WO2013139159A1 - Method for forwarding packet in network and provider edge device - Google Patents

Abstract

The present invention relates to the network transmission field. Disclosed are a method for forwarding a packet in a network and a provider edge (PE) device. The network comprises: a first PE and a second PE. The first PE and the second PE belong to a redundancy group (RG), the RG corresponds to a virtual PE, the virtual PE has an IP address, the virtual PE is a member PE in a first VPLS instance, and a first CE is a local CE of the virtual PE in the first VPLS instance. The method comprises: the first PE receiving an Ethernet frame from the first CE; and the first PE encapsulating the Ethernet frame according to the type of the Ethernet frame to obtain a packet after encapsulation, and sending the packet after encapsulation to a remote end PE of the virtual PE in the first VPLS instance. The first PE uses the IP address of the virtual PE corresponding to the RG that includes the first PE and the second PE for packet encapsulation, thereby ensuring that when the remote end PE receives the packet and learns a MAC address, a next-hop PE corresponding to the MAC is the virtual PE all the time. Therefore, no transition of the next-hop PE occurs, thereby ensuring the stability of a MAC table.

Description

 Method for forwarding messages in the network and operator edge device This application claims to be filed on March 22, 2012, the Chinese Patent Office, the application number is CN 201210079144.1, and the invention name is "Methods for Forwarding Messages in the Network and Operator Edges" The priority of the Chinese Patent Application, the entire disclosure of which is incorporated herein by reference. Technical field

 The present invention relates to the field of network transmission, and in particular, to a method for forwarding a message in a network and a carrier edge device.

BACKGROUND With the continuous development of Ethernet technology, Ethernet networks are becoming larger and larger. At present, the carrier-grade Ethernet circuit is widely used. The virtual private LAN service (English full name is Virtual Private LAN Service, VPLS in English) is in the public Internet protocol (English full name is Internet Protocol, English abbreviated as IP) / Multi-Protocol Label Switching (English for Multi-Protocol Label Switching, English abbreviated as MPLS) The point-to-multipoint Layer 2 virtual private network provided in the network (English name is Level-2 Virtual Private Network, English abbreviated as L2VPN) business. VPLS utilizes Ethernet technology and MPLS technology to connect multiple Ethernet LANs (English called Local Area Network, English abbreviated as LAN) distributed in different regions through the carrier's IP/MPLS network, so that they are connected to each other. The same as an Ethernet bridge. VPLS is based on media access control (English name is called Media Access Control, MAC for short) or MAC address plus virtual local area network (English name is called Virtual Local Area Network, English abbreviated as VLAN) to make forwarding decisions, user edge devices ( The English full name is Custom Edge, which is abbreviated as CE. The device can communicate with other CEs belonging to the same VPLS instance. From the perspective of the CE device, the VPLS service provided by the service provider network is an Ethernet bridge (or Ethernet switch). Currently, the existing implementation of VPLS requires the establishment of a fully interconnected pseudowire (in English, Pseudo Wire, PW in English) to simulate an Ethernet network. In addition, the existing VPLS solution only supports the multicast service in the ingress replication mode. Therefore, in the network with a large number of VPLS instances, the existing VPLS technology has serious scalability problems, such as maintaining a large number of PW states and a large number of ingress replications. Bandwidth wasted and intermediate system to intermediate system (English full name Intermediate) System- Intermediate System (English for short) IVP) VPLS implements VPLS reachability information advertisement by extending the ISIS type length value (abbreviated as TLV), that is, completes the automatic discovery and signaling function of the neighbor, avoiding the existence of the existing VPLS technology. The above problem.

 In the existing ISIS VPLS technology, if two uplinks of a CE are connected to the PE1 and the inter-chassis link aggregation group (English name is Multi-Chasis Link Aggegation Group, English abbreviated as MC-LAG) respectively. PE2, then, for the CE, when receiving the packet sent from the CE and performing MAC address learning, the next hop PE corresponding to the MAC hops between the first PE and the second PE, resulting in a MAC table. Unstable.

SUMMARY OF THE INVENTION The embodiments of the present invention provide a method for forwarding a message in a network and an operator edge device. The technical solution of the embodiment of the present invention includes:

 A method for forwarding a packet in a network, the network includes: a first carrier edge device PE, a second PE, the first PE and the second PE belong to a redundancy group RG, and the RG corresponds to a virtual PE, the virtual PE has an IP address, and the virtual PE is a member PE of the first virtual private local area network VPLS instance, and the first VPLS instance is an intermediate system to an intermediate system ISIS VPLS instance, the first user The edge device CE is a local CE of the virtual PE in the first VPLS instance, and the first CE is connected to the first PE and the second PE in an inter-frame link aggregation group MC-LAG manner. , the method includes:

 The first PE receives an Ethernet frame from the first CE;

 The first PE encapsulates the encapsulated packet according to the type of the Ethernet frame, and encapsulates the encapsulated packet into the virtual PE in the first VPLS instance. The remote PE sends, where the tunnel source address of the encapsulated packet is the IP address of the virtual PE.

 A method for forwarding a message in a network, the network includes: a third PE and a virtual PE, where a member PE of the RG corresponding to the virtual PE includes a first PE and a second PE, and the virtual PE has an IP address The virtual PE and the third PE are both member PEs of the first VPLS instance, the first VPLS instance is an intermediate system to an intermediate system ISIS VPLS instance, and the first PE is the distance of the RG medium The third PE is the shortest member PE, and the second CE is the local CE of the third PE in the first VPLS instance, and the method includes:

Receiving, by the third PE, an Ethernet frame sent from the second CE; The third PE encapsulates the Ethernet frame by using the IP address of the virtual PE or the multicast IP address of the multicast tree corresponding to the first VPLS instance as the tunnel destination address, and sends the encapsulated packet to the The virtual PE is configured to enable the member PE of the RG corresponding to the virtual PE to receive the encapsulated packet and decapsulate the packet, and forward the decapsulated packet to the first CE, where the first CE is The virtual CE is in the local CE of the first VPLS instance, and the first CE is connected to the first PE and the second PE in an inter-frame link aggregation group MC-LAG manner.

 A first carrier edge device, the first carrier edge device PE belongs to a redundancy group RG, the RG further includes a second PE, and the RG corresponds to one virtual PE, and the virtual PE has an IP address. The virtual PE is a member PE of the first VPLS instance, the first VPLS instance is an intermediate system to an intermediate system ISIS VPLS instance, and the first user edge device CE is the locality of the virtual PE in the first VPLS instance. The CE is connected to the first PE and the second PE in the MC-LAG mode of the inter-frame link aggregation group, where the first PE includes:

 a first receiving module, configured to receive an Ethernet frame from the first CE,

 a first sending module, configured to encapsulate the encapsulated packet according to the type of the Ethernet frame, and encapsulate the encapsulated packet to the virtual PE in the first VPLS instance. The remote PE sends the encapsulated source address of the encapsulated >3⁄4 text as the IP address of the virtual PE.

 A third carrier edge device, the third carrier edge device PE is used in the network, the network further includes a virtual PE, and the member PE of the RG corresponding to the virtual PE includes the first PE and the second a PE, the virtual PE has an IP address, and the first VPLS instance is an intermediate system to an intermediate system ISIS VPLS instance, and the virtual PE and the third PE are both member PEs of the first VPLS instance, where the A PE is a member of the RG that is the shortest member of the third PE. The second CE is the local CE of the third PE in the first VPLS instance, and the third PE includes:

a second receiving module, configured to receive an Ethernet frame sent from the second CE, and a second sending module, configured to use an IP address of the virtual PE or a multicast IP of a multicast tree corresponding to the first VPLS instance The address is used as the tunnel destination address to encapsulate the Ethernet frame, and the encapsulated packet is sent to the virtual PE, so that the member PE of the RG corresponding to the virtual PE receives the encapsulated packet and performs solution. Encapsulating, and forwarding the decapsulated packet to the first CE, where the first CE is a local CE of the virtual PE in the first VPLS instance, and the first CE is aggregated by a cross-frame link The group MC-LAG mode is respectively associated with the first PE and the second PE Even. According to the method of the present embodiment, the first PE encapsulates the packet by using the IP address of the virtual PE corresponding to the RG of the first PE and the second PE, thereby ensuring that the remote PE receives the packet. When the MAC address learning is performed, the next hop PE corresponding to the MAC is always a virtual PE. Therefore, the next hop PE does not jump between the first PE and the second PE, ensuring the stability of the MAC table.

BRIEF DESCRIPTION OF THE DRAWINGS In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings used in the description of the embodiments will be briefly described below. Obviously, the drawings in the following description are only some of the present invention. For the embodiments, those skilled in the art can obtain other drawings according to the drawings without any creative work.

 1 is a flowchart of a method for forwarding a packet in a network according to an embodiment of the present invention; FIG. 2 is a structural diagram of an ISIS VPLS network according to an embodiment of the present invention;

 FIG. 3 is a flowchart of a method for managing an operator edge device in a network according to an embodiment of the present invention; FIG.

 FIG. 3 is a flowchart of a method for forwarding a packet in a network according to an embodiment of the present invention; FIG. 4 is a flowchart of a method for forwarding a packet in a network according to an embodiment of the present invention; A schematic diagram of a structure of a first carrier edge device according to an embodiment of the invention; FIG. 6 is a schematic structural diagram of a third carrier edge device according to an embodiment of the present invention.

DETAILED DESCRIPTION OF THE EMBODIMENTS In order to make the objects, technical solutions and advantages of the present invention more comprehensible, the embodiments of the present invention will be further described in detail with reference to the accompanying drawings.

FIG. 1 is a flowchart of a method for forwarding a message in a network according to an embodiment of the present invention. The network includes: a first carrier edge device (hereinafter referred to as Provider Edge in English, abbreviated as PE), and a second PE, where the first PE and the second PE belong to a redundancy group (English name is Redundancy Group, English abbreviation is RG ), the RG corresponds to a virtual PE, that is, The virtual PE includes the first PE and the second PE, and the virtual PE has an IP address, and the virtual PE is a first virtual private local area network service. The member VP of the VPLS instance, the first VPLS instance is a VPLS instance of the Intermediate System-Intermediate System (ISIS), and the first user-side edge device (English name is Customer Edge, English abbreviation CE) The first CE is in the local CE of the first VPLS instance, and the first CE is in the multi-Chasis Link Aggegation Group (MC-LAG). And connecting to the first PE and the second PE. Referring to FIG. 1, the method includes:

 101. The first PE receives an Ethernet frame from the first CE.

 The first PE encapsulates the encapsulated packet according to the type of the Ethernet frame, and encapsulates the encapsulated packet to the virtual PE in the first VPLS instance. The remote source PE sends the encapsulated source address of the encapsulated packet as the IP address of the virtual PE.

 According to the method of the present embodiment, the first PE encapsulates the packet by using the IP address of the virtual PE corresponding to the RG of the first PE and the second PE, so that the remote PE receives the packet. When the MAC address learning is performed, the next hop PE corresponding to the MAC is always a virtual PE. Therefore, the next hop PE does not jump between the first PE and the second PE, ensuring stability of the MAC table.

 FIG. 2 is a structural diagram of an ISIS VPLS network according to an embodiment of the present invention. Referring to FIG. 2, the VPLS network includes at least: a first PE, a second PE, a third PE, and a fourth PE, where the first PE, the second PE, the third PE, and the fourth PE are mainly responsible for accessing VPN services. The mapping from the private network to the public network tunnel and the packet from the public network tunnel to the private network is performed. The CE can be a switch or a host. The first CE is in the MC-LAG mode and the first PE. The second PE is connected. Optionally, the VPLS network shown in Figure 2 may also not include the third PE or the fourth PE.

 FIG. 3a is a flowchart of a method for managing an operator edge device in a network according to an embodiment of the present invention. This embodiment uses the ISIS VPLS network shown in FIG. 2 as an example for description. Referring to FIG. 3a, the embodiment specifically includes:

301. The first PE obtains a redundancy group (English name: Redundancy Group, abbreviated as RG). The RG information includes an IP address of the virtual PE corresponding to the RG, and a VPLS identifier of the first VPLS instance corresponding to the RG. The VPLS label assigned by the virtual PE to the first VPLS instance and/or the IP address of other member PEs belonging to the RG, where the first VPLS instance is ISIS VPLS instance. Optionally, the virtual PE corresponding to the RG includes a member PE in the RG.

 The RG includes the first PE, and further includes a second PE, where the virtual PE is a member PE of the first VPLS instance, and the virtual PE includes the first PE and the second PE, and the network further includes The first CE is a local CE of the virtual PE in the first VPLS instance. The first CE is connected to the first PE and the second PE in an MC-LAG manner. Specifically, the first CE needs to be described that each member PE that belongs to the RG obtains the RG information. The acquisition of a PE is described as an example.

 The first PE obtains RG information, which can be used to guide subsequent packet forwarding.

 The VPLS label allocated by the virtual PE to the VPLS instance is generated by any member PE of the virtual PE or configured by a technician.

 In the VPLS network architecture shown in Figure 2, as an example, the actual IP address of the first PE is 1.1.1.2, and the actual IP address of the second PE is 1.1.1.3. The IP address of the virtual PE of the virtual PE can be used. It is set to 1丄1.1, and the setting is performed by a technician, which is not specifically limited in the embodiment of the present invention.

 302. The first PE advertises route reachability information corresponding to the IP address of the virtual PE by using a routing protocol.

 In this embodiment, the second PE also advertises the reachability information corresponding to the IP address of the virtual PE through a routing protocol.

 The VPLS packet of the virtual IP address of the virtual PE (that is, 1.1.1.1) is sent from the remote PE of the virtual PE in the VPLS instance (in this embodiment, the remote PE is the third PE or the PE4). The text will be forwarded to a member PE that is closer to the remote PE.

 303. After the first PE determines that the first PE is the primary PE, the VPLS identifier of the first VPLS instance is advertised. The reachability information of the first VPLS instance includes the VPLS identifier of the first VPLS instance corresponding to the RG. And the virtual PE is the VPLS label assigned to the first VPLS instance and the IP address of the virtual PE.

Optionally, the virtual router redundancy protocol is run between the first PE and the second PE through a communication interface between the two PEs (in English, the Virtual Router Redundancy Protocol is abbreviated as VRRP) or the cross-frame communication protocol. For the Inter-Chassis Communication Protocol (ICCP) protocol, determine the primary PE and the standby PE for any VPLS instance. In this embodiment, for the first VPLS instance to which the first CE belongs, the first PE is used as the primary PE, and the second PE is used as the standby PE, and the first PE determines that the first PE is the primary PE. The first PE represents the reachability information of the first VPLS instance, and the reachability information of the first VPLS instance includes the VPLS identifier of the first VPLS instance and the virtual PE is the first VPLS instance. The assigned VPLS label and the IP address of the virtual PE.

 Preferably, the first PE and the second PE are required to allocate the same VPLS label to the same VPLS instance. To ensure that the first VPLS label is assigned to the same VPLS instance, the VPLS label can be set to the same value as the VPLS identifier. For example, the VPLS label corresponding to the VPLS ID = 100 is equal to 100.

 After the foregoing steps 301-303 are performed, an RG including the first PE and the second PE and a virtual PE corresponding to the RG are formed in the first VPLS instance, and the VPLS instance and the virtual device are sent to other devices in the network. The information of the PE, so that the first PE encapsulates the packet by using the IP address of the virtual PE corresponding to the RG of the first PE and the second PE, so that the remote PE receives the packet and performs the packet. When the MAC address is learned, the next hop PE corresponding to the MAC is always a virtual PE. Therefore, the next hop PE does not hop between the first PE and the second PE, ensuring the stability of the MAC table. Therefore steps 301-303 can be used to help ensure the stability of the MAC table. FIG. 3b is a flowchart of a method for forwarding a message in a network according to an embodiment of the present invention, where the method is applied to a network as shown in FIG. 2. This embodiment may or may not be based on the results obtained by performing steps 301-303 as in Figure 3a. The first PE in the RG corresponding to the virtual PE receives the Ethernet frame sent by the first CE through the aggregation link as an example. The PE of any member of the RG receives the Ethernet sent by the first CE through the aggregation link. The processing method after the frame is the same as the processing method of the first PE. Referring to Figure 3b, this embodiment includes the following.

 304. The first PE receives an Ethernet frame from the first CE (English name is an Ethernet frame); the Ethernet frame may be a known destination unicast Ethernet frame, an unknown destination unicast Ethernet frame, a multicast Ethernet frame, or a broadcast Ethernet Any one or more of the frames.

 305. The first PE encapsulates the encapsulated packet according to the type of the Ethernet frame, and encapsulates the encapsulated packet to the virtual PE in the first VPLS instance. The remote source PE sends the encapsulated source address of the encapsulated packet as the IP address of the virtual PE. Processing method 305-1

When the Ethernet frame is a known destination unicast Ethernet frame, the first PE uses the IP address of the virtual PE as the tunnel source address and the IP address of the destination remote PE as the tunnel destination address. The packet is encapsulated, and the encapsulated packet is sent to the destination remote PE. The destination remote PE is the next hop PE corresponding to the destination MAC address of the Ethernet frame in the MAC forwarding table of the first VPLS instance. ;

 The encapsulation is a VPLS data encapsulation, and the encapsulated packet carries the VPLS label allocated by the destination remote PE to the first VPLS instance, and the label type is a downstream allocation label, so that the destination remote PE that receives the packet can Indicates which VPLS instance the packet encapsulated in the tunnel belongs to.

 Processing method 305-2

 If the Ethernet frame is an unknown destination unicast Ethernet frame, a multicast Ethernet frame, or a broadcast Ethernet frame, if the first VPLS instance uses the ingress replication mode, the Ethernet frame is copied, and the number is obtained with the virtual PE. A plurality of the Ethernet frames of the same number of the remote PEs in the first VPLS instance are encapsulated to obtain a plurality of encapsulated packets, where the tunnel source address of each encapsulated text is obtained. The IP addresses of the virtual PEs, and the tunnel destination address of each of the encapsulated packets is the IP address of a PE in the remote PE, and the tunnel destination addresses of the plurality of encapsulated packets are not mutually And sending the plurality of encapsulated packets according to the tunnel destination address of the plurality of encapsulated packets;

 The first VPLS instance uses the ingress replication mode. Therefore, the received Ethernet frame needs to be replicated to obtain the same number of remote Ethernets as the virtual PE in the first VPLS instance. Frames are encapsulated separately and sent to each remote PE.

 In addition, each encapsulated packet carries a VPLS label allocated by the corresponding remote PE to the first VPLS instance, and the label type is a downstream allocation label, so that the remote PE that receives the packet can identify the tunnel. Which VPLS instance the encapsulated packet belongs to.

 In this embodiment, for the remote PE, the tunnel source address of the received encapsulated packet is the IP address of the virtual PE, and the packet is forwarded by the PE of the RG corresponding to the virtual PE. When the remote PE performs MAC address learning, the virtual PE is used as the next hop PE of the MAC address, that is, the stability of the MAC forwarding table is guaranteed.

 Treatment method 305-3

If the Ethernet frame is an unknown destination unicast Ethernet frame, a multicast Ethernet frame, or a broadcast Ethernet frame, if the first VPLS instance uses the non-aggregated multicast tree mode, the first PE uses the IP address of the virtual PE as a tunnel. The source address and the multicast IP address of the multicast tree corresponding to the first VPLS instance are encapsulated as the tunnel destination address, and the Ethernet frame is encapsulated and sent to the PE other than the first PE added to the multicast tree. Encapsulated message; After receiving the encapsulated packet, the second PE discards the encapsulated packet.

 The non-aggregated multicast tree mode has a multicast tree corresponding to a VPLS instance, and the VPLS instance and the multicast tree are corresponding relationships.

 Because the VPLS instance and the multicast tree are in the corresponding relationship, the VPLS label does not need to be carried. The PE that receives the encapsulated packet can know which VPLS instance the file belongs to by using the destination multicast IP address.

 Further, the first PE sends the Ethernet frame to the second PE, so that the second PE uses the IP address of the virtual PE as the tunnel source address and the multicast IP address of the multicast tree corresponding to the first VPLS instance as the tunnel destination address. Encapsulating the Ethernet frame, and transmitting, by the multicast tree corresponding to the first VPLS instance, a packet encapsulated by the second PE to a PE other than the second PE that is added to the multicast tree.

 It should be noted that the first PE and the second PE are member PEs of the RG corresponding to the virtual PE, and the Ethernet frame interaction between the first PE and the second PE may be through the communication interface between the first PE and the second PE. get on.

 Processing method 305-4

 If the Ethernet frame is an unknown destination unicast Ethernet frame, a multicast Ethernet frame, or a broadcast Ethernet frame, if the first VPLS instance uses the aggregate multicast tree mode, the IP address of the virtual PE is used as the tunnel source address and used. The multicast IP address of the multicast tree corresponding to the first VPLS instance is used as the tunnel destination address to encapsulate the Ethernet frame, and the encapsulated packet carries the first VPLS label, where the first VPLS label is the virtual PE. a VPLS label allocated by a VPLS instance, where the label type of the first VPLS label is an upstream label, and the encapsulated packet is sent to a PE other than the first PE that is added to the multicast tree.

 In the aggregation multicast tree mode, multiple VPLS instances share a multicast tree. The VPLS instance and the multicast tree have a many-to-one relationship.

 In the aggregated multicast tree mode, the VPLS instance is in a multi-to-one relationship with the multicast tree, so that the first VPLS label allocated by the virtual PE for the first VPLS instance is carried in the encapsulated packet, so that The PE that receives the encapsulated packet can know which VPLS instance the packet belongs to. When the non-aggregation multicast tree method is applied, the first VPLS label may not be carried.

Further, the first PE sends the Ethernet frame to the second PE, so that the second PE uses the IP address of the virtual PE as the tunnel source address and the group of the multicast tree corresponding to the first VPLS instance. The broadcast IP address is encapsulated as the tunnel destination address, and the encapsulated packet is encapsulated. The first VPLS label carries the first VPLS label, and the first VPLS label is a VPLS label that is allocated to the first VPLS instance, and the label type of the first VPLS label is an upstream allocation label, and the group corresponding to the first VPLS instance is configured. The broadcast tree sends the packet encapsulated by the second PE to the PEs other than the second PE that are added to the multicast tree. The first PE and the second PE are member PEs of the RG corresponding to the virtual PE, and the Ethernet frame interaction between the first PE and the second PE may be performed through a communication interface between the first PE and the second PE.

 According to the method of the present embodiment, the first PE encapsulates the packet by using the IP address of the virtual PE corresponding to the RG of the first PE and the second PE, so that the remote PE receives the packet. When the MAC address learning is performed, the next hop PE corresponding to the MAC is always a virtual PE. Therefore, the next hop PE does not jump between the first PE and the second PE, ensuring stability of the MAC table.

 FIG. 4 is a flowchart of a method for forwarding a message in a network according to an embodiment of the present invention. The network includes: a third PE and a virtual PE, where the member PE of the RG corresponding to the virtual PE includes at least a first PE and a second PE, that is, the virtual PE includes the first PE and the second The PE, the virtual PE has an IP address, and the virtual PE and the third PE are both member PEs of the first VPLS instance, and the first PE is the member PE of the RG that is the shortest route from the third PE. The second CE is a local CE of the third PE in the first VPLS instance, and the first VPLS instance is an intermediate system to an intermediate system ISIS VPLS instance. The third PE that is the remote PE of the virtual PE receives the packet as an example. Referring to FIG. 4, the embodiment includes the following content:

 401. The third PE receives an Ethernet frame sent from the second CE.

 Optionally, before the step 401, the embodiment further includes: the third PE receives the route reachability information corresponding to the IP address of the virtual PE that is advertised by the one or more member PEs of the virtual PE by using the routing protocol. The third PE receives the reachability information of the VPLS instance corresponding to the RG that is advertised by the first PE that is the primary PE, and the reachability information of the VPLS instance includes at least the VPLS of the VPLS instance corresponding to the RG. The VPLS label assigned to the VPLS instance and the IP address of the virtual PE.

The third PE encapsulates the Ethernet frame by using the IP address of the virtual PE or the multicast IP address of the multicast tree corresponding to the first VPLS instance as the tunnel destination address, and sends the encapsulated packet. Giving the virtual PE, the member PE of the RG corresponding to the virtual PE, receiving the encapsulated packet, decapsulating the packet, and forwarding the decapsulated packet to the first CE, the first CE The local CE of the virtual PE is in the first VPLS instance, and the first CE is connected to the first PE and the second PE in an MC-LAG manner. The first PE Routing the shortest member PE for the third PE from the RG. Processing method 402-1

 If the Ethernet frame is a known destination unicast Ethernet frame, if the next hop corresponding to the destination MAC address of the Ethernet frame is the virtual PE, the third PE uses the IP address of the virtual PE as the tunnel destination address and The IP address of the third PE is used as the tunnel source address to encapsulate the Ethernet frame, and the encapsulated > PDU carries the first VPLS label, and the first VPLS label is the VPLS label allocated by the virtual PE to the first VPLS instance. The label type of the first VPLS label is a downstream label, and the encapsulated packet is sent to the first PE, so that the first PE receives the encapsulated packet and decapsulates the packet, and decapsulates the packet. The packet is forwarded to the first CE, and the first CE is the local CE of the virtual PE in the first VPLS instance.

 It should be noted that the encapsulated packet is received by the shortest member PE of the virtual PE or RG from the third PE.

 In this embodiment, only the first PE is the member PE of the RG corresponding to the virtual PE and the shortest route from the third PE.

 Processing method 402-2

 If the Ethernet frame is an unknown destination unicast Ethernet frame, a multicast Ethernet frame, or a broadcast Ethernet frame, if the first VPLS instance uses the ingress replication mode, the third PE copies the Ethernet frame to obtain the quantity and And the plurality of Ethernet frames in the first VPLS instance are the same number of the Ethernet frames; and the plurality of Ethernet frames are respectively encapsulated to obtain a plurality of encapsulated packets, where each of the encapsulated packets The source address of the tunnel is the IP address of the third PE, and the tunnel destination address of each encapsulated packet is the IP address of a PE in the remote PE of the third PE. The tunnel destination address of the subsequent packet is different from each other; the encapsulated packet is sent to the corresponding remote PE, so that the first PE receives the encapsulated packet whose destination address is the IP address of the virtual PE and performs the encapsulation. Decapsulating, and forwarding the decapsulated packet to the first CE, where the encapsulated packet whose destination address is the IP address of the virtual PE carries the first VPLS label, where the first VPLS label is the virtual PE is the first V The VPLS label assigned by the PLS instance. The label type of the first VPLS label is the downstream allocation label.

 It should be noted that the encapsulated packet is received by the shortest member PE of the virtual PE or RG from the third PE.

In this embodiment, only the first PE is the RG in the virtual PE corresponding to the third PE route. The short member PE is described as an example.

 Processing method 402-3

 When the first PE is the primary PE, the second PE is the standby PE, the first PE is added to the multicast tree corresponding to the first VPLS, and the Ethernet frame is an unknown destination unicast Ethernet frame, multicast If the first VPLS instance is in the non-aggregated multicast tree mode, the third PE uses the multicast IP address of the multicast tree corresponding to the first VPLS instance as the tunnel destination address and uses the first The IP address of the third PE is used as the tunnel source address to encapsulate the Ethernet frame, and the encapsulated packet is forwarded by the multicast tree corresponding to the first VPLS instance, so that the first PE receives the encapsulated packet and The decapsulation is performed, and the decapsulated packet is forwarded to the first CE.

 Processing method 402-4

 When the first PE is the primary PE, the second PE is the standby PE, the first PE is added to the multicast tree corresponding to the first VPLS, and the Ethernet frame is an unknown destination unicast Ethernet frame, multicast If the first VPLS instance is in the aggregate multicast tree mode, the third PE uses the multicast IP address of the multicast tree corresponding to the first VPLS instance as the tunnel destination address and uses the first The IP address of the third PE is used as the tunnel source address to encapsulate the Ethernet frame. The encapsulated packet carries the third VPLS label, and the third VPLS label is the VPLS label allocated by the third PE to the first VPLS instance. The label type of the VPLS label is an upstream label, and the encapsulated packet is forwarded through the multicast tree, so that the first PE receives the encapsulated packet and decapsulates the packet, and decapsulates the packet. Forwarded to the first CE.

 In addition, in the aggregation multicast tree mode, the VPLS instance is in a many-to-one relationship with the multicast tree, and the encapsulated packet carries the third VPLS label, and the third VPLS label is the third PE is the first VPLS. The VPLS label is assigned by the instance. The label type of the third VPLS label is an upstream label. The PE that receives the encapsulated packet knows which VPLS instance the packet comes from.

It should be noted that, in the aggregated or non-aggregated multicast tree mode, when the first PE is the primary PE and the second PE is the standby PE, the second PE may also join the first VPLS. In the multicast tree corresponding to the instance, when the first PE and the second PE are added to the multicast tree corresponding to the first VPLS, in order to prevent the first CE from receiving the repeated packets forwarded from the second PE, After receiving the encapsulated packet from the multicast tree corresponding to the first VPLS, the second PE discards the encapsulated packet. Optionally, after the second PE that is the primary PE fails, the second PE is switched to the primary PE, and the second PE receives the encapsulated report by using the multicast tree corresponding to the first VPLS. The second PE decapsulates the encapsulated packet and forwards the decapsulated packet to the first CE. According to the method of the present embodiment, in the member PE of the RG corresponding to the virtual PE, only one member PE receives the encapsulated packet sent from the third PE, and decapsulates the packet and forwards the packet to the first CE, thereby avoiding The first CE receives the duplicated message. FIG. 5 is a schematic structural diagram of a first carrier edge device according to an embodiment of the present invention. The first carrier edge device (English Provider Edge, abbreviated as PE) belongs to a redundancy group (English name is Redundancy Group, abbreviated as RG), and the RG further includes a second PE, and the RG corresponds to one Virtual PE, that is, the virtual PE includes the first PE and the second PE, the virtual PE has an IP address, and the virtual PE is a first virtual private local area network service (English name is Virtual Private) The local area network service (abbreviated as VPLS) is the member PE of the instance. The first VPLS instance is the VPLS instance of the Intermediate System-Intermediate System (English). The first user edge device (English full name is Custom) Edge, abbreviated as CE, is the local CE of the virtual PE in the first VPLS instance. Referring to FIG. 5, the first PE includes:

 The first receiving module 501 is configured to receive an Ethernet frame from the first CE.

 The first sending module 502 is configured to encapsulate the encapsulated packet according to the type of the Ethernet frame, and encapsulate the encapsulated packet to the virtual PE in the first VPLS. The remote PE is sent by the instance, where the tunnel source address of the encapsulated text is the IP address of the virtual PE.

 Optionally, the function of the first receiving module 501 is implemented by a first physical receiving port, and the function of the first sending module 502 is implemented by a first physical circuit, where the first physical circuit includes a first processor and a first physical sending port, where the first processor is configured to perform the encapsulating process in the first sending module 502, where the first physical sending port is configured to perform the sending in the first sending module 502 deal with.

Optionally, the first sending module 502 includes one or more of the following units: a first sending unit, configured to use the virtual PE when the Ethernet frame is a known destination unicast Ethernet frame The IP address is used as the tunnel source address and the IP address of the destination remote PE is used as the tunnel destination address to encapsulate the Ethernet frame, and the encapsulated packet is sent to the destination remote PE, where the destination is far The end PE is the next hop PE corresponding to the destination MAC address of the Ethernet frame in the MAC forwarding table of the first VPLS instance. Optionally, the first sending unit is configured to execute the content of the “processing mode 305-1”. a second sending unit, configured to: if the Ethernet frame is an unknown destination unicast Ethernet frame, a multicast Ethernet frame, or a broadcast Ethernet frame, if the first VPLS instance uses an ingress copy mode, perform the Copying, obtaining a plurality of the Ethernet frames having the same number of remote PEs as the virtual PE in the first VPLS instance; and encapsulating the multiple Ethernet frames to obtain a plurality of encapsulated packets The tunnel source address of each encapsulated packet is an IP address of the virtual PE, and the tunnel destination address of each encapsulated packet is an IP of a PE in the remote PE. The address, the tunnel destination address of the plurality of encapsulated packets is different from each other; and the plurality of encapsulated packets are sent according to the tunnel destination address of the plurality of encapsulated packets. Optionally, the second sending unit is configured to execute the content of the “processing mode 305-2”.

 a third sending unit, configured to: when the Ethernet frame is an unknown destination unicast Ethernet frame, a multicast Ethernet frame, or a broadcast Ethernet frame, if the first VPLS instance uses a non-aggregated multicast tree mode, use the The IP address of the virtual PE is used as the tunnel source address and the multicast IP address of the multicast tree corresponding to the first VPLS instance is used as the tunnel destination address to encapsulate the Ethernet frame and join the multicast tree. The PE other than the first PE sends the encapsulated text. Optionally, the third sending unit is configured to execute the content of the “processing mode 305-3”.

 a fourth sending unit, configured to: when the Ethernet frame is an unknown destination unicast Ethernet frame, a multicast Ethernet frame, or a broadcast Ethernet frame, if the first VPLS instance uses an aggregate multicast tree mode, use the virtual The IP address of the PE is used as the tunnel source address and the multicast IP address of the multicast tree corresponding to the first VPLS instance is used as the tunnel destination address to encapsulate the packet, and the encapsulated packet carries the first packet. a VPLS label, the first VPLS label is a VPLS label allocated by the virtual PE to the first VPLS instance, and the label type of the first VPLS label is an upstream allocation label, and is added to the multicast tree. The PE other than the first PE sends the encapsulated packet. Optionally, the fourth sending unit is configured to execute the content of the “processing mode 305-4”.

 Optionally, the functions of each of the first sending unit, the second sending unit, the third sending unit, and the fourth sending unit are all implemented by the first processor and the first physical sending port.

Optionally, the third sending unit is further configured to send the Ethernet frame to the second PE, so that the second PE uses an IP address of the virtual PE as a tunnel source address, and uses the Encapsulating the Ethernet frame with the multicast IP address of the multicast tree corresponding to the VPLS instance as the tunnel destination address, and adding the multicast tree corresponding to the first VPLS instance to the multicast tree The PE other than the second PE sends the packet encapsulated by the second PE. Optionally, the fourth sending unit is further configured to send the Ethernet frame to the second PE, so that the second PE uses an IP address of the virtual PE as the tunnel source address and uses The multicast IP address of the multicast tree corresponding to the first VPLS instance encapsulates the Ethernet frame as a tunnel destination address, and the packet encapsulated by the second PE carries a first VPLS label, where the first VPLS The label is a VPLS label that is allocated by the virtual PE to the first VPLS instance, and the label type of the first VPLS label is an upstream allocation label, and is added to the foregoing by a multicast tree corresponding to the first VPLS instance. A PE other than the second PE in the multicast tree sends the packet encapsulated by the second PE.

 Optionally, the first carrier edge device further includes:

 The first obtaining module 503 is configured to obtain RG information, where the RG information includes an IP address of the virtual PE corresponding to the RG, a VPLS identifier of the first VPLS instance corresponding to the RG, and the virtual PE is The IP address of the VPLS label assigned by the first VPLS instance and/or other member PEs belonging to the RG;

 a publishing module 504, configured to advertise routing reachability information corresponding to an IP address of the virtual PE by using a routing protocol;

 The issuing module 504 is further configured to: after determining that the operator edge device is the primary PE, advertise the reachability information of the first VPLS instance, where the reachability information of the first VPLS instance includes the The VPLS identifier of the VPLS instance corresponding to the RG and the virtual PE are the VPLS label allocated by the VPLS instance and the IP address of the virtual PE.

 Optionally, the function of the first obtaining module 503 is implemented by a second physical circuit, and the second physical circuit may include a second physical receiving port to receive the RG information. The second physical receiving port may be the same port as the first physical receiving port, or may be a different port. The second physical circuit may further include a second processor to identify the RG information. The second processor may be the same processor as the first processor, or may be a different processor.

 Optionally, the first PE further includes:

 The active/standby negotiation module 505 is configured to determine an active PE and a standby PE with the second PE. Optionally, the function of the active/standby negotiation module 505 is implemented by one processor.

In this embodiment, the first PE encapsulates the IP address of the virtual PE corresponding to the RG of the first PE and the second PE, thereby ensuring that the remote PE receives the packet When the MAC address is learned, the next hop PE corresponding to the MAC is always a virtual PE. Therefore, the next hop PE does not jump between the first PE and the second PE, ensuring the stability of the MAC table. In an embodiment of the present invention, a first carrier edge device (English full name

Provider Edge (abbreviated as PE) belongs to a redundancy group (English name is Redundancy Group, abbreviated as RG), and the RG further includes a second PE, where the RG corresponds to a virtual PE, that is, the virtual PE The first PE and the second PE, where the virtual PE has an IP address, and the virtual PE is an instance of a first virtual private local area network service (hereinafter referred to as VPLS) The member PE, the first VPLS instance is a VPLS instance of the Intermediate System-Intermediate System (ISIS), and the first user edge device (referred to as Custom Edge in English, CE) is the virtual PE. Local CE within the first VPLS instance.

 The first PE includes the following content.

 An acquiring module, configured to obtain RG information, where the RG information includes an IP address of the virtual PE corresponding to the RG, a VPLS identifier of the first VPLS instance corresponding to the RG, and the virtual PE is the The VPLS label assigned by a VPLS instance and/or the IP address of other member PEs belonging to the RG. Optionally, the obtaining module is configured to perform step 301 shown in Figure 3a.

 And a publishing module, configured to advertise the route reachability information corresponding to the IP address of the virtual PE by using a routing protocol. Optionally, the publishing module is configured to perform step 302 shown in Figure 3a.

 The issuance module is further configured to: after determining that the operator edge device is the primary PE, advertise the reachability information of the first VPLS instance, where the reachability information of the first VPLS instance includes the RG corresponding The VPLS identifier of the VPLS instance and the VPLS label assigned by the virtual PE to the VPLS instance and the IP address of the virtual PE. Optionally, the publishing module is further configured to perform step 303 shown in Figure 3a.

 The first PE is configured to form an RG including the first PE and the second PE and a virtual PE corresponding to the RG in the first VPLS instance, and send the VPLS instance and the virtual PE to other devices in the network. The information is such that the first PE uses the IP address of the virtual PE corresponding to the RG of the first PE and the second PE to encapsulate the packet, thereby ensuring that the remote PE receives the packet and performs the packet. When the MAC address is learned, the next hop PE corresponding to the MAC is always a virtual PE, so that the next hop PE does not jump between the first PE and the second PE, ensuring the stability of the MAC table. Therefore the first PE can be used to help ensure MAC table stability.

FIG. 6 is a schematic structural diagram of a third carrier edge device according to an embodiment of the present invention. The third carrier edge device (PE) is used in a network, the network including the In addition to the third carrier edge device (PE), the virtual PE also includes a virtual PE, and the member PE of the RG corresponding to the virtual PE includes a first PE and a second PE, that is, the virtual PE includes the first PE and The second PE, the virtual PE has an IP address, and the virtual PE and the third PE are both member PEs of the first VPLS instance, and the first PE is the third PE of the RG. The second CE is the local CE of the third PE in the first VPLS instance, and the first VPLS instance is the intermediate system to the intermediate system ISIS VPLS instance. The third PE includes: a second receiving module 601, configured to receive an Ethernet frame sent from the second CE; and a second sending module 602, configured to use an IP address of the virtual PE or a first VPLS instance The multicast IP address of the corresponding multicast tree is used as the tunnel destination address to encapsulate the Ethernet frame, and the encapsulated packet is sent to the virtual PE, so that the member PE of the RG corresponding to the virtual PE receives the The encapsulated packet is decapsulated and decapsulated, and the decapsulated packet is forwarded to the first CE, where the first CE is the local CE of the virtual PE in the first VPLS instance, and the The first CE is connected to the first PE and the second PE in an MC-LAG manner.

 Optionally, the second receiving module 601 and the second sending module 602 are respectively configured to perform steps 401 and 402 shown in FIG.

 The first PE in the embodiment is the shortest member of the third PE route in the RG.

PE.

 Optionally, the function of the second receiving module 601 is implemented by a third physical receiving port, and the function of the second sending module 602 is implemented by a third physical circuit, where the third physical circuit includes a third processor and a third physical sending port, where the third processor is configured to perform the encapsulating process in the second receiving module 602, where the third physical sending port is configured to perform the sending in the second sending module 602 deal with.

Optionally, the second sending module 602 includes one or more of the following units: a fifth sending unit, configured to: when the Ethernet frame is a known destination unicast Ethernet frame, if the Ethernet frame is If the next hop PE corresponding to the destination MAC address is the virtual PE, the IP address of the virtual PE is used as the tunnel destination address, and the IP address of the third carrier edge device is used as the tunnel source address to the Ethernet frame. Encapsulating, the encapsulated packet carries a first VPLS label, the first VPLS label is a VPLS label allocated by the virtual PE to the first VPLS instance, and the label type of the first VPLS label is a downstream label. Sending the encapsulated packet to the first PE, so that the first PE receives the encapsulated packet and decapsulates the packet, and forwards the decapsulated packet to the first packet. CE. Optionally, the fifth sending unit is configured to perform The "processing mode 402-1".

 a sixth sending unit, configured to: when the Ethernet frame is an unknown destination unicast Ethernet frame, a multicast Ethernet frame, or a broadcast Ethernet frame, if the first VPLS instance uses an ingress copy mode, the Ethernet frame is used Performing the copying to obtain a plurality of the Ethernet frames that are the same as the number of the remote PEs in the first VPLS instance of the third carrier edge device; and separately encapsulating the multiple Ethernet frames to obtain multiple The encapsulated packet, wherein the tunnel source address of each encapsulated packet is an IP address of the third carrier edge device, and the tunnel destination address of each encapsulated packet is The IP address of a PE in the remote PE of the third carrier edge device, and the tunnel destination address of the plurality of encapsulated packets are different from each other; and the encapsulated packet is sent to the corresponding remote PE. And performing decapsulation, and forwarding the decapsulated packet to the first CE, where the encapsulated packet whose destination address is the IP address of the virtual PE carries the first VPLS label, where a VPLS label For the VPLS label that is allocated to the first VPLS instance by the virtual PE, the label type of the first VPLS label is a downstream allocation label. Optionally, the sixth sending unit is configured to execute the "processing mode 402-2".

 a seventh sending unit, configured to: the first PE is a primary PE, the second PE is a standby PE, the first PE is added to a multicast tree corresponding to the first VPLS, and the Ethernet frame is When unicasting an Ethernet frame, a multicast Ethernet frame, or a broadcast Ethernet frame for an unknown destination, if the first VPLS instance uses the non-aggregated multicast tree mode, the multicast IP address of the multicast tree corresponding to the first VPLS instance is used. Encapsulating the Ethernet frame as a tunnel destination address and using an IP address of the third carrier edge device as a tunnel source address, and forwarding the encapsulated packet to a multicast tree corresponding to the first VPLS instance The first PE receives the encapsulated packet and decapsulates the packet, and forwards the decapsulated packet to the first CE. Optionally, the seventh sending unit is configured to execute the “processing mode 402-3”.

An eighth sending unit, configured to: when the first PE is a primary PE, the second PE is a standby PE, the first PE is added to a multicast tree corresponding to the first VPLS, and the ether is If the frame is an unknown destination unicast Ethernet frame, a multicast Ethernet frame, or a broadcast Ethernet frame, if the first VPLS instance uses the aggregate multicast tree mode, the multicast of the multicast tree corresponding to the first VPLS instance is used. The IP address is used as the tunnel destination address and the IP address of the edge device of the third carrier is used as the tunnel source address to encapsulate the Ethernet frame, and the encapsulated packet carries a third VPLS label, where the third VPLS label is a VPLS label allocated by the third carrier edge device to the first VPLS instance. The label type of the third VPLS label is an upstream label, and the encapsulated packet is forwarded by the multicast tree, so that the first PE receives the encapsulated packet and decapsulates the packet. The encapsulated packet is forwarded to the first CE. Optionally, the eighth sending unit is configured to execute the “processing mode 402-4”.

 Optionally, the functions of each of the fifth sending unit, the sixth sending unit, the seventh sending unit, and the eighth sending unit are all implemented by the third processor and the third physical sending port. Optionally, the third carrier edge device further includes:

 The third receiving module 603 is configured to receive route reachability information corresponding to the IP address of the virtual PE that is advertised by the PE or the member PE through the routing protocol;

 The third receiving module 603 is further configured to receive reachability information of the VPLS instance corresponding to the RG that is advertised by the first PE that is the primary PE, where the reachability information of the VPLS instance includes the VPLS identifier of the VPLS instance corresponding to the RG, the virtual PE is a VPLS label allocated by the VPLS instance and an IP address of the virtual PE.

 Optionally, the function of the third receiving module 603 is implemented by a fourth physical receiving port, and the fourth physical receiving port may be the same port as the third physical receiving port, or may be a different port.

 In this implementation, the member CE of the RG corresponding to the virtual PE receives only the encapsulated packet sent from the third PE, and decapsulates the packet and forwards the packet to the first CE to avoid the first CE. Received duplicate messages. A person skilled in the art may understand that all or part of the steps of implementing the above embodiments may be completed by hardware, or may be instructed by a program to execute related hardware, and the program may be stored in a computer readable storage medium. The storage medium mentioned may be a read only memory, a magnetic disk or an optical disk or the like.

 The above is only the preferred embodiment of the present invention, and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., which are within the spirit and scope of the present invention, should be included in the protection of the present invention. Within the scope.

Claims

Rights request

A method for forwarding a packet in a network, the network includes: a first carrier edge device PE, a second PE, and the first PE and the second PE belong to a redundancy group RG, the RG corresponds to a virtual PE, the virtual PE has an IP address, the virtual PE is a member PE of the first virtual private local area network service VPLS instance, and the first VPLS instance is an intermediate system to an intermediate system ISIS VPLS instance. The first user edge device CE is the local CE of the virtual PE in the first VPLS instance, and the first CE is in the inter-frame link aggregation group MC-LAG manner and the first PE and the The second PE is connected, and the method includes:

 The first PE receives an Ethernet frame from the first CE;

 The first PE encapsulates the encapsulated packet according to the type of the Ethernet frame, and encapsulates the encapsulated packet into the virtual PE in the first VPLS instance. The remote PE sends, where the tunnel source address of the encapsulated packet is the IP address of the virtual PE.

The method according to claim 1, wherein the first PE encapsulates the Ethernet frame according to the type of the Ethernet frame, and obtains the encapsulated packet, and encapsulates the encapsulated packet. And sending, by the virtual PE, the remote PE in the first VPLS instance, specifically: when the Ethernet frame is a known destination unicast Ethernet frame, using the IP address of the virtual PE as the The tunnel source address and the IP address of the destination remote PE are used as the tunnel destination address to encapsulate the Ethernet frame, and the encapsulated packet is sent to the destination remote PE, where the destination remote PE is the The next ϋ PE corresponding to the destination MAC address of the Ethernet frame in the MAC forwarding table of a VPLS instance;

 and / or,

 When the Ethernet frame is an unknown destination unicast Ethernet frame, a multicast Ethernet frame, or a broadcast Ethernet frame, if the first VPLS instance uses the ingress copy mode, the Ethernet frame is copied, and the quantity is obtained. a plurality of the Ethernet frames of the same number of the remote PEs in the first VPLS instance; the plurality of Ethernet frames are respectively encapsulated to obtain a plurality of encapsulated packets, where each of the The tunnel source address of the encapsulated packet is the IP address of the virtual PE, and the tunnel destination address of each encapsulated packet is the IP address of a PE in the remote PE. The tunnel destination addresses of the encapsulated packets are different from each other; the plurality of encapsulated packets are sent according to the tunnel destination address of the plurality of encapsulated packets;

and / or, When the Ethernet frame is an unknown destination unicast Ethernet frame, a multicast Ethernet frame, or a broadcast Ethernet frame, if the first VPLS instance uses a non-aggregated multicast tree mode, the IP address of the virtual PE is used as the Encapsulating the Ethernet frame with the tunnel source address and the multicast IP address of the multicast tree corresponding to the first VPLS instance as a tunnel destination address, and adding to the first PE in the multicast tree The PE sends the encapsulated packet;

 and / or,

 When the Ethernet frame is an unknown destination unicast Ethernet frame, a multicast Ethernet frame, or a broadcast Ethernet frame, if the first VPLS instance uses an aggregate multicast tree mode, the IP address of the virtual PE is used as the The tunnel source address and the multicast IP address of the multicast tree corresponding to the first VPLS instance are used as the tunnel destination address to encapsulate the Ethernet frame, and the encapsulated packet carries the first VPLS label, where the first The VPLS label is a VPLS label that is allocated to the first VPLS instance by the virtual PE, and the label type of the first VPLS label is an upstream allocation label, and is added to the first PE that is added to the multicast tree. The PE sends the encapsulated packet.

3. The method of claim 2, wherein

 When the Ethernet frame is an unknown destination unicast Ethernet frame, a multicast Ethernet frame, or a broadcast Ethernet frame, if the first VPLS instance uses a non-aggregated multicast tree mode, the IP address of the virtual PE is used as the Encapsulating the Ethernet frame with the tunnel source address and the multicast IP address of the multicast tree corresponding to the first VPLS instance as a tunnel destination address, and adding to the first PE in the multicast tree After the PE sends the encapsulated packet, the PE further includes:

 The first PE sends the Ethernet frame to the second PE, so that the second PE uses the IP address of the virtual PE as the tunnel source address and uses the multicast tree corresponding to the first VPLS instance. The multicast IP address is encapsulated as the tunnel destination address, and the Ethernet frame is encapsulated by the multicast tree corresponding to the first VPLS instance to be sent to the PE other than the second PE added to the multicast tree. The packet encapsulated by the second PE;

 and / or,

When the Ethernet frame is an unknown destination unicast Ethernet frame, a multicast Ethernet frame, or a broadcast Ethernet frame, if the first VPLS instance uses an aggregate multicast tree mode, the IP address of the virtual PE is used as the The tunnel source address and the multicast IP address of the multicast tree corresponding to the first VPLS instance are used as the tunnel destination address to encapsulate the Ethernet frame, and the encapsulated packet carries the first VPLS label, where the first The VPLS label is that the virtual PE is allocated to the first VPLS instance. The VPLS label, the label type of the first VPLS label is an upstream allocation label, and the encapsulated packet is sent to a PE other than the first PE that is added to the multicast tree, and the method further includes:

 Sending, by the first PE, the Ethernet frame to the second PE, so that the second PE uses the IP address of the virtual PE as the tunnel source address and the group corresponding to the first VPLS instance. The multicast IP address of the broadcast tree is used as the tunnel destination address to encapsulate the Ethernet frame, and the packet encapsulated by the second PE carries the first VPLS label, where the first VPLS label is the virtual PE. a VPLS label allocated by the first VPLS instance, the label type of the first VPLS label is an upstream allocation label, and the second is added to the multicast tree by using a multicast tree corresponding to the first VPLS instance. The PE other than the PE sends the >3⁄4 text after the second PE encapsulation.

The method according to claim 1, wherein before the first PE receives the Ethernet frame from the first CE, the method includes:

 The first PE acquires RG information, where the RG information includes an IP address of the virtual PE corresponding to the RG, a VPLS identifier of the first VPLS instance corresponding to the RG, and the virtual PE is the first a VPLS label assigned by a VPLS instance and/or an IP address of another member PE belonging to the RG;

 The first PE advertises route reachability information corresponding to the IP address of the virtual PE by using a routing protocol;

 After the first PE determines the first PE as the primary PE, the privilege information of the first VPLS instance is advertised, and the reachability information of the first VPLS instance includes the VPLS corresponding to the RG. IP address of the PE.

A method for forwarding a packet in a network, where the network includes: a third PE and a virtual PE, where the member PE of the RG corresponding to the virtual PE includes a first PE and a second PE, The virtual PE has an IP address, and the virtual PE and the third PE are both member PEs of the first VPLS instance, and the first VPLS instance is an intermediate system to an intermediate system ISIS VPLS instance, and the first PE is the The SG is the local CE that is the shortest member of the third PE, and the second CE is the local CE of the third PE in the first VPLS instance, and the method includes:

 Receiving, by the third PE, an Ethernet frame sent from the second CE;

The third PE uses the IP address of the virtual PE or the multicast corresponding to the first VPLS instance. The multicast IP address of the tree is used as the tunnel destination address to encapsulate the Ethernet frame, and the encapsulated packet is sent to the virtual PE, so that the member PE of the RG corresponding to the virtual PE receives the encapsulated And decapsulating the packet, and forwarding the decapsulated packet to the first CE, where the first CE is the local CE of the virtual PE in the first VPLS instance, and the first CE is The inter-frame link aggregation group MC-LAG mode is respectively connected to the first PE and the second PE.

The method according to claim 5, wherein the third PE uses the IP address of the virtual PE or the multicast IP address of the multicast tree corresponding to the first VPLS instance as the tunnel destination address pair. Encapsulating the Ethernet frame, and sending the encapsulated packet to the virtual PE, so that the member PE of the RG corresponding to the virtual PE receives the encapsulated packet and decapsulates the packet, and decapsulates the packet. The packet is forwarded to the first CE, including:

 When the Ethernet frame is a known destination unicast Ethernet frame, if the next hop PE corresponding to the destination MAC address of the Ethernet frame is the virtual PE, the third PE uses the IP of the virtual PE. The address is used as the tunnel destination address, and the IP address of the third PE is used as the tunnel source address to encapsulate the Ethernet frame. The encapsulated packet carries the first VPLS label, and the first VPLS label is the virtual PE. The VPLS label allocated by the first VPLS instance, the label type of the first VPLS label is a downstream allocation label, and the encapsulated packet is sent to the first PE, so that the first PE receives the Encapsulating the packet and decapsulating the packet, and forwarding the decapsulated packet to the first CE;

When the Ethernet frame is an unknown destination unicast Ethernet frame, a multicast Ethernet frame, or a broadcast Ethernet frame, if the first VPLS instance uses the ingress replication mode, the third PE replicates the Ethernet frame. And obtaining, by the plurality of the Ethernet frames, the number of the remote PEs of the third PE in the first VPLS instance, and encapsulating the plurality of Ethernet frames to obtain a plurality of encapsulated packets. The tunnel source address of each encapsulated packet is the IP address of the third PE, and the tunnel destination address of each encapsulated document is the remote PE of the third PE. The IP address of a PE, the tunnel destination address of the multiple encapsulated packets is different from each other; the encapsulated packet is sent to the corresponding remote PE, so that the first PE receives the tunnel destination address as The encapsulated packet of the IP address of the virtual PE is decapsulated, and the decapsulated packet is forwarded to the first CE, where the tunnel destination address is an encapsulation of the IP address of the virtual PE. The subsequent packet carries the first VPLS label, and the first VPLS label The virtual PE is assigned to the first VPLS VPLS instance tag, the first tag type VPLS downstream label distribution standard 佥

 When the first ΡΕ is the primary ΡΕ, the second ΡΕ is the standby ΡΕ, the first ΡΕ is added to the multicast tree corresponding to the first VP LS, and the Ethernet frame is an unknown destination unicast If the first VPLS instance is in the non-aggregated multicast tree mode, the third PE uses the multicast IP address of the multicast tree corresponding to the first VPLS instance as the Ethernet frame. Encapsulating the Ethernet frame with the tunnel destination address and using the IP address of the third PE as the tunnel source address, and forwarding the encapsulated packet to the multicast tree corresponding to the first VPLS instance, so that the The first PE receives the encapsulated packet and decapsulates the packet, and forwards the decapsulated packet to the first CE;

 When the first PE is the primary PE, the second PE is the standby PE, the first PE is added to the multicast tree corresponding to the first VPLS, and the Ethernet frame is an unknown destination unicast If the first VPLS instance is in the aggregate multicast tree mode, the third PE uses the multicast IP address of the multicast tree corresponding to the first VPLS instance, in the case of the Ethernet frame, the multicast Ethernet frame, or the broadcast Ethernet frame. Encapsulating the Ethernet frame as a tunnel destination address and using the IP address of the third PE as a tunnel source address, the encapsulated packet carrying a third VPLS label, and the third VPLS label being the third PE The VPLS label allocated by the first VPLS instance, the label type of the third VPLS label is an upstream allocation label, and the encapsulated packet is forwarded through the multicast tree, so that the first PE receives the encapsulation The subsequent packet is decapsulated and the decapsulated packet is forwarded to the first CE.

The method according to claim 5, wherein the third PE uses the IP address of the virtual PE or the multicast IP address of the multicast tree corresponding to the first VPLS instance as the tunnel destination address pair. After the Ethernet frame is encapsulated, and the encapsulated packet is sent to the virtual PE, the method further includes:

 When the first PE is the primary PE and the second PE is the standby PE, and the first PE and the second PE are both added to the multicast tree corresponding to the first VPLS, the second PE is in the secondary PE. After receiving the encapsulated packet, the multicast tree corresponding to the first VPLS discards the encapsulated packet;

 and / or

After the first PE fails, the second PE switches to the primary PE, and if the second PE receives the encapsulated packet by using the multicast tree corresponding to the first VPLS, The second PE decapsulates the encapsulated packet and forwards the decapsulated packet to the first CE.

The method according to claim 5, 6 or 7, wherein, before the third PE receives the Ethernet frame sent from the second CE, the method further includes:

 The third PE receives the route reachability information corresponding to the IP address of the virtual PE that is sent by the PE or the member PE of the virtual PE through the routing protocol;

 The third PE receives the reachability information of the VPLS instance corresponding to the RG that is advertised by the first PE that is the primary PE, and the reachability information of the VPLS instance includes the VPLS instance corresponding to the RG. VPLS identifier, the virtual PE is a VPLS label allocated by the VPLS instance and an IP address of the virtual PE.

A first carrier edge device, wherein the first carrier edge device PE belongs to a redundancy group RG, the RG further includes a second PE, and the RG corresponds to a virtual PE, The virtual PE has an IP address, the virtual PE is a member PE of the first VPLS instance, the first VPLS instance is an intermediate system to an intermediate system ISIS VPLS instance, and the first user edge device CE is the virtual PE at the first a local CE in a VPLS instance, where the first CE is connected to the first PE and the second PE in an inter-frame link aggregation group MC-LAG manner, where the first PE includes:

 a first receiving module, configured to receive an Ethernet frame from the first CE,

 a first sending module, configured to encapsulate the encapsulated packet according to the type of the Ethernet frame, and encapsulate the encapsulated packet to the virtual PE in the first VPLS instance. The remote PE sends the encapsulated source address of the encapsulated >3⁄4 text as the IP address of the virtual PE.

10. The first carrier edge device of claim 9, wherein the first transmitting module comprises one or more of the following:

 a first sending unit, configured to use the IP address of the virtual PE as the tunnel source address and the IP address of the destination remote PE as the tunnel destination address when the Ethernet frame is a known destination unicast Ethernet frame Encapsulating the Ethernet frame, and sending the encapsulated packet to the destination remote PE, where the destination remote PE is the destination MAC address of the Ethernet frame in the MAC forwarding table of the first VPLS instance. Corresponding next hop PE;

a second sending unit, configured to: unicast an Ethernet frame, a multicast ethernet when the Ethernet frame is an unknown destination If the first VPLS instance uses the ingress replication mode to copy the Ethernet frame, the number of the remote PEs in the first VPLS instance is the same as the number of the remote PEs in the first VPLS instance. And the plurality of the Ethernet frames are encapsulated to obtain the plurality of encapsulated packets, wherein the tunnel source address of each of the encapsulated packets is the IP of the virtual PE. Address, the tunnel destination address of each of the encapsulated texts is an IP address of a PE in the remote PE, and the tunnel destination addresses of the plurality of encapsulated 4 艮 texts are different from each other; Transmitting the plurality of encapsulated packets by using a tunnel destination address of the plurality of encapsulated packets;

 a third sending unit, configured to: when the Ethernet frame is an unknown destination unicast Ethernet frame, a multicast Ethernet frame, or a broadcast Ethernet frame, if the first VPLS instance uses a non-aggregated multicast tree mode, use the The IP address of the virtual PE is used as the tunnel source address and the multicast IP address of the multicast tree corresponding to the first VPLS instance is used as the tunnel destination address to encapsulate the Ethernet frame and join the multicast tree. The PE other than the first PE sends the encapsulated packet;

 a fourth sending unit, configured to: when the Ethernet frame is an unknown destination unicast Ethernet frame, a multicast Ethernet frame, or a broadcast Ethernet frame, if the first VPLS instance uses an aggregate multicast tree mode, use the virtual The IP address of the PE is used as the tunnel source address and the multicast IP address of the multicast tree corresponding to the first VPLS instance is used as the tunnel destination address to encapsulate the Ethernet frame, and the encapsulated packet carries the first packet. a VPLS label, the first VPLS label is a VPLS label allocated by the virtual PE to the first VPLS instance, and the label type of the first VPLS label is an upstream allocation label, and is added to the multicast tree. The PE other than the first PE sends the encapsulated packet.

The first carrier edge device according to claim 10, wherein the third sending unit is further configured to send the Ethernet frame to the second PE, so that the second PE uses Encapsulating the Ethernet frame by using the IP address of the virtual PE as the tunnel source address and the multicast IP address of the multicast tree corresponding to the first VPLS instance as the tunnel destination address, and corresponding to the first VPLS instance Transmitting, by the multicast tree, a packet encapsulated by the second PE to a PE other than the second PE that is added to the multicast tree;

The fourth sending unit is further configured to send the Ethernet frame to the second PE, so that the second PE uses an IP address of the virtual PE as the tunnel source address and uses the first The multicast IP address of the multicast tree corresponding to the VPLS instance encapsulates the Ethernet frame as the tunnel destination address, and the packet encapsulated by the second PE carries the first VPLS label, and the first VPLS label The VPLS label that is allocated to the first VPLS instance by the virtual PE, the label type of the first VPLS label is an upstream allocation label, and the multicast tree corresponding to the first VPLS instance is added to the A PE other than the second PE in the multicast tree sends the packet encapsulated by the second PE.

The first carrier edge device according to claim 9, wherein the first PE comprises:

 a first acquiring module, configured to acquire RG information, where the RG information includes an IP address of the virtual PE corresponding to the RG, a VPLS identifier of the first VPLS instance corresponding to the RG, and the virtual PE is a The VPLS label allocated by the first VPLS instance and/or the IP address of other member PEs belonging to the RG;

 a publishing module, configured to advertise a route reachability information corresponding to an IP address of the virtual PE by using a routing protocol;

 The issuing module is further configured to: after determining that the first PE is a primary PE, advertise the reachability information of the first VPLS instance, where the reachability information of the first VPLS instance includes the RG The VPLS identifier of the VPLS instance and the VPLS label assigned by the virtual PE to the VPLS instance and the IP address of the virtual PE.

A third carrier edge device, wherein the third carrier edge device PE is used in a network, the network further includes a virtual PE, and the member PE of the RG corresponding to the virtual PE includes The first PE and the second PE, the virtual PE has an IP address, and the first VPLS instance is an intermediate system to an intermediate system ISIS VPLS instance, and the virtual PE and the third PE are both the first VPLS instance. a member PE, the first PE is a member PE of the RG that is the shortest route from the third PE, and the second CE is a local CE of the third PE in the first VPLS instance, and the third PE includes :

a second receiving module, configured to receive an Ethernet frame sent from the second CE, and a second sending module, configured to use an IP address of the virtual PE or a multicast IP of a multicast tree corresponding to the first VPLS instance The address is used as the tunnel destination address to encapsulate the Ethernet frame, and the encapsulated packet is sent to the virtual PE, so that the member PE of the RG corresponding to the virtual PE receives the encapsulated packet and performs solution. Encapsulating, and forwarding the decapsulated packet to the first CE, where the first CE is a local CE of the virtual PE in the first VPLS instance, and the A CE is connected to the first PE and the second PE in an inter-frame link aggregation group MC-LAG manner.

The third carrier edge device according to claim 13, wherein the second sending module comprises at least one of the following units:

 a fifth sending unit, configured to: when the Ethernet frame is a known destination unicast Ethernet frame, if the next hop PE corresponding to the destination MAC address of the Ethernet frame is the virtual PE, use the virtual PE The IP address is used as the tunnel destination address and the IP address of the edge device of the third carrier is used as the tunnel source address to encapsulate the Ethernet frame, and the encapsulated packet carries the first VPLS label, where the first VPLS label is The virtual PE is a VPLS label that is allocated to the first VPLS instance, and the label type of the first VPLS label is a downstream allocation label, and the encapsulated packet is sent to the first PE, so that the first Receiving, by the PE, the encapsulated packet and decapsulating the packet, and forwarding the decapsulated packet to the first CE;

 a sixth sending unit, configured to: when the Ethernet frame is an unknown destination unicast Ethernet frame, a multicast Ethernet frame, or a broadcast Ethernet frame, if the first VPLS instance uses an ingress copy mode, the Ethernet frame is used Performing the copying to obtain a plurality of the Ethernet frames having the same number of remote PEs as the third PE in the first VPLS instance; and encapsulating the plurality of Ethernet frames to obtain a plurality of encapsulated a packet, where the tunnel source address of each encapsulated packet is an IP address of the third PE, and the tunnel destination address of each encapsulated packet is the third PE The IP address of a PE in the remote PE, and the tunnel destination addresses of the plurality of encapsulated packets are different from each other; the plurality of encapsulated packets are sent to the corresponding remote PE, so that the The first PE receives the encapsulated packet whose tunnel destination address is the IP address of the virtual PE, decapsulates the packet, and forwards the decapsulated packet to the first CE, where the tunnel destination address is Encapsulation of the IP address of the virtual PE The packet carries the first VPLS label, and the first VPLS label is a VPLS label that is allocated by the virtual PE to the first VPLS instance, and the label type of the first VPLS label is a downstream label.

a seventh sending unit, configured to: the first PE is a primary PE, the second PE is a standby PE, the first PE is added to a multicast tree corresponding to the first VPLS, and the Ethernet frame is When unicasting an Ethernet frame, a multicast Ethernet frame, or a broadcast Ethernet frame for an unknown destination, if the first VPLS instance uses the non-aggregated multicast tree mode, the multicast IP address of the multicast tree corresponding to the first VPLS instance is used. As the tunnel destination address and using the IP address of the third PE as the tunnel source address to the ether The frame is encapsulated, and the encapsulated packet is forwarded by the multicast tree corresponding to the first VPLS instance, so that the first PE receives the encapsulated packet and decapsulates the packet, and decapsulates the packet. The message is forwarded to the first CE;

 An eighth sending unit, configured to: when the first PE is a primary PE, the second PE is a standby PE, the first PE is added to a multicast tree corresponding to the first VPLS, and the ether is If the frame is an unknown destination unicast Ethernet frame, a multicast Ethernet frame, or a broadcast Ethernet frame, if the first VPLS instance uses the aggregate multicast tree mode, the multicast of the multicast tree corresponding to the first VPLS instance is used. The IP address is used as the tunnel destination address, and the IP address of the third PE is used as the tunnel source address to encapsulate the Ethernet frame. The encapsulated packet carries a third VPLS label, and the third VPLS label is the third. The PE is the VPLS label allocated by the first VPLS instance, and the label type of the third VPLS label is an upstream allocation label, and the encapsulated packet is forwarded through the multicast tree, so that the first PE receiving station The encapsulated packet is decapsulated and the decapsulated packet is forwarded to the first CE.

The third carrier edge device according to claim 12, 13 or 14, wherein the third carrier edge device further comprises:

 a third receiving module, configured to receive route reachability information corresponding to an IP address of the virtual PE that is advertised by the PE in the virtual PE by the routing protocol;

 The third receiving module is further configured to receive reachability information of the VPLS instance corresponding to the RG that is advertised by the first PE that is the primary PE, where the reachability information of the VPLS instance includes the RG The VPLS identifier of the corresponding VPLS instance, the virtual PE is the VPLS label allocated by the VPLS instance and the IP address of the virtual PE.