTEP，其中，所述多归属组中的VTEP之间串行建立端至端链路Peer link，在将所述报文发送至所述多归属组中的其他VTEP时，所述处理器被所述机器可执行指令促使：
    确定所述报文携带的VXLAN标识VNI；
    对所述报文进行VXLAN解封装，得到解封装后的报文；
    将所述VNI对应的VLAN标识携带在所述解封装后的报文中并通过已建立的Peer link发送。
  10. 根据权利要求9所述的VTEP，其中，
    当所述VTEP与所述指定角色之间的链路故障时，所述处理器被所述机器可执行指令促使：
    在路由表中确定满足以下条件的路由表项：该路由表项的出端口为所述VTEP上连接故障链路的端口；
    将确定的路由表项中的出端口修改为所述VTEP上的Peer link端口；
    当将所述VNI对应的VLAN标识携带在解封装后的报文中并通过已建立的Peer link发送时，所述处理器被所述机器可执行指令促使：
    基于VNI和所述报文的目的地址在路由表中确定路由表项；
    将所述VNI对应的VLAN标识携带在解封装后的报文中并通过确定的路由表项中的出端口发送，所述出端口为Peer link端口。
  11. 根据权利要求9所述的VTEP，其中，所述VTEP为边界Board网关；
    当所述VTEP与所述指定角色之间的链路故障时，所述处理器被所述机器可执行指令促使：
    在路由表中找到满足条件的路由表项并删除，所述条件为：该路由表项的出端口为所述VTEP上连接故障链路的端口；
    当将所述VNI对应的VLAN标识携带在解封装后的报文中并通过已建立的Peer link发送时，所述处理器被所述机器可执行指令促使：
    基于VNI和报文的目的地址在路由表中确定设定的第一默认路由，将所述VNI对应的VLAN标识携带在解封装后的报文中并通过第一默认路由的出端口发送，第一默认路由的出端口为Peer link端口。
  12. 根据权利要求8所述的VTEP，其中，所述多归属组中的VTEP之间串行建立VXLAN隧道；
    在将所述报文发送至所述多归属组中其他VTEP时，所述处理器被所述机器可执行指令促使：
    修改所述报文的外层互联网协议IP头并通过已建立的VXLAN隧道发送，所述修改所述报文的外层IP头包括：修改外层IP头中的源IP地址为所述VTEP的环回接口Loopback地址，目的IP地址为所述多归属组中与所述VTEP建立了VXLAN隧道的其他VTEP的Loopback地址。
  13. 根据权利要求12所述的VTEP，其中，
    当所述VTEP与所述指定角色之间的链路故障时，所述处理器被所述机器可执行指令促使：
    在路由表中确定满足以下条件的路由表项：该路由表项的出端口为所述VTEP上连接故障链路的端口；
    将确定的路由表项中的出端口修改为所述多归属组中与所述VTEP已建立了VXLAN隧道的其他VTEP的Loopback地址；
    在修改所述报文的外层IP头并通过已建立的VXLAN隧道发送时，所述处理器被所述机器可执行指令促使：
    基于所述报文携带的所述VNI和所述报文的目的地址在本地路由表中确定路由表项；
    修改所述报文的外层IP头的源IP地址为所述VTEP的Loopback地址，目的IP地址为所确定的路由表项中的出端口，并向所述出端口发送，其中，所确定的路由表项中的出端口为Loopback地址。
  14. 根据权利要求12所述的VTEP，其中，所述VTEP为边界Board网关；
    当所述VTEP与所述指定角色之间的链路故障时，所述处理器被所述机器可执行指令促使：
    在路由表中确定满足条件的路由表项并删除，所述条件为：该路由表项的出端口为所述VTEP上连接故障链路的端口；
    当修改所述报文的外层IP头并通过已建立的VXLAN隧道发送时，所述处理器被所述机器可执行指令促使：
    基于所述报文携带的所述VNI和所述报文的目的地址在路由表中确定设定的第二默认路由，修改所述报文的外层IP头的源IP地址为所述VTEP的Loopback地址，目的IP地址为第二默认路由的出端口，并向所述第二默认路由的出端口发送；第二默认路由的出端口为所述多归属组中与所述VTEP建立了VXLAN隧道的其他VTEP的Loopback地址。
PCT/CN2017/110883 2016-11-17 2017-11-14 报文转发 WO2018090907A1 (zh)

Priority Applications (3)

Application Number Priority Date Filing Date Title
EP17872007.4A EP3528441B1 (en) 2016-11-17 2017-11-14 Message forwarding
US16/461,563 US11159341B2 (en) 2016-11-17 2017-11-14 Forwarding packet
JP2019526244A JP6682045B2 (ja) 2016-11-17 2017-11-14 パケット転送

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201611033162.0 2016-11-17
CN201611033162.0A CN108075969B (zh) 2016-11-17 2016-11-17 报文转发方法和装置

Publications (1)

Publication Number Publication Date
WO2018090907A1 true WO2018090907A1 (zh) 2018-05-24

Family

ID=62146097

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2017/110883 WO2018090907A1 (zh) 2016-11-17 2017-11-14 报文转发

Country Status (5)

Country Link
US (1) US11159341B2 (zh)
EP (1) EP3528441B1 (zh)
JP (1) JP6682045B2 (zh)
CN (1) CN108075969B (zh)
WO (1) WO2018090907A1 (zh)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2021530158A (ja) * 2018-06-30 2021-11-04 華為技術有限公司Ｈｕａｗｅｉ Ｔｅｃｈｎｏｌｏｇｉｅｓ Ｃｏ．，Ｌｔｄ． Ｂｇｐメッセージ送信方法、ｂｇｐメッセージ受信方法、及びデバイス

Families Citing this family (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110636583B (zh) * 2018-06-21 2021-08-13 华为技术有限公司 路径变更方法及装置
TWI821373B (zh) 2018-08-23 2023-11-11 美商阿爾克斯股份有限公司 網路運算環境中的第一跳轉閘道的冗餘機制系統
CN113382370A (zh) 2018-10-08 2021-09-10 华为技术有限公司 一种通信方法及装置
US10805210B1 (en) * 2019-09-20 2020-10-13 Juniper Networks, Inc. GRE tunneling with reduced packet encryption at intermediate routers using loose source routing
CN111277500A (zh) * 2020-01-15 2020-06-12 山东汇贸电子口岸有限公司 一种数据中心evpn组网多出口设计方法及系统
CN113472645A (zh) * 2020-03-30 2021-10-01 华为技术有限公司 发送路由、处理路由的方法、装置和设备
CN114124617A (zh) * 2020-08-31 2022-03-01 华为技术有限公司 一种通信方法、装置、设备及系统
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

Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104270298A (zh) * 2014-09-30 2015-01-07 杭州华三通信技术有限公司 一种vxlan网络中的报文转发方法及装置
CN104378296A (zh) * 2013-08-15 2015-02-25 杭州华三通信技术有限公司 一种报文转发方法及设备
CN104378297A (zh) * 2013-08-15 2015-02-25 杭州华三通信技术有限公司 一种报文转发方法及设备
US20150195178A1 (en) * 2014-01-09 2015-07-09 Ciena Corporation Method for resource optimized network virtualization overlay transport in virtualized data center environments
CN105553848A (zh) * 2014-10-22 2016-05-04 瞻博网络公司 针对数据中心互联的协议独立多播稀疏模式（pim-sm）的支持
CN105577417A (zh) * 2014-11-06 2016-05-11 杭州华三通信技术有限公司 基于vxlan网络的报文转发方法及装置
CN106789635A (zh) * 2016-03-25 2017-05-31 新华三技术有限公司 一种报文转发方法及装置

Family Cites Families (12)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
KR101344015B1 (ko) * 2007-02-09 2013-12-23 삼성전자주식회사 ＰＩＭ―ＳＳＭ을 이용하는 네트워크에서 Ｊｏｉｎ 메시지부하 조절 시스템 및 방법
JP6036459B2 (ja) 2013-03-25 2016-11-30 富士通株式会社 情報処理装置、情報処理システム、ゲートウェイ管理方法及びゲートウェイ管理プログラム
CN103166858B (zh) * 2013-03-26 2016-06-08 杭州华三通信技术有限公司 一种报文传输方法和设备
US9374323B2 (en) * 2013-07-08 2016-06-21 Futurewei Technologies, Inc. Communication between endpoints in different VXLAN networks
US9374294B1 (en) 2013-11-05 2016-06-21 Cisco Technology, Inc. On-demand learning in overlay networks
EP2966815A4 (en) * 2014-05-29 2016-04-13 Huawei Tech Co Ltd PACKET PROCESSING AND VXLAN GATEWAY
CN104468394B (zh) * 2014-12-04 2018-02-09 新华三技术有限公司 一种vxlan网络中报文转发方法及装置
CN105791072A (zh) * 2014-12-22 2016-07-20 华为数字技术（苏州）有限公司 以太虚拟网络的接入方法及装置
US9900250B2 (en) * 2015-03-26 2018-02-20 Cisco Technology, Inc. Scalable handling of BGP route information in VXLAN with EVPN control plane
US10666500B2 (en) * 2016-03-30 2020-05-26 Juniper Networks, Inc. Failure handling for active-standby redundancy in EVPN data center interconnect
CN106059932B (zh) * 2016-08-08 2020-12-11 新华三技术有限公司 一种路由表项生成方法及装置
US10425325B2 (en) * 2017-10-30 2019-09-24 Dell Products Lp Optimizing traffic paths to orphaned hosts in VXLAN networks using virtual link trunking-based multi-homing

Patent Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104378296A (zh) * 2013-08-15 2015-02-25 杭州华三通信技术有限公司 一种报文转发方法及设备
CN104378297A (zh) * 2013-08-15 2015-02-25 杭州华三通信技术有限公司 一种报文转发方法及设备
US20150195178A1 (en) * 2014-01-09 2015-07-09 Ciena Corporation Method for resource optimized network virtualization overlay transport in virtualized data center environments
CN104270298A (zh) * 2014-09-30 2015-01-07 杭州华三通信技术有限公司 一种vxlan网络中的报文转发方法及装置
CN105553848A (zh) * 2014-10-22 2016-05-04 瞻博网络公司 针对数据中心互联的协议独立多播稀疏模式（pim-sm）的支持
CN105577417A (zh) * 2014-11-06 2016-05-11 杭州华三通信技术有限公司 基于vxlan网络的报文转发方法及装置
CN106789635A (zh) * 2016-03-25 2017-05-31 新华三技术有限公司 一种报文转发方法及装置

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
See also references of EP3528441A4 *

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2021530158A (ja) * 2018-06-30 2021-11-04 華為技術有限公司Ｈｕａｗｅｉ Ｔｅｃｈｎｏｌｏｇｉｅｓ Ｃｏ．，Ｌｔｄ． Ｂｇｐメッセージ送信方法、ｂｇｐメッセージ受信方法、及びデバイス
JP7216120B2 (ja) 2018-06-30 2023-01-31 華為技術有限公司 Ｂｇｐメッセージ送信方法、ｂｇｐメッセージ受信方法、及びデバイス
US11973795B2 (en) 2018-06-30 2024-04-30 Huawei Technologies Co., Ltd. BGP message sending method, BGP message receiving method, and device

Also Published As

Publication number Publication date
JP2019536366A (ja) 2019-12-12
JP6682045B2 (ja) 2020-04-15
EP3528441A4 (en) 2019-11-06
EP3528441B1 (en) 2022-07-13
US11159341B2 (en) 2021-10-26
CN108075969B (zh) 2020-01-03
CN108075969A (zh) 2018-05-25
EP3528441A1 (en) 2019-08-21
US20200067731A1 (en) 2020-02-27

Similar Documents

Publication Publication Date Title
WO2018090907A1 (zh) 报文转发
US12106160B2 (en) First hop gateway redundancy in a network computing environment
US20200296025A1 (en) Route Processing Method and Apparatus, and Data Transmission Method and Apparatus
US9992154B2 (en) Layer 3 convergence for EVPN link failure
US10237163B2 (en) Static route advertisement
US10050877B2 (en) Packet forwarding method and apparatus
EP3065342B1 (en) Update of mac routes in evpn single-active topology
US10148566B2 (en) Split-horizon packet forwarding in a multi-home PBB-EVPN network
EP3151509B1 (en) Enhanced evpn mac route advertisement having mac (l2) level authentication, security and policy control
US9860150B2 (en) Fast convergence of EVPN networks for multi homing topologies
US8787149B1 (en) MAC address synchronization for multi-homing with multichassis link aggregation
US9019814B1 (en) Fast failover in multi-homed ethernet virtual private networks
US8953590B1 (en) Layer two virtual private network having control plane address learning supporting multi-homed customer networks
US9860169B1 (en) Neighbor resolution for remote EVPN hosts in IPV6 EVPN environment
WO2017114196A1 (zh) 一种报文处理方法、相关装置及nvo3网络系统
US8170033B1 (en) Virtual private local area network service (VPLS) flush mechanism for BGP-based VPLS networks
US9178816B1 (en) Control plane messaging in all-active multi-homed ethernet virtual private networks
US8125926B1 (en) Inter-autonomous system (AS) virtual private local area network service (VPLS)
US9300524B2 (en) Message forwarding between geographically dispersed network sites
CN109218178A (zh) 一种报文处理方法及网络设备
US10523466B1 (en) Aliasing in an active-active multi-homed PBB-EVPN network
CN107070789B (zh) 主动-主动pbb-evpn冗余的流量黑洞避免和快速融合
US9019973B1 (en) Static MAC address propagation in multipoint network services
US20150334057A1 (en) Packet forwarding
EP3301868B1 (en) Symmetric intersubnet traffic load balancing in multihomed networks

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: 17872007

Country of ref document: EP

Kind code of ref document: A1

ENP Entry into the national phase

Ref document number: 2019526244

Country of ref document: JP

Kind code of ref document: A

ENP Entry into the national phase

Ref document number: 2017872007

Country of ref document: EP

Effective date: 20190514

NENP Non-entry into the national phase

Ref country code: DE