CN110050445B

CN110050445B - Method, device and system for sending and receiving message

Info

Publication number: CN110050445B
Application number: CN201680091456.0A
Authority: CN
Inventors: 潘望; 于勇; 张彦兵; 冯锦良; 许志军; 王歆平
Original assignee: Huawei Technologies Co Ltd
Current assignee: Huawei Technologies Co Ltd
Priority date: 2016-12-09
Filing date: 2016-12-09
Publication date: 2021-01-05
Anticipated expiration: 2036-12-09
Also published as: CN110050445A; WO2018103099A1

Abstract

The application provides a method, a device and a system for sending and receiving messages. The method is applied to EVPN, the user side equipment is respectively accessed to the scenes of at least three PE devices through at least three links, the at least three links form a group of links, the group of links are link sections, and the at least three PE devices comprise first PE devices. The method comprises the following steps: when the redundant mode of the link segment is a multi-active mode, the first PE device generates and sends a first mode notification message to the remote PE device, where the first mode notification message includes the multi-active mode and an identifier of the link segment, and the identifier of the link segment is used to uniquely identify the link segment. The first mode advertisement message is used for advertising the far-end PE device, and a part of links in the link segment can be used for forwarding data streams, and the number of the part of links is greater than 1 and less than the maximum number of links in the link segment. The far-end PE device receives a first mode notification message sent by the first PE device.

Description

Method, device and system for sending and receiving message

Technical Field

The present invention relates to the field of communications, and in particular, to a method, device, and system for sending and receiving a message.

Background

An Ethernet Virtual Private Network (EVPN) is a Virtual Private Network (VPN) technology that provides two-layer Network interconnection over a Multi-Protocol Label Switching (MPLS) Network. EVPN is currently used in network design of large operators as a mainstream solution for two-layer traffic bearer. The EVPN technology uses a Border Gateway Protocol (BGP) as a Protocol of a Control plane, performs Media Access Control (MAC) learning between Provider Edge (PE) devices, and transfers a MAC address learning and issuing process from a conventional data plane to the Control plane, thereby greatly reducing MAC address diffusion in a traffic flooding manner, supporting multi-homing EVPN Access of a user end device, and facilitating management of MAC addresses to implement load sharing.

The user-side device is usually connected to the PE device by using an access method such as an ethernet link (e.g., Virtual Local Area Network (VLAN)), so as to access the EVPN. As can be seen from the above, one of the important advantages of the EVPN scheme is that multi-homing access of the user side device is achieved. However, in the current access EVPN technology such as ethernet link, the notification of the multi-active redundancy mode cannot be supported between PE devices, and thus the advantages of the EVPN technology cannot be effectively exerted.

Disclosure of Invention

The method, the device and the system for sending and receiving the message solve the problems that in the multi-homing access of the user side equipment to the EVPN, a multi-active redundancy mode is not supported, so that the advantages of the EVPN technology cannot be effectively exerted, and multi-active single standby or multi-active multi-standby cannot be supported.

In order to solve the above problem, a first aspect of the embodiments of the present invention provides a method for sending a message, where the method is applied to an ethernet virtual private network EVPN, a user-side device is respectively connected to at least three provider edge PE devices through at least three links, where the at least three links form a group of links, the group of links is a link segment, and the at least three PE devices include a first PE device, where the method includes: the first PE equipment acquires a redundancy mode of the link section; when the redundancy mode of the link segment is a multi-active mode, the first PE device sends a mode advertisement message to a remote PE device, where the mode advertisement message includes information indicating that the redundancy mode of the link segment is the multi-active mode and an identifier of the link segment, where the identifier of the link segment is used to uniquely identify the link segment, and the multi-active mode indicates that a part of links in the link segment can be used to forward a data stream, and the number of the part of links is greater than 1 and less than the maximum number of links in the link segment.

The multi-active mode is notified to the far-end PE equipment by the at least three PE equipment, and the embodiment of the invention provides a method for supporting multi-active redundancy mode notification in a multi-homing access EVPN scene, thereby effectively exerting the advantages of the EVPN technology.

In a first possible implementation manner of the first aspect, the method further includes: and the first PE device sends a destination MAC address and a next hop network address to the far-end PE device, wherein the destination MAC address is the MAC address of a terminal device accessed to the user side device, and the next hop network address is the network address of the first PE device.

In a second possible implementation manner of the first aspect, the method further includes: the first PE device sends an MAC routing message to the far-end PE device, wherein the MAC routing message comprises a destination MAC address and a next hop network address, the destination MAC address is an MAC address of a terminal device accessed to the user side device, and the next hop network address is a network address of the first PE device.

The embodiment of the invention provides a method for sharing multiple active loads, which not only provides transmission capability with larger bandwidth through load sharing, but also improves the reliability of load sharing and forwarding by utilizing backup protection.

In a third possible implementation manner of the first aspect, when the redundancy mode of the link segment is a multi-active mode, the first PE device obtains a state of a first link between the user side device and the first PE device, where the state of the first link is active or inactive; correspondingly, the mode advertisement message further includes a state of the first link and a next hop network address, where the next hop network address is a network address of the first PE device; the first PE device further sends an MAC routing message to the far-end PE device, wherein the MAC routing message comprises a destination MAC address, and the destination MAC address is an MAC address of a terminal device accessed to the user-side device.

The embodiment of the invention provides a method for sharing multiple active loads, which can realize load sharing more quickly and make the load sharing more balanced.

With reference to the third possible implementation manner of the first aspect, in a fourth possible implementation manner of the first aspect, the link is an ethernet link, the link section is an ethernet segment ES, a redundancy mode of the link section is a redundancy mode of the ES, and the first link is a first ethernet link; correspondingly, when the state of the first ethernet link is active, the mode advertisement message is used to advertise the remote PE device, and a part of ethernet links in the ES may be used to forward data streams, where the first ethernet link may be used to forward data streams; or when the state of the first ethernet link is inactive, the mode advertisement message is used to advertise the remote PE device, and a part of ethernet links in the ES may be used to forward data streams, where the first ethernet link may not be used to forward data streams.

The at least three PE devices notify the far-end PE device of the multi-active mode and the state of the Ethernet link, and the embodiment of the invention provides a method for supporting the multi-active redundancy mode in an Ethernet link multi-homing EVPN scene, thereby supporting multi-active single-standby or multi-active multi-standby, providing the Ethernet link transmission capability with larger bandwidth through load sharing, and improving the reliability of load sharing and forwarding by utilizing backup protection. The method can realize the load sharing of the Ethernet link relatively quickly, thereby ensuring that the effect of the load sharing of the Ethernet link is more balanced.

With reference to the first aspect or any one of the first to the second possible implementation manners of the first aspect, in a fifth possible implementation manner of the first aspect, the mode advertisement message is a BGP Update message, where the BGP Update message carries an Ethernet auto-discovery Ethernet a-D route, and the Ethernet a-D route includes an Ethernet segment identifier ESI field; the BGP Update message also carries ESI tag extended community attributes, the ESI tag extended community attributes comprise Flag fields, and the Flag fields are used for bearing the redundancy mode of the ES.

By carrying the redundancy mode in the BGP message, the embodiment of the invention can effectively utilize the existing protocol to realize the notification of the redundancy mode of the Ethernet link.

With reference to the first aspect or the first or third or fourth possible implementation manner of the first aspect, in a sixth possible implementation manner of the first aspect, the mode advertisement message is a BGP Update message, where the BGP Update message carries an Ethernet auto discovery Ethernet a-D route, and the Ethernet a-D route includes an Ethernet segment identifier ESI field; the BGP Update message further carries an ESI tag extended community attribute, where the ESI tag extended community attribute includes a Flag field, and the Flag field is used to carry a redundancy mode of the ES and a state of the first ethernet link.

By carrying the redundancy mode in the BGP message, the embodiment of the invention can effectively utilize the existing protocol to realize the Ethernet link redundancy mode and the notice of the link state.

With reference to the third possible implementation manner of the first aspect, in a seventh possible implementation manner of the first aspect, the link is a pseudo wire PW, the link segment is a pseudo wire segment PWs, a redundancy mode of the link segment is a redundancy mode of the PWs, and the first link is a first PW; accordingly, when the state of the first PW is active, the mode advertisement message is used to advertise the far-end PE device, and a part of PWs in the PWs may be used to forward data flows, where the first PW may be used to forward data flows; or when the state of the first PW is inactive, the mode advertisement message is used to advertise the far-end PE device, and a part of PWs in the PWs may be used to forward a data flow, where the first PW may not be used to forward a data flow.

The embodiment of the invention provides a method for supporting a multi-active redundancy mode in a PW multi-homing access EVPN scene, thereby supporting multi-active single-standby or multi-active multi-standby, providing PW transmission capability with larger bandwidth through load sharing, and improving the reliability of load sharing and forwarding by utilizing backup protection. The method can realize PW load sharing more quickly, thereby ensuring that the PW load sharing effect is more balanced.

In an eighth possible implementation manner of the first aspect, the link is a pseudo wire PW, the link segment is a pseudo wire segment PWs, and a redundancy mode of the link segment is a redundancy mode of the PWs; when the redundancy mode of the PWS is a full-active mode, the first PE device sends a mode notification message to a remote PE device, wherein the mode notification message comprises the full-active mode and an identifier of the link segment, the identifier of the link segment is used for uniquely identifying the link segment, and the full-active mode indicates that all PWs in the PWS can be used for forwarding data streams.

The full-active mode is notified to the far-end PE equipment through the at least three PE equipment, and the embodiment of the invention provides a method for supporting the full-active redundancy mode in a PW multi-homing access EVPN scene, thereby providing PW transmission capability with larger bandwidth through load sharing.

With reference to the seventh possible implementation manner of the first aspect, in a ninth possible implementation manner of the first aspect, the mode advertisement message further includes a state of the first PW, where the state of the first PW is active.

The at least three PE devices inform the far-end PE device of the full-active mode and the state of the PW, and the embodiment of the invention provides a method for supporting the full-active redundancy mode in a PW multi-homing access EVPN scene, thereby providing PW transmission capability with larger bandwidth through load sharing, and realizing load sharing more quickly so that the effect of load sharing is more balanced.

In a tenth possible implementation manner of the first aspect, the link is a pseudo wire PW, the link segment is a pseudo wire segment PWs, and a redundancy mode of the link segment is a redundancy mode of the PWs; when the redundancy mode of the PWS is a single active mode, the first PE device sends a mode notification message to a remote PE device, wherein the mode notification message comprises the single active mode and an identifier of the link segment, the identifier of the link segment is used for uniquely identifying the link segment, and the single active mode indicates that only one PW in the PWS can be used for forwarding data streams.

The single-active mode is notified to the far-end PE device by the at least three PE devices, and the embodiment of the invention provides a method for supporting the single-active redundancy mode in a PW multi-homing access EVPN scene, so that single-active single standby or single-active multi-standby of the PW can be supported.

With reference to the eleventh possible implementation manner of the first aspect, in a tenth possible implementation manner of the first aspect, the first PE device further obtains a state of the first PW, where the state of the first PW is active or inactive; the mode advertisement message further includes a status of the first PW; when the state of the first PW is active, the mode advertisement message is used to advertise the remote PE device, and only one PW in the PWs may be used to forward a data stream, where the first PW may be used to forward a data stream; or when the state of the first PW is inactive, the mode advertisement message is used to advertise the remote PE device, and only one PW in the PWs may be used to forward a data stream, where the first PW may not be used to forward a data stream.

The at least three PE devices notify the far-end PE device of the single active mode and the state of the PW, and the embodiment of the invention provides a method for supporting the single active redundancy mode in a PW multi-homing access EVPN scene, so that single active single standby or single active multi standby of the PW can be supported.

With reference to any one of the seventh to eleventh possible implementations of the first aspect, in an eleventh possible implementation of the first aspect, the mode advertisement message is a BGP Update message, where the BGP Update message carries an Ethernet auto discovery Ethernet a-D route, and the Ethernet a-D route includes a field indicating PWSI, where the field indicating PWSI is an Ethernet segment identity ESI field or a newly added field; the BGP Update message further carries an ESI tag extended community attribute, where the ESI tag extended community attribute includes a Flag field, and the Flag field is used to carry the redundancy mode of the PWS or the redundancy mode of the PWS and the state of the first PW.

With reference to any one of the seventh to eleventh possible implementations of the first aspect, in a twelfth possible implementation of the first aspect, the mode advertisement message is a BGP Update message, where the BGP Update message carries an Ethernet auto discovery Ethernet a-D route, and the Ethernet a-D route includes a field indicating PWSI, and the field indicating PWSI is an Ethernet segment identity ESI field or a newly added field; the BGP Update message also carries ESI tag extended community attributes, the ESI tag extended community attributes comprise Flag fields, and the Flag fields are used for bearing the redundancy mode of the PWS and the state of the first PW.

By carrying the redundancy mode in the BGP message, the embodiment of the invention can effectively utilize the existing protocol to realize the PW redundancy mode and the notification of the PW state.

A second aspect of the embodiments of the present invention provides a method for receiving a message, which is applied to an ethernet virtual private network EVPN, where a user-side device is respectively accessed to a scenario of at least three provider edge PE devices through at least three links, where the at least three links form a group of links, the group of links is a link segment, the at least three PE devices include a first PE device, and the method includes: a far-end PE device receives a first mode notification message sent by the first PE device, wherein the first mode notification message comprises a redundant mode of the link segment and an identifier of the link segment, the identifier of the link segment is used for uniquely identifying the link segment, and the redundant mode of the link segment is a multi-active mode; and the remote equipment PE acquires the identifier of the link section, and when the redundancy mode of the link section is a multi-active mode, confirms that a part of links in the link section identified by the identifier of the link section can be used for forwarding data streams according to the multi-active mode, wherein the number of the part of links is more than 1 and less than the maximum number of links in the link section.

The embodiment of the invention provides a method for supporting multi-active redundancy mode notification in a multi-homing access EVPN scene by the far-end PE equipment receiving the multi-active modes notified by the at least three PE equipment, thereby effectively exerting the advantages of the EVPN technology.

In a first possible implementation manner of the second aspect, the at least three PE devices further include a second PE device, and the method further includes: when the state of a first link between the first PE device and the user side device and the state of a second link between the second PE device and the user side device are both active, the remote PE device receives a first MAC routing message and a second MAC routing message sent from the first PE device and the second PE device, respectively, where the first MAC routing message includes a destination MAC address and a first next hop network address, and the second MAC routing message includes the destination MAC address and a second next hop network address; the destination MAC address is an MAC address of a terminal device accessing the user side device, the first next hop network address is a network address of the first PE device, and the second next hop network address is a network address of the second PE device; the far-end PE device determines, according to the multi-active mode in the first mode advertisement message, the identifier of the link segment, the first MAC routing message, and the second MAC routing message, that the first link and the second link can forward the data stream destined for the destination MAC in a load-sharing manner in the link segment.

With reference to the first possible implementation manner of the second aspect, in a second possible implementation manner of the second aspect, the determining, by the far-end PE device according to the multi-active mode in the first mode advertisement message, the identifier of the link segment, the first MAC routing message, and the second MAC routing message, that a data flow to the destination MAC can be forwarded in a load-sharing manner by the first link and the second link in the link segment includes: the far-end PE device generates an MAC forwarding table according to the multi-active mode, the identifier of the link segment, the destination MAC address, the first next hop network address, and the second next hop network address in the first mode notification message, where the MAC forwarding table includes the destination MAC address and an outgoing interface list, the outgoing interface list includes a first outgoing interface of the far-end PE device and a second outgoing interface of the far-end device, and the far-end PE device forwards a data flow load destined for the destination MAC from the first outgoing interface and the second outgoing interface; the first output interface is obtained according to the first next hop network address, and the second output interface is obtained according to the second next hop network address.

In a third possible implementation manner of the second aspect, the method further includes: the first mode advertisement message further comprises a state of a first link, the state of the first link being active or inactive; the far-end PE device determines whether the first link between the first PE device and the user side device can be used for forwarding data flow according to the multi-active mode, the identification of the link section and the state of the first link.

The embodiment of the invention provides a method for sharing multiple active loads, which can realize load sharing more quickly and make the load sharing effect more balanced by the far-end PE equipment receiving the corresponding link states announced by the at least three PE equipment.

With reference to the third possible implementation manner of the second aspect, in a fourth possible implementation manner of the second aspect, the at least three PE devices further include a second PE device, and the method further includes: the far-end PE equipment receives a second mode notification message sent by the second PE equipment, wherein the second mode notification message comprises that the redundant mode of the link section is a multi-active mode, the state of the second link is active and the link section identifier; the far-end PE device determines that the second link between the second PE device and the user-side device can be used for forwarding data flow according to the multi-active mode, the identification of the link segment and the state of the second link in the second mode notification message; when the state of the first link is active, the far-end PE device determines that the first link and the second link can carry out load sharing forwarding of data flow in the link section.

With reference to the fourth possible implementation manner of the second aspect, in a fifth possible implementation manner of the second aspect, when the state of the first link is active, the determining, by the far-end PE device, that the first link and the second link can forward the data stream in load sharing in the link segment includes: the first mode advertisement message further includes a first next hop network address, where the first next hop network address is a network address of the first PE device; the second mode advertisement message further includes a second next hop network address, where the second next hop network address is a network address of the second PE device; the far-end PE device generates a mapping relationship between link information and next hop network address information according to the multi-active mode, the link segment identifier, the state of the first link being active, the state of the second link being active, the first next hop network address and the second next hop network address, where the mapping relationship indicates a forwarding data stream that the first link and the second link in the link segment identified by the link segment identifier can share load, the data stream reaches the first link through the first PE device identified by the first next hop network address, and the data stream reaches the second link through the second PE device identified by the second next hop network address; correspondingly, the remote PE device further receives an MAC routing message from the first PE device, where the MAC routing message includes a destination MAC address, and the destination MAC address is an MAC address of a terminal device accessing the user side device; the remote PE device further generates an MAC forwarding table according to the destination MAC address and the mapping relationship, where the MAC forwarding table includes the destination MAC address and an egress interface list, the egress interface list includes a first egress interface of the remote PE device and a second egress interface of the remote PE device, and the remote PE device forwards a data stream load destined for the destination MAC from the first egress interface and the second egress interface in a load sharing manner; the first output interface is obtained according to the first next hop network address, and the second output interface is obtained according to the second next hop network address.

With reference to any one of the third to fourth possible implementation manners of the second aspect, in a sixth possible implementation manner of the second aspect, the link is an ethernet link, the link section is an ethernet section ES, a redundancy mode of the link section is a redundancy mode of the ES, and the first link is a first ethernet link; accordingly, the determining, by the far-end PE device, whether the first link between the first PE device and the user-side device can be used for forwarding a data flow according to the multi-active mode, the identifier of the link segment, and the state of the first link includes: when the state of the first ethernet link is active, the far-end PE device determines that the first ethernet link is an active link in the ES, and may be used to forward a data stream; or when the state of the first ethernet link is inactive, the far-end PE device determines that the first ethernet link is not available for forwarding a data stream as an inactive link in the ES.

The embodiment of the invention provides a method for supporting a multi-active redundancy mode in an Ethernet link multi-homing access EVPN scene by receiving the multi-active mode and the state of the Ethernet link announced by the at least three PE devices through the far-end PE device, thereby supporting multi-active single-standby or multi-active multi-standby, not only providing the Ethernet link transmission capability with larger bandwidth through load sharing, but also improving the reliability of load sharing and forwarding by utilizing backup protection. The method can realize the load sharing of the Ethernet link relatively quickly, thereby ensuring that the effect of the load sharing of the Ethernet link is more balanced.

With reference to any one of the first to sixth possible implementation manners of the second aspect, in a seventh possible implementation manner of the second aspect, the first mode advertisement message is a BGP Update message, where the BGP Update message carries an Ethernet auto discovery Ethernet a-D route, and the Ethernet a-D route includes an Ethernet segment identifier ESI field; the BGP Update message further carries an ESI tag extended community attribute, where the ESI tag extended community attribute includes a Flag field, and the Flag field is used to carry the redundancy mode of the ES or the redundancy mode of the ES and the state of the first ethernet link.

By carrying the redundancy mode in the BGP message, the embodiment of the present invention can effectively utilize the existing protocol to implement the notification of the ethernet link redundancy mode or the notification of the ethernet link redundancy mode and the ethernet link status.

With reference to any one of the third to fourth possible implementation manners of the second aspect, in an eighth possible implementation manner of the second aspect, the link is a pseudo wire PW, the link segment is a pseudo wire segment PWs, a redundancy mode of the link segment is a redundancy mode of the PWs, and the first link is a first PW; accordingly, the determining, by the far-end PE device, whether the first link between the first PE device and the user-side device can be used for forwarding a data flow according to the multi-active mode, the identifier of the link segment, and the state of the first link includes: when the state of the first PW is active, the far-end PE device determines that the first PW is an active link in the PWs and can be used to forward a data stream; or when the state of the first PW is inactive, the far-end PE device determines that the first PW is not available for forwarding a data stream as an inactive link in the PWs.

The embodiment of the invention provides a method for supporting a multi-active redundancy mode in a PW multi-homing access EVPN scene, thereby supporting multi-active single-standby or multi-active multi-standby, not only providing PW transmission capability with larger bandwidth through load sharing, but also improving the reliability of load sharing and forwarding by utilizing backup protection. The method can realize PW load sharing more quickly, thereby ensuring that the PW load sharing effect is more balanced.

In a ninth possible implementation manner of the second aspect, the link is a pseudo wire PW, the link segment is a pseudo wire segment PWs, the redundancy mode of the link segment is a redundancy mode of the PWs, and the identifier of the link is an identifier of the PWs; and when the redundancy mode of the link section is the full-active mode, the far-end PE confirms that all PWs identified by the PWS identification can be used for forwarding data streams according to the full-active mode and the PWS identification.

The embodiment of the invention provides a method for supporting an all-active redundancy mode in a PW multi-homing access EVPN scene by the far-end PE equipment receiving the all-active mode announced by the at least three PE equipment, thereby providing a PW transmission capability with larger bandwidth through load sharing.

With reference to the third possible implementation manner of the second aspect, in a tenth possible implementation manner of the second aspect, the link is a pseudo wire PW, the link segment is a pseudo wire segment PWs, a redundancy mode of the link segment is a redundancy mode of the PWs, an identifier of the link is an identifier of the PWs, the first link is a first PW, and a state of the first PW is active; and when the redundancy mode of the link section is the full-active mode, the far-end PE confirms that all PWs in the PWS can be used for forwarding data streams according to the full-active mode and the PWS identifier.

The embodiment of the invention provides a method for supporting the full-active redundancy mode in a PW multi-homing access EVPN scene by the far-end PE equipment receiving the full-active mode and the state of the PW notified by the at least three PE equipment, thereby providing PW transmission capability with larger bandwidth through load sharing, realizing load sharing more quickly and balancing the effect.

In an eleventh possible implementation manner of the second aspect, the link is a pseudo wire PW, the link segment is a pseudo wire segment PWs, the redundancy mode of the link segment is a redundancy mode of the PWs, and the identifier of the link is an identifier of the PWs; and when the redundancy mode of the link section is a single-active mode, the far-end PE confirms that only one PW in the PWS identified by the PWS identification can be used for forwarding data streams according to the single-active mode and the PWS identification.

The embodiment of the invention provides a method for supporting a single active redundancy mode in a PW multi-homing access EVPN scene by the far-end PE equipment receiving the single active mode announced by the at least three PE equipment, thereby supporting single active single standby or single active multiple standby of the PW.

With reference to the third possible implementation manner of the second aspect, in a twelfth possible implementation manner of the second aspect, the link is a pseudo wire PW, the link segment is a pseudo wire segment PWs, a redundancy mode of the link segment is a redundancy mode of the PWs, an identifier of the link is an identifier of the PWs, the first link is a first PW, and a state of the first PW is active or inactive; and when the redundant mode of the link segment is a single active mode and the state of the first PW is active, the far-end PE confirms that the first PW in the PWS identified by the PWS identification can be used for forwarding data flow according to the single active mode, the state of the first PW is active and the PWS identification.

With reference to the first, ninth, or eleventh implementation manner of the second aspect, in a thirteenth possible implementation manner of the second aspect, the mode advertisement message is a BGP Update message, where the BGP Update message carries an Ethernet auto-discovery Ethernet a-D route, and the Ethernet a-D route includes a field indicating PWSI, where the field indicating PWSI is an Ethernet segment identity ESI field or a newly added field; the BGP Update message also carries ESI tag extended community attributes, the ESI tag extended community attributes comprise Flag fields, and the Flag fields are used for bearing the redundancy mode of the PWS.

By carrying the redundancy mode in the BGP message, the embodiment of the invention can effectively utilize the existing protocol to realize the notification of the PW redundancy mode.

With reference to the third or fourth or eighth or ninth or twelfth implementation manner of the second aspect, in a fourteenth possible implementation manner of the second aspect, the mode advertisement message is a border gateway protocol BGP Update message, where the BGP Update message carries Ethernet auto discovery Ethernet a-D routes, and the Ethernet a-D routes include a field indicating PWSI, where the field indicating PWSI is an Ethernet segment identity ESI field or a newly added field; the BGP Update message also carries ESI tag extended community attributes, the ESI tag extended community attributes comprise Flag fields, and the Flag fields are used for bearing the redundancy mode of the PWS and the state of the first PW.

In a third aspect, the present invention provides a first PE device, where the first PE device is configured to perform the method in the first aspect or any possible implementation manner of the first aspect. In particular, the first PE device comprises means for performing the first aspect or the method in any possible implementation manner of the first aspect.

In a fourth aspect, the present invention provides a first PE device, where the first PE device includes: a processor, a transmitter, a random access memory, a read only memory, and a bus. The processor is coupled with the transmitter, the random access memory and the read-only memory through the buses respectively. When the first PE device needs to be operated, the first PE device is guided to enter a normal operation state by starting a basic input and output system solidified in a read only memory or a bootloader guiding system in an embedded system. After the first PE device enters the normal operation state, the application and the operating system are run in the random access memory, so that the processor performs the method of the first aspect or any possible implementation manner of the first aspect.

In a fifth aspect, the present invention provides a computer-readable medium for storing a computer program comprising instructions for performing the method of the first aspect or any possible implementation manner of the first aspect.

In a sixth aspect, a first PE device is provided, where the first PE device includes: the main control board and the interface board, further, can also include the exchange network board. The first PE device is configured to perform the first aspect or the method in any possible implementation manner of the first aspect. In particular, the first PE device comprises means for performing the first aspect or the method in any possible implementation manner of the first aspect.

A seventh aspect provides a first PE device, where the first PE device includes a controller and a first PE forwarding device. The first PE forwarding device includes: the interface board further can also comprise a switching network board. The first PE device is configured to execute a function of the interface board in the sixth aspect, and further, may execute a function of the switch network board in the sixth aspect. The controller includes a receiver, a processor, a transmitter, a random access memory, a read only memory, and a bus. The processor is coupled to the receiver, the transmitter, the random access memory and the read only memory through the bus respectively. When the controller needs to be operated, the basic input and output system solidified in the read-only memory or the bootloader guide system in the embedded system is started, and the guide controller enters a normal operation state. After the controller enters the normal operation state, the application program and the operating system are operated in the random access memory, so that the processor executes the functions of the main control board in the sixth aspect.

In an eighth aspect, the present invention provides a far-end PE device, where the far-end PE device is configured to execute the method in the second aspect or any possible implementation manner of the second aspect. In particular, the remote PE device comprises means for performing the method of the second aspect or any possible implementation of the second aspect.

In a ninth aspect, the present invention provides a remote PE device, including: a receiver, a processor, a transmitter, a random access memory, a read only memory, and a bus. The processor is coupled to the receiver, the transmitter, the random access memory and the read only memory through the bus respectively. When the remote PE equipment needs to be operated, the remote PE equipment is guided to enter a normal operation state by starting a basic input and output system solidified in a read only memory or a bootloader guiding system in an embedded system. After the remote PE device enters the normal operation state, the application program and the operating system are executed in the random access memory, so that the processor executes the method of the second aspect or any possible implementation manner of the second aspect.

In a tenth aspect, the invention provides a computer-readable medium for storing a computer program comprising instructions for performing the method of the second aspect or any possible implementation of the second aspect.

In an eleventh aspect, a remote PE device is provided, the remote PE device comprising: the main control board and the interface board, further, can also include the exchange network board. The remote PE device is configured to perform the second aspect or the method in any possible implementation manner of the second aspect. In particular, the remote PE device comprises means for performing the method of the second aspect or any possible implementation of the second aspect.

In a twelfth aspect, a far-end PE device is provided, where the far-end PE device includes a controller and a far-end PE forwarding device, where the far-end PE forwarding device includes: the interface board further can also comprise a switching network board. The far-end PE forwarding device is configured to execute a function of the interface board in the eleventh aspect, and further, may also execute a function of the switching network board in the eleventh aspect. The controller includes a receiver, a processor, a transmitter, a random access memory, a read only memory, and a bus. The processor is coupled to the receiver, the transmitter, the random access memory and the read only memory through the bus respectively. When the controller needs to be operated, the basic input and output system solidified in the read-only memory or the bootloader guide system in the embedded system is started, and the guide controller enters a normal operation state. After the controller enters the normal operation state, the application program and the operating system are operated in the random access memory, so that the processor performs the functions of the main control board in the eleventh aspect.

In a thirteenth aspect, a system is provided, which includes the first PE device of any of the third to seventh aspects and the remote PE device of any of the eighth to twelfth aspects.

Drawings

In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below. It is obvious that the following figures reflect only some embodiments of the invention, and that other embodiments of the invention can be obtained by those skilled in the art without inventive exercise. And all such embodiments or implementations are within the scope of the present invention.

Fig. 1a is a schematic network diagram of an ethernet link multi-homing EVPN according to an embodiment of the present invention;

fig. 1b is a schematic diagram of a PW multi-homed access EVPN network according to an embodiment of the present invention;

fig. 2 is a schematic flow chart of a redundant mode notification method according to an embodiment of the present invention;

fig. 3a is an interaction diagram of a redundant mode advertisement process without ethernet link status according to an embodiment of the present invention;

fig. 3b is an interaction diagram of a redundant mode advertisement flow with ethernet link status according to an embodiment of the present invention;

fig. 3c is a schematic diagram of an MP _ REACH _ NLRI field format according to an embodiment of the present invention;

fig. 3d is a schematic diagram of an EVPN NLRI field format according to an embodiment of the present invention;

FIG. 3e is a diagram illustrating the format of an Ethernet A-D routing field according to an embodiment of the present invention;

FIG. 3f is a diagram illustrating an ESI Label Extended Community field format according to an embodiment of the present invention;

fig. 4a is an interactive schematic diagram of a redundant mode advertisement process without a PW status in an embodiment of the present invention;

fig. 4b is an interactive schematic diagram of a redundant mode advertisement flow carrying a PW status in an embodiment of the present invention;

FIG. 4c is a diagram illustrating the format of the ESI field in an embodiment of the present invention;

fig. 5a is a schematic structural diagram of a first PE device according to an embodiment of the present invention;

fig. 5b is a schematic structural diagram of another first PE device and a controller according to an embodiment of the present invention;

fig. 5c is a schematic structural diagram of another first PE device according to an embodiment of the present invention;

fig. 5d is a schematic structural diagram of yet another first PE device according to an embodiment of the present invention;

fig. 6a is a schematic structural diagram of a remote PE device according to an embodiment of the present invention;

fig. 6b is a schematic structural diagram of another remote PE device and a controller according to an embodiment of the present invention;

fig. 6c is a schematic structural diagram of another remote PE device according to an embodiment of the present invention;

fig. 6d is a schematic structural diagram of another remote PE device according to an embodiment of the present invention;

fig. 7 is a schematic diagram of a network system according to an embodiment of the present invention;

Detailed Description

The technical solution in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. It is to be understood that the described embodiments are merely exemplary of the invention, and not restrictive of the full scope of the invention. All other embodiments, which can be obtained by a person skilled in the art without inventive effort based on the embodiments of the present invention, are within the scope of the present invention.

The network architecture and the service scenario described in the embodiment of the present invention are for more clearly illustrating the technical solution of the embodiment of the present invention, and do not form a limitation on the technical solution provided in the embodiment of the present invention, and it can be known by those skilled in the art that the technical solution provided in the embodiment of the present invention is also applicable to similar technical problems along with the evolution of the network architecture and the appearance of a new service scenario.

The technical scheme described by the invention can be applied to EVPN based on BGP MPLS. The EVPN technology adopts a mechanism similar to BGP/MPLS Internet Protocol (IP) VPN, and by expanding the BGP Protocol and using the expanded reachability information, the MAC address learning and publishing process between two layers of networks of different sites is transferred from a data plane to a control plane. The function of the L2VPN is realized by learning MAC addresses at the control plane. The MAC address is learned in the control layer, so that the problems that the multi-homing of the user side equipment is difficult to realize and the load sharing cannot be supported due to the MAC address learning in the data plane can be solved.

For different application requirements and different network design requirements, the deployment of the multi-homing EVPN of the user side equipment generally includes two scenarios of the ethernet link multi-homing EVPN and the Pseudo Wire (PW). The deployment that one user side device is respectively connected to a plurality of network devices through a plurality of links is called user side device multi-homing access. The user-side device may be a user Edge (CE) device or a lower PE (UPE) device in different deployment scenarios. Both scenarios are explained in detail below.

Fig. 1a shows a typical scenario of Ethernet Link multiple access EVPN, and CE1 is connected to PE1-1, PE1-2 and PE1-3 through Ethernet Links (EL) 1, EL2 and EL3, respectively. A group of Ethernet links formed by the three Ethernet links forms an Ethernet Segment (english: Ethernet Segment, ES). The Ethernet Segment Identifier (ESI) is a unique non-zero Identifier that identifies the Ethernet Segment ES. The MAC address of UE1 is flooded to PE1-1, PE1-2 and PE1-3 through CE1, so that PE1-1, PE1-2 and PE1-3 learn the MAC address of UE1 from CE1 connected to them, and PE2 is connected to CE1 through PE1-1, PE1-2 or PE 1-3. Thus PE1-1, PE1-2 and PE1-3 are near-end PEs from the perspective of being close to CE1 or from the perspective of the MAC address of the UE1 being the home or Local (English: Local) MAC address of PE1-1, PE1-2 and PE 1-3. In contrast, PE2 is a remote (English) PE device from the perspective of being far from CE1 or from the perspective of the MAC address of the UE1 being the Local MAC address of PE1-1, PE1-2, and PE 1-3. The sites (English: site)1 and site2 of the VPN1 are announced between the PE1-1, PE1-2, PE1-3 and PE2 through BGP MAC announcement routing messages, for example, the MAC address of the User Equipment (UE) 1 in site1 and the MAC address of the UE2 in site2, so that the UE1 and the UE2 in the VPN1 are intercommunicated. Moreover, interworking between the UE1 and the UE2 can be achieved through multi-homed links between the CE1 and the PE1-1, PE1-2 and PE 1-3.

However, although one of the important advantages of the EVPN scheme is that the multi-homed access of the user-side device is implemented, in the current Ethernet link access EVPN technology, messages are automatically discovered between PE devices through BGP Ethernet (english: Ethernet Auto-Discovery, Ethernet a-D), only a single-active or full-active redundancy mode can be advertised, the advertising of a multi-active redundancy mode cannot be supported, the advantages of the EVPN technology cannot be effectively exerted, and a multi-active single-standby or multi-active multi-standby cannot be supported.

In the ethernet link multi-homing deployment scenario, the single active redundancy mode (referred to as the single active mode) refers to that the state of only one ethernet link in the ethernet link segment is active, and the state of one or more other ethernet links is inactive. Active state means that the ethernet link can be used to carry, forward data streams. In the scenario of active/standby protection, it is usually used as an active ethernet link. Accordingly, the state of the ethernet link may also be inactive, which means that the ethernet link may not be used to carry, forward data streams, and is typically used as a backup ethernet link. When the active Ethernet link fails, the standby Ethernet link is switched to carry and forward the data stream. Therefore, the deployment scenario in the single active mode may include single active (only one EL in the ES), single active and single standby (two ELs in the ES, one being active, the other being inactive), and single active and multiple standby (at least three ELs in the ES, one being active, the other at least two being inactive). As further explained in conjunction with fig. 1a, if only one ethernet link EL1 in the ES is active and serves as the primary EL, and the other ELs 2 and EL3 are inactive and serve as backup ELs, then this redundancy mode is single active dual standby (single active multi standby).

In the ethernet link multi-homing deployment scenario, the active-all redundancy mode (referred to as active-all mode for short) refers to that all ethernet links in the ethernet link segment have active states, i.e. no inactive ethernet links. All these active ethernet links can implement load-shared forwarding of data streams, thereby providing greater bandwidth transmission capability. However, the scenario of the active mode does not support backup, that is, there is no standby ethernet link, and when one or more active ethernet links fail, the standby ethernet link cannot be switched to for redundancy protection. As explained further in connection with fig. 1a, this redundancy mode is the fully active mode if there is no backup EL if all three ethernet links EL1, EL2, and EL3 in the ES are active.

Therefore, although the single active mode has at least one standby ethernet link to protect a single active ethernet link, only one ethernet link for carrying data streams cannot share load, and cannot flexibly provide transmission capability with larger bandwidth. The active mode can support load sharing and provide transmission capability with larger bandwidth, but has no protection of the standby ethernet link, and when the active ethernet link fails, the active/standby switching cannot be performed, and the reliability is poor.

The invention expands the multi-active mode in the multi-homing scene of the Ethernet link and informs the multi-active mode to the far-end PE, thereby guiding the far-end PE equipment to transmit the data stream load sharing to the near-end PE equipment, further utilizing the Ethernet link between the near-end PE equipment and the CE to realize the forwarding of the load sharing, and simultaneously utilizing the inactive Ethernet link to carry out backup protection, thereby not only providing the transmission capability with larger bandwidth through the load sharing, but also utilizing the backup protection to improve the reliability.

It should be noted that in the ethernet link multihoming deployment scenario, the active-multiple redundancy mode (abbreviated as active-multiple mode) refers to that the states of a part of ethernet links in the ethernet link segment are active, and the states of another part of ethernet links are inactive. These active ethernet links (acting as active ethernet links) may enable load-shared forwarding of data streams, thereby providing greater bandwidth transmission capability. And the other part of the Ethernet links in the inactive state are used for backup, and when one or more active Ethernet links are in failure, the standby Ethernet links can be switched to for redundancy protection. As further explained in conjunction with fig. 1a, if two ethernet links EL1 and EL2 in the ES are active and EL3 is inactive, then EL1 and EL2 jointly forward the data stream for load sharing, and EL3 provides backup protection for EL1 or EL 2.

Fig. 1b shows a PW multiple homing EVPN scenario. In this hierarchical deployment scenario, PE devices are subdivided into two categories, namely UPE and upper PE (SPE). The UPE is a convergence device of a user, i.e. a device directly connected to a CE, also called a user-side PE. The SPE is a device connected to the UPE and located inside the network, also called a switching PE. A PW may be referred to as a pseudowire, and may also be referred to as a virtual link. The meaning of these terms will be understood by those skilled in the art. In the scenario shown in FIG. 1b, the UPE device is connected to SPE1-1, SPE1-2, SPE1-3 and SPE1-4 through PW1, PW2, PW3 and PW4, respectively. A group of PW links consisting of these four PWs is called a Pseudo Wire Segment (PWs). The Pseudo-Segment Identifier (PWSI) is a unique non-zero Identifier that identifies a Pseudo-Segment. The MAC address (which is a Local MAC address) of the UE1 is flooded to SPEs 1-1, 1-2, 1-3 and 1-4 through CE1 and then UPE, so that SPE1-1, SPE1-2, SPE1-3 and SPE1-4 learn the MAC address of the UE1 from the UPE connected with the UE, and PE2 is connected with the UPE through SPE1-1, SPE1-2, SPE1-3 or SPE 1-4. Thus SPE1-1, SPE1-2, SPE1-3 and SPE1-4 are near end PEs from the perspective of being close to UPE or CE1 or from the perspective of the MAC address of the UE1 being the Local MAC address of SPE1-1, SPE1-2, SPE1-3 and SPE 1-4. In contrast, PE2 is a far-end PE device from the perspective of being far from the UPE or from the perspective of the MAC address of the UE1 as the Local MAC address of SPE1-1, SPE1-2, SPE1-3, and SPE 1-4. The site (English: site)1 and the MAC route in the site2 of the VPN1, such as the MAC address of a User Equipment (UE) 1 in the site1 and the MAC address of a UE2 in the site2, are notified among the SPEs 1-1, SPE1-2, SPE1-3, SPE1-4 and PE2 through a BGP protocol, so that the intercommunication between the UE1 in the VPN1 and the UE2 is realized. Moreover, the UE1 and the UE2 can realize the intercommunication through the multi-home link between the UPE and the SPE1-1, the SPE1-2, the SPE1-3 and the SPE 1-4.

However, although one of the important advantages of the EVPN scheme is that the multi-homing access of the user-side device is implemented, in the PW access EVPN technology, a single-active, multi-active, or fully-active redundancy mode cannot be notified between PE devices, the advantage that the EVPN supports the multi-homing access cannot be effectively exerted, multi-active single-standby or multi-active multi-standby cannot be supported, single-active single-standby or single-active multi-standby cannot be supported, and PW load sharing and forwarding for data streams in the multi-active mode and the fully-active mode cannot be supported.

In the PW multi-homing deployment scenario, the meaning and description of the so-called single active mode, full active mode, and multi-active mode are similar to those in the ethernet link multi-homing deployment scenario, and only the ethernet link is changed into the PW.

The invention expands the multi-active mode, the full-active mode and the single-active mode in the PW multi-homing scene and informs the multi-active mode, the full-active mode and the single-active mode to the far-end PE. Therefore, in the multi-active mode, the notification guides the PE device at the far end to send the data stream load sharing to the SPE device at the near end, further utilizes the PW between the SPE device at the near end and the UPE to realize the forwarding of the load sharing, and simultaneously can also use the inactive PW to perform backup protection, thereby not only providing the transmission capability of larger bandwidth through the PW load sharing, but also improving the reliability of the PW load sharing forwarding by utilizing the backup protection. In the full-live mode, the announcement guides the PE device at the far end to send the data stream load sharing to all the SPE devices at the near end, and further utilizes all PWs between all the SPE devices at the near end and the UPE respectively to realize the maximum PW load sharing and forwarding and provide the maximum bandwidth within the transmission capability range. In the single active mode, the announcement guides the PE device at the far end to send the data stream to a near-end SPE device, and further, the PW in an active state between the near-end SPE device and the UPE is used for forwarding the data stream, and meanwhile, the PW in an inactive state can be used for carrying out backup protection on the data stream, so that the reliability is improved by using the backup protection.

It should be noted that the CE device, the PE device, the UPE device, and the SPE device may be routers or switches. The CE device is usually connected to the PE or UPE device on one side and to the UE on the other side, so as to transfer and access the user equipment to the operator network. The UE may include various handheld devices, vehicle-mounted devices, wearable devices, computer devices or other processing devices connected to a wireless modem with wireless communication functions, as well as various forms of user Equipment, Mobile Station (MS), Terminal (Terminal), Terminal Equipment (TE), and so on. For convenience of description, the above-mentioned devices are collectively referred to as user equipment or UE in this application.

It should be further noted that the PE and the PE device have the same meaning in the embodiments of the present invention, and may be used together. Similarly, CE and CE devices, UPE and UPE devices, and SPE devices may also be used with one another. The data stream described in the present invention may be a known unicast data stream.

Several possible application scenarios to which the present invention relates are described above, and embodiments of the present invention will be further detailed based on these scenarios.

Fig. 2 is a flowchart illustrating a method for notifying in a multi-active mode according to an embodiment of the present invention. The method is applied to an Ethernet virtual private line network, user side equipment is respectively accessed to scenes of at least three Provider Edge (PE) equipment through at least three links, the at least three links form a group of links, the group of links are link sections, and the at least three PE equipment comprise first PE equipment.

The scheme provided by the embodiment of the invention comprises a part 201, a part 202 and a part 203, wherein the part 201 and the part 202 are executed on a first PE device, and the part 203 is executed on a far-end PE device, which are respectively explained below.

At element 201, the first PE device obtains a redundancy pattern for the link segment.

The redundant mode of the link segment may be a multi-active mode, a full active mode, or a single active mode. The manner in which the first PE device obtains the redundancy pattern of the link segment may include, but is not limited to, the following:

first, the network administrator configures the redundant mode of the link segment (e.g., 1 for multi-active mode, 2 for full-active mode, and 3 for single-active mode) on the first PE device in advance through a command line.

And secondly, the first PE device receives messages sent by other control management devices, wherein the messages carry the redundancy mode of the link section. And the first PE equipment carries out configuration according to the redundancy mode carried in the message.

And thirdly, the first PE equipment operates the redundancy mode generation algorithm, automatically generates the redundancy mode according to the algorithm software, and completes configuration according to the generated redundancy mode.

In a fourth mode, when software running on the first PE device is developed, a default value of a redundancy mode parameter is set in advance in a software code, for example, the mode is a multi-active mode.

At 202 and 203, when the redundant mode of the link segment is a multi-active mode, the first PE device generates and sends a first mode advertisement message to a remote PE device, the first mode advertisement message including the multi-active mode and an identification of the link segment, the identification of the link segment uniquely identifying the link segment. The first mode advertisement message is used for advertising the far-end PE device, and a part of the links in the link segment can be used for forwarding a known unicast data stream, and the number of the part of links is greater than 1 and less than the maximum number of links in the link segment. The far-end PE device receives a first mode notification message sent by the first PE device.

The at least three PE devices further include a second PE device and a third PE device, where the second PE device and the third PE device also respectively obtain the redundant mode of the link segment (the obtaining manner is the same as that of the first PE device, please refer to it, and is not described herein again), and then respectively generate and send a second mode notification message and a third mode notification message to the remote PE device, where the second mode notification message and the third mode notification message also respectively include the identity of the multi-active mode and the link segment. The second mode advertisement message and the third mode advertisement message are also used for advertising the far-end PE device, respectively, a part of links in the link segment may be used for forwarding a known unicast data stream, and the number of the part of links is greater than 1 and less than the maximum number of links in the link segment. And the far-end PE equipment receives a second mode notification message sent by the second PE equipment and a third mode notification message sent by the third PE equipment. Therefore, in the EVPN, the notification of the multi-active mode is realized among the PE devices, and the support of multi-active single standby or multi-active multi-standby is facilitated. It should be understood that in an actual deployment, the at least three PE devices may also include a fourth PE device, a fifth PE device, and so on.

In addition, the PE device connected to the active link of the at least three PE devices further sends a MAC routing message to the remote PE device, where the MAC routing message includes a MAC address of the UE, and the UE accesses the EVPN through the user side device. If the states of the first link and the second link are both active and the state of the third link is inactive, the first PE device and the second PE device encapsulate the MAC address of the UE into the first MAC routing message and the second MAC routing message, respectively, and notify the first MAC routing message and the second MAC routing message to the remote PE after receiving the MAC broadcast message (the source MAC address in the message is the MAC address of the UE) sent by the user side device. Because the third link is inactive, the third PE device connected to the third link does not need to send MAC routing messages. The remote PE device receives the first MAC routing message from the first PE device and the second MAC routing message from the second PE device, respectively, and uses the first PE device and the second PE device as a next hop node of an MAC address (which is a destination MAC address) to the UE, thereby implementing a load-sharing forwarding data stream by using the first link connected to the first PE device and the second link connected to the second PE device. The method needs the remote PE device to use all active links to jointly perform load sharing forwarding data stream after receiving the MAC routing messages sent by all PE devices connected to the active links in the at least three PE devices. That is, when the far-end PE device receives only the first MAC routing message sent by the first PE device and forwards a data stream destined for the UE, only the first link can be used. When the far-end PE device continues to receive the second MAC routing message sent by the second PE device and forwards the data stream destined for the UE, the far-end PE device can use the first link and the second link to perform forwarding for load sharing at the same time.

When one of the first link and the second link fails (e.g., the second link fails), a standby third link is enabled. And the second PE device sends a MAC route withdrawal message to the far-end device, and the third PE device sends a third MAC route message to the far-end device. And the far-end PE equipment continues to carry out load sharing forwarding data flow by utilizing the first link and the third link. Therefore, the transmission capability of larger bandwidth is provided through load sharing, and the reliability of load sharing and forwarding is improved by utilizing backup protection.

The above describes a mode advertisement method not carrying a link status, and the following is a mode advertisement method carrying a link status. In one example, when the redundancy mode of the link segment is a multi-active mode, the first PE device further obtains a state of a first link between the user-side device and the first PE device, where the state of the first link is active or inactive. Accordingly, the first mode advertisement message further includes a status of the first link. Similarly, the second PE device and the third PE device also obtain a state of a second link between the user-side device and the second PE device and a state of a third link between the user-side device and the third PE device, respectively, where the state of the second link is active or inactive, and the state of the third link is active or inactive. Correspondingly, the second mode advertisement message further includes a state of the second link, and the third mode advertisement message further includes a state of the third link.

The far-end PE device receives the first mode advertisement message from the first PE device from a first interface, the second mode advertisement message from the second PE device from a second interface, and the third mode advertisement message from the third PE device from a third interface. If the state of the first link carried in the first mode notification message is active, the state of the second link carried in the second mode notification message is active, and the state of the third link carried in the third mode notification message is inactive, the far-end PE device determines that the first link and the second link can be combined for forwarding of load sharing, and the third link is used for backing up the first link or the second link.

In addition, the PE device connected to the active link of the at least three PE devices further sends a MAC routing message to the remote PE device, where the MAC routing message includes a MAC address of the UE, and the UE accesses the EVPN through the user side device. Therefore, after receiving the MAC broadcast message sent by the UE (the source MAC address in the message is the MAC address of the UE), the first PE device and the second PE device encapsulate the MAC address of the UE into the first MAC routing message and the second MAC routing message, respectively, and notify the first MAC routing message and the second MAC routing message to the remote PE. Because the third link is inactive, the third PE device connected to the third link does not need to send MAC routing messages. And the far-end PE equipment obtains the MAC address of the UE after receiving the first MAC routing message from the first PE equipment. And the remote PE generates an MAC forwarding table entry using the MAC address of the UE as a destination MAC address and the first interface and the second interface as outgoing interfaces according to the determined first link and the determined second link, which can be used as two load sharing links. That is, on the far-end PE device, the first PE device and the second PE device are used as next hop nodes to the MAC address of the UE, and then the first link connected to the first PE device and the second link connected to the second PE device are used to implement a forwarding data stream for load sharing.

In the example of advertising the link status, the mode advertisement message carries status information of the link, so that the remote PE device may determine in advance an active link that can be used for load sharing forwarding. Then, when receiving the MAC routing message sent by any one of the at least three PE devices, the far-end PE device may completely generate an MAC forwarding table entry for load sharing, and may forward the data stream using all links that can be used for load sharing, so as to achieve load sharing more quickly, improve the effect of load sharing, and make load sharing more balanced.

When one of the first link and the second link fails (e.g., the second link fails), a standby third link is enabled. And the second PE device sends a fourth mode notification message to the remote device, wherein the state of the second link carried in the fourth mode notification message is inactive. And the third PE device sends a fifth mode notification message to the remote device, wherein the state of the third link carried in the fifth mode notification message is active. And the remote PE equipment updates the outgoing interfaces of the MAC forwarding table entry into the first interface and the third interface. And the far-end PE equipment continues to carry out load sharing forwarding data flow by utilizing the first link and the third link. Therefore, the transmission capability of larger bandwidth is provided through load sharing, and the reliability of load sharing and forwarding is improved by utilizing backup protection.

It should be noted that, in the above two embodiments of the method for not announcing the link status and announcing the link status, the deployment mode of dual active devices and single standby devices is used as an example. In actual deployment, a dual-active dual-standby deployment mode can be adopted. For example, there are four links, where the first link and the second link are active (serving as active links) and the third link and the fourth link are inactive (serving as standby links). When both the active first link and the active second link fail, the active first link and the active second link may be switched to the standby third link and the standby fourth link at the same time, and the third link and the standby fourth link are continuously used for forwarding load sharing. Compared with a deployment mode of double active devices and single standby devices, the deployment mode of double active devices and double standby devices has higher reliability. It should be further noted that the multi-active single standby may include a double-active single standby, a triple-active single standby, a quadruple-active single standby, and the like, and the multi-active refers to two or more active links. The multi-active multi-standby can comprise multi-active two-standby, multi-active three-standby, multi-active four-standby and the like, and the multi-standby means that the number of the backup links is more than or equal to two.

While fig. 2 describes the embodiment of multiple active mode notification and multiple active load sharing in EVPN, it should be noted that the link may be an ethernet link or a PW, the link segment may be an ethernet link segment or a pseudo-line segment, and the link segment identifier may be ESI or PWSI. The method shown in fig. 2 can be applied in particular to the two deployment scenarios shown in fig. 1a and 1 b. Embodiments in both scenarios of ethernet link access EVPN and PW access EVPN will be described below based on the embodiment described in fig. 2, and with reference to fig. 1a, 3a, and 3b, and fig. 1b, 4a, and 4b, respectively.

Taking the scenario of ethernet link access EVPN shown in fig. 1a as an example, fig. 3a shows an interactive schematic diagram of a method flow of multi-active mode advertisement and load sharing without a link state provided in an embodiment of the present invention. As shown in FIG. 1a, CE1 is mostly accessed to PE1-1, PE1-2 and PE1-3 through EL1, EL2 and EL3 respectively, UE1 is accessed to EVPN through CE1, and the far-end PE is PE 2. EL1, EL2, and EL3 form an ethernet segment, and ESI is the identity of the ethernet segment. As shown in FIG. 3a, the following steps are performed on PE 1-1:

3a-1101, obtaining the redundant mode of the ES as a multi-active mode (the obtaining mode is consistent with that described in fig. 2, and is not described here again).

3a-1102, generating a first mode advertisement message, the first mode advertisement message including the multi-activity mode and the ESI. The first mode advertisement message is used to advertise PE2 that a portion of the ethernet links in the ES may be used to forward a data stream known to be unicast.

3a-1103, sending the first mode advertisement message to PE 2.

Similar steps to those of PE1-1 are performed on PE1-2 and PE1-3, which have the same principle, and are briefly described below, and please refer to the description of PE1-1, which is not repeated herein.

PE1-2 and PE1-3 respectively obtain the redundant mode of ES as multi-active mode (the obtaining manner is the same as that described in FIG. 2, and is not described here again).

PE1-2 generates a second mode advertisement message that includes the multi-active mode, the EL2 being an active state, and the ESI. The second mode advertisement message is used to advertise PE2 that a portion of the ethernet links in the ES may be used to forward a data stream known to be unicast. And sends the second mode advertisement message to PE 2.

PE1-3 generates a third mode advertisement message that includes the multi-active mode and the ESI. The third mode advertisement message is used to advertise PE2 that a portion of the ethernet links in the ES may be used to forward a data stream known to be unicast. And sends the third mode advertisement message to PE 2.

The following steps are performed on PE 2:

3a-201, receiving from Interface (Intf) 1, said first mode advertisement message from PE1-1, said first mode advertisement message comprising said multi-active mode and said ESI. PE2 validates that the ESI identified redundancy mode of the ES is a multi-active mode.

3a-202, receiving from Intf 2 said second mode advertisement message from PE1-2, said second mode advertisement message comprising said multi-active mode and said ESI. PE2 validates that the ESI identified redundancy mode of the ES is a multi-active mode.

3a-203, receiving from Intf 3 said third mode advertisement message from PE1-3, said third mode advertisement message comprising said multi-active mode and said ESI. PE2 validates that the ESI identified redundancy mode of the ES is a multi-active mode.

In addition, the PE1-1 and PE1-2 connected to the ELI and EL2 in the active state receive the MAC broadcast message (for MAC learning) sent from the CE1, where the source MAC address of the MAC broadcast message is the MAC address of the UE1, and the destination MAC address is the broadcast address. PE1-1 and PE1-2 also generate and send MAC routing messages to PE2, respectively, as follows.

PE1-1 also performs the steps of:

3a-1104, generating a first MAC routing message, the first MAC routing message including a MAC address of the UE1 and a next hop network address 1, the next hop network address 1 being a loopback (english: loopback) address of the PE 1-1. It should be noted that the loopback address described in the present invention is an IP address configured on a loopback interface of a network device (e.g., a router, a switch, etc.), and is usually used as a network device identifier (e.g., a 32-bit masked IPv4 address: 10.10.1.1/32), as will be understood by those skilled in the art.

3a-1105, sending the first MAC routing message.

PE1-2 also performs the steps of:

3a-1204, generating a second MAC routing message, the second MAC routing message including the MAC address of the UE1 and the next hop network address 2, the next hop network address 2 being a loopback address of PE 1-2.

3a-1205, sending the second MAC routing message.

PE2 also performs the steps of:

3a-204, receives the first MAC routing message from PE1-1 from Intf 1. PE2 regards PE1-1 as the next hop node to UE1 according to the next hop network address 1 (being the loopback address of PE 1-1) carried in the first MAC routing message. A control plane of PE2 (e.g., a control board of PE2) generates a MAC routing entry (as shown in table 1), where a destination MAC address of the MAC routing entry is a MAC address of the UE1, and a next hop network address is a loopback address of PE 1-1. The control plane of PE2 then generates a MAC forwarding entry (as shown in table 2) from the MAC routing entry and sends the MAC forwarding entry to the forwarding plane of PE2 (e.g., the forwarding board of PE 2). The destination MAC address of the MAC forwarding entry is a destination address in the MAC routing entry (i.e., the MAC address of the UE 1), and the output interface of the MAC forwarding entry is Intf 1. The obtaining method of Intf1 may be: firstly, the PE2 looks up a mapping table (also called FTN mapping table or FTN Forwarding table) from Forwarding Equivalence Class (FEC) to Next Hop Label Forwarding Entry (NHLFE) with the loopback address of the PE1-1 in the MAC routing table Entry as the destination IP address (also called matching Entry or key value), and obtains a Tunnel Identifier (Tunnel ID) of a Tunnel whose exit interface is PE2 to PE1-1 corresponding to the loopback address of the PE 1-1; then, the Tunnel forwarding table is searched by using the Tunnel ID, and an outbound interface corresponding to the Tunnel ID is Intf1 (i.e., an interface of the Tunnel from PE2 to PE1-1 on PE2) is obtained. It should be noted that the tunnel may be a Label Switched Path (LSP) tunnel, or a Resource Reservation Protocol-Traffic Engineering (RSVP-TE) tunnel. The tunnel is used to carry known unicast data streams, and for the sake of brevity, the embodiments of the present invention are not shown in fig. 1a and 1b, and will be understood by those skilled in the art.

Table 1: MAC routing table

Destination MAC	Next hop network address list
		MAC Address of UE1	Loopback address of PE1-1

Table 2: MAC forwarding table

Destination MAC	Exit interface list
		MAC Address of UE1	Intf	1

When PE2 receives a known unicast data stream from CE2 destined for UE1 (the destination MAC address carried in the data packet is the MAC address of UE 1), PE2 looks up the MAC forwarding table (as in table 2) to obtain Intf1 as the egress interface. Then, the PE2 forwards the packet in the data flow from the Intf1, and the packet reaches the PE1-1 through the tunnel from the PE2 to the PE1-1, so as to be forwarded to the UE1 by using the EL 1.

3a-205, receives the second MAC routing message from PE1-2 from Intf 2. PE2 takes PE1-2 as the next hop node to UE1, as with the steps 3a-204 described above. The control plane of PE2 adds the loopback address of PE1-2 to the list of next hop network addresses in the MAC routing table (as shown in table 1-1). Then, the control plane of PE2 obtains Intf 2 according to the loopback address of PE1-2, and adds Intf 2 in the outgoing interface list of MAC forwarding table entry (as shown in table 2-1).

Tables 1 to 1: MAC routing table

Table 2-1: MAC forwarding table

Destination MAC	Exit interface list
		MAC Address of UE1	Intf	1、Intf 2

At this time, the MAC forwarding table entry is used to instruct the PE2 to forward the data stream (known unicast data stream) destined for the UE1 through Intf1 and Intf 2 load sharing. For example, the packet 1 in the data flow is forwarded from the Intf1, and reaches the PE1-1 through the tunnel from the PE2 to the PE1-1, so as to be forwarded to the UE1 by using the EL 1. The packet 2 in the data flow is forwarded from the Intf 2, and reaches the PE1-2 through the tunnel from the PE2 to the PE1-2, so as to be forwarded to the UE1 by using the EL 2.

From the above, after receiving the MAC routing messages sent from PE1-1 and PE1-2, PE2 uses all active links to jointly perform load sharing on the forwarded data stream. EL1 can only be used when PE2 forwards a data stream destined for the UE by receiving only the first MAC routing message from PE 1-1. When PE2 continues to receive the second MAC routing message from PE1-2 and forwards the data stream to the UE, it can use EL1 and EL2 for load sharing forwarding.

In one example, the first, second, and third mode advertisement messages are BGP Update (english: Update) messages (also referred to as BGP Update messages) that contain Ethernet Auto-Discovery routes (english: Ethernet Auto-Discovery, Ethernet a-D Route). The Ethernet automatic discovery route belongs to a route type in EVPN Network Layer Reachable Information (NLRI) defined by BGP protocol. The EVPN NLRI is loaded in the attribute of multi-protocol accessible information (MP _ REACH _ NLRI). The MP _ REACH _ NLRI attribute is an attribute defined in the BGP Update message, and is specifically formatted as shown in fig. 3c, and includes an Address Family Identifier (AFI) field and a Sub Address Family Identifier (SAFI) field. The value of the AFI field is used to indicate L2VPN, e.g. 25. The value of the SAFI field is used to indicate EVPN, e.g., 70. The MP _ REACH _ NLRI attribute also includes the Next Hop Network Address Length (in English) and Next Hop Network Address (in English) fields. The next hop network address field is used to carry the network address (e.g., the loopback address) of the next hop node described in the above embodiments. The MP _ REACH _ NLRI attribute further includes an NLRI field, which indicates EVPN in L2VPN in combination with the values of the AFI and SAFI, and the NLRI field is an EVPN NLRI field. As shown in fig. 3d, the EVPN NLRI field includes a 1-byte Route Type (english: Route Type) field, a 1-byte Length (english: Length) field, and a variable Length Route Type detail (english: Route Type specific) field. The Route Type field includes the ethernet auto discovery Route, for example, the value is 1. The Route Type specific field carries details of the ethernet auto discovery Route. As shown in fig. 3e, the Ethernet auto-discovery Route includes an 8-byte routing specifier (RD) field, a 10-byte Ethernet Segment Identifier (ESI) field, a 4-byte Ethernet Tag Identifier (Ethernet Tag ID) field, and a 3-byte MPLS Tag (Label) field.

Also advertised with the Ethernet auto discovery route is the ESI Label Extended Community (English) attribute, whose format is shown in FIG. 3 f. The ESI Label Extended Community includes a 1-byte Type (English: Type) field, a 1-byte Sub-Type (English: Sub-Type) field, a 1-byte flag (English: Flags) field, a 2-byte Reserved (English: Reserved) field, and a 3-byte ESI Label field. The Flags field is used to indicate a redundancy mode under the multi-homing condition, for example, when the value of Flags is 0x00(0x represents hexadecimal), a single active mode is indicated; when the value of Flags is 0x01, indicating a full-live mode; the invention expands the value of Flags and increases the support for the multi-active mode, for example, when the value is 0x02(0x represents hexadecimal), the multi-active mode is indicated. It should be noted that the values herein are merely examples, and other values are possible, and do not limit the present invention.

In another example, the first and second MAC routing messages are BGP Update messages, which contain MAC/IP Advertisement routes (english). This MAC/IP advertised Route is of another Route Type defined by Route Type in EVPN NLRI field described in the above example (as shown in fig. 3 d) for advertising MAC/IP reachable address information. Similar to the ethernet auto-discovery route, the EVPN NLRI of the MAC/IP advertisement route type is also carried in the MP _ REACH _ NLRI attribute, and the MP _ REACH _ NLRI attribute includes a next hop network address field for carrying the network address (e.g., loopback address) of the next hop node in the above embodiment. The description of the MP _ REACH _ NLRI attribute and the EVPN NLRI is the same as the above example, please refer to this example, and will not be repeated here.

It should be noted that, in the present invention, For details about EVPN NLRI, ethernet auto discovery routing, ESI Label Extended connectivity, and MAC/IP advertisement routing, please refer to Request For Comments (RFC) 7432 issued by the Internet Engineering Task Force (IETF), and the contents of the document and the relevant parts are generally incorporated into the present document as if they were copied in their entirety, and are not described herein again For brevity.

Through the announcement of the ES multi-active mode and the distribution of the MAC route, it is achieved that in the ES multi-homing scenario, a known unicast data stream to the UE1 is forwarded through EL1 and EL2 load sharing, providing a larger bandwidth of transmission capability. When any one of the EL1 and the EL2 fails, backup protection can be performed by switching to the EL3, and reliability of EL load sharing forwarding is improved.

Taking the scenario of ethernet link access EVPN shown in fig. 1a as an example, fig. 3b shows a schematic interaction diagram of a method flow for multi-active mode advertisement and load sharing with a link state provided in an embodiment of the present invention. As shown in FIG. 1a, CE1 is mostly accessed to PE1-1, PE1-2 and PE1-3 through EL1, EL2 and EL3 respectively, UE1 is accessed to EVPN through CE1, and the far-end PE is PE 2. EL1, EL2, and EL3 form an ethernet segment, and ESI is the identity of the ethernet segment. As shown in FIG. 3b, the following steps are performed on PE 1-1:

3b-1101, obtaining the redundant mode of the ES as a multi-active mode (the obtaining mode is consistent with that described in part in FIG. 2 and is not described herein again).

3b-1102, the state of the get EL1 may be active or inactive (see fig. 2 for details), for example, the state of the EL1 is active.

3b-1103, generating a first mode advertisement message comprising the multi-active mode, EL1 being active state, and the ESI. The first mode advertisement message is used to advertise PE2, and part of ethernet links in the ES may be used to forward data streams known to be unicast, wherein EL1 may be used to forward data streams known to be unicast. It should be noted that, when the state of the EL1 is inactive, the first mode advertisement message is used to advertise PE2, and part of ethernet links in the ES may be used to forward data streams of known unicast, where EL1 may not be used to forward data streams of known unicast, but is used as a backup link.

The first mode advertisement message further includes a next hop network address 1, where the next hop network address 1 is a loopback address (e.g., IP address: 10.10.1.1/32) of PE 1-1.

3b-1104, sends the first mode advertisement message to PE 2.

Similar steps to those of PE1-1 are performed on PE1-2 and PE1-3, the principle is the same, and the description is briefly provided below, and for details, refer to the description of PE1-1, and are not repeated.

PE1-2 and PE1-3 respectively acquire the redundant mode of the ES as a multi-active mode.

PE1-2 and PE1-3 obtain the states of EL2 and EL3, respectively, and may be active or inactive, e.g., the state of EL2 is active and the state of EL3 is inactive.

PE1-2 generates a second mode advertisement message that includes the multi-active mode, the EL2 being an active state, and the ESI. The second mode advertisement message is used to advertise PE2, and part of ethernet links in the ES may be used to forward data streams known to be unicast, wherein EL2 may be used to forward data streams known to be unicast. The second mode advertisement message also includes next hop network address 2, and next hop network address 2 is a loopback address (e.g., IP address: 20.20.1.1/32) of PE 1-2. PE1-2 sends the second mode advertisement message to PE 2.

PE1-3 generates a third mode advertisement message that includes the multi-active mode, the EL3 being inactive, and the ESI. The third mode advertisement message is used to advertise PE2, and part of ethernet links in the ES may be used to forward data streams of known unicast, wherein EL3 may not be used to forward data streams of known unicast. The third mode advertisement message further includes next hop network address 3, and next hop network address 3 is a loopback address (e.g., IP address: 30.30.1.1/32) of PE 1-3. PE1-3 sends the third mode advertisement message to PE 2.

The following steps are performed on PE 2:

3b-201, receiving from Intf1 said first mode advertisement message from PE1-1, said first mode advertisement message comprising said multi-active mode, EL1 as active state and said ESI. PE2 confirms that the ESI identifies the redundant mode of the ES as multi-active mode, and EL1 connected to PE1-1 is an ethernet link in the ES with active stripe, which can be used to forward data streams known as unicast.

3b-202, receiving from Intf 2 said second mode advertisement message from PE1-2, said second mode advertisement message comprising said multi-active mode, EL2 as active state and said ESI. PE2 confirms that the ESI identifies the redundant mode of the ES as multi-active mode, and EL2 connected to PE1-2 is an ethernet link in the ES with active stripe, which can be used to forward data streams known as unicast.

3b-203, receiving from Intf 3 said third mode advertisement message from PE1-3, said third mode advertisement message comprising said multi-active mode, EL3 being inactive state and said ESI. PE2 confirms that the ESI identifies the redundant mode of the ES as multi-active mode, and EL3 connected to PE1-3 is an ethernet link in the ES with a bar state inactive and cannot be used to forward data streams known to be unicast.

3b-204 EL1, EL2 and EL3 obtained from said first, second and third mode advertisement messages have the same said ESI and thus belong to the same ES. PE2 also determines that the redundancy mode of the ES is a multi-active mode according to the multi-active modes carried in the first, second, and third mode advertisement messages, respectively. PE2 further determines that EL1 and EL2 in the ES are used as links for load sharing forwarding and EL3 is used as a backup link according to EL1 and EL2 being active and EL3 being inactive. PE2 further generates and stores a mapping relationship between link information and next hop node information according to the next hop network address information carried in the first, second, and third mode advertisement messages, for example, as shown in table 3. The mapping relationship is used to indicate an association relationship between information of each link in the link segment and information of a next hop node of a forwarding device (e.g., PE2) forwarding the data packet. The link information may include link segment identification, link status, etc. The next hop node information may include a network address (e.g., the network address is an IP address), typically a loopback address, of the next hop node. It should be noted that the network address of the next hop node and the network address of the next hop are the same in the embodiments of the present invention, and may be used with each other.

Table 3: mapping relation between link information and next hop network address information

Link segment identification	Ethernet link identification	Status of state	Next hop node information
				ESI	EL1	Activating	Loopback address of PE1-1
ESI	EL2	Activating	Loopback address of PE1-2
				ESI	EL3	Is inactive	Loopback address of PE1-3

In addition, the PE1-1 and PE1-2 connected to the EL1 and EL2 in the active state receive the MAC broadcast message (for MAC learning) sent from the CE1, where the source MAC address of the MAC broadcast message is the MAC address of the UE1, and the destination MAC address is the broadcast address. PE1-1 and PE1-2 also generate and send MAC routing messages to PE2, respectively, as follows.

PE1-1 also performs the steps of:

3b-1104, generating a first MAC routing message that includes the MAC address of the UE 1.

3b-1105, sending the first MAC routing message.

PE1-2 also performs the steps of:

3b-1204, generating a second MAC routing message including the MAC address of the UE 1.

3b-1205, sending the second MAC routing message.

PE2 also performs the steps of:

3b-205, receiving the first MAC routing message sent by PE1-1 from Intf1, where the first MAC routing message carries a MAC address of UE 1. A control plane (e.g., a control board of PE2) of the PE2 generates a MAC forwarding entry (e.g., table 4) according to the MAC address of the UE1 and the mapping relationship (e.g., table 3) between the link information and the next hop network address information, where a destination MAC address of the MAC forwarding entry is the MAC address of the UE1, and outgoing interfaces of the MAC forwarding entry are Intf1 and Intf 2. Intf1 and Intf 2 are obtained by PE2 according to the loopback address of PE1-1 and loopback address of PE1-2 corresponding to EL1 and EL2 in active state. For the sake of brevity, please refer to the description of the steps 3a-204, which is similar to the above description.

The MAC forwarding entry is used to direct the PE2 to forward a data flow (known as a unicast data flow) destined for the UE1 through Intf1 and Intf 2 load sharing. For example, the packet 1 in the data flow is forwarded from the Intf1, and reaches the PE1-1 through the tunnel from the PE2 to the PE1-1, so as to be forwarded to the UE1 by using the EL 1. The packet 2 in the data flow is forwarded from the Intf 2, and reaches the PE1-2 through the tunnel from the PE2 to the PE1-2, so as to be forwarded to the UE1 by using the EL 2.

Table 4: MAC forwarding table

Destination MAC	Outlet interface
		MAC Address of UE1	Intf	1、Intf 2

PE2 then also receives the second MAC routing message from PE1-2 from Intf 2, but this message has no effect on the generation of the load-sharing MAC forwarding table. It should be understood that this is only an example, and PE2 may also receive the second MAC routing message sent by PE1-2 from Intf 2 and then receive the first MAC routing message sent by PE1-1 from Intf 1. When receiving the second MAC routing message sent by PE1-2 from Intf 2, the MAC forwarding table shown in table 4 is generated in the same manner.

From the above, when the PE2 receives a MAC routing message sent by any one of the end point PE1-1 of the active link EL1 and the end point PE1-2 of the active link EL2, the PE2 can completely generate a MAC forwarding table entry for load sharing (without waiting for receiving MAC routing messages sent by end points corresponding to all active links), and can forward data streams by using all links EL1 and EL2 that can be used for load sharing, so that load sharing can be realized faster, the effect of load sharing is improved, and load sharing is more balanced.

In one example, the first, second, and third mode advertisement messages are BGP Update messages that contain ethernet auto-discovery routes. The related description is consistent with the description of the multi-active mode advertisement not carrying link status in the above-mentioned portion of fig. 3a, please refer to this description, and will not be described again here. In contrast, embodiments of the present invention also advertise the status of the ethernet link. In a specific embodiment, the implementation is realized by extending the value of the Flags field in the ESI Label Extended Community. For example, one value is added to indicate the multi-active mode and the link status is active (e.g., Flags is 0x02, 0x represents hexadecimal), and another value is added to indicate the multi-active mode and the link status is inactive (e.g., Flags is 0x03, 0x represents hexadecimal). It should be noted that the values herein are merely examples, and other values are possible, and do not limit the present invention.

In another example, the first and second MAC routing messages are BGP Update messages, and the BGP Update messages include MAC/IP advertisement routes. The related description is consistent with the description of the multi-active mode advertisement not carrying link status in the above-mentioned portion of fig. 3a, please refer to this description, and will not be described again here.

Next, taking the scenario of PW accessing EVPN shown in fig. 1b as an example, fig. 4a shows an interactive schematic diagram of a method flow of multiple active mode advertisement and load sharing without carrying a link state provided in the embodiment of the present invention. As shown in FIG. 1b, the UPE is respectively accessed to SPE1-1, SPE1-2, SPE1-3 and SPE1-4 through PW1, PW2, PW3 and PW4, the UE1 is accessed to the EVPN through CE1 and UPE, and the far-end PE is PE 2. PW1, PW2, PW3, and PW4 constitute a PWs, the PWSI being the identity of the PWs. As shown in FIG. 4a, the following steps are performed on SPE 1-1:

4a-1101, obtaining the redundant mode of the PWS is a multi-active mode (the obtaining mode is the same as the obtaining mode of the redundant mode of the ES in the part of fig. 2, and the description is omitted here).

4a-1102, generating a first mode advertisement message comprising the multi-active mode and the PWSI. The first mode advertisement message is used to advertise PE2, and a portion of the PWS may be used to forward a data stream known to be unicast.

4a-1103, sending the first mode advertisement message to PE 2.

A similar following step is performed on SPE 1-2:

4a-1201, acquiring the PWS redundancy mode as a multi-active mode (the acquiring method is the same as that of the ES redundancy mode acquired in fig. 2, and is not described herein again).

4a-1202, generating a second mode advertisement message comprising the multi-active mode, a PW2 as active, and the PWSI. The second mode advertisement message is used to advertise PE2 that a portion of the PWS may be used to forward a data stream known to be unicast.

4a-1203 sending the second mode notification message to PE 2.

A similar following step is performed on SPE 1-3:

4a-1301, acquiring the redundant mode of the PWS as a multi-active mode (the acquiring mode is the same as the acquiring mode of the redundant mode of the ES in the part of FIG. 2, and the description is omitted here).

4a-1302, generating a third mode advertisement message comprising the multi-active mode and the PWSI. The third mode advertisement message is used to advertise PE2, and a portion of the PWS may be used to forward a data stream known to be unicast.

4a-1303, sending the third mode advertisement message to PE 2.

A similar following step is performed on SPE 1-4:

4a-1401, obtaining the redundant mode of the PWS is a multi-active mode (the obtaining mode is the same as the obtaining mode of the redundant mode of the ES in fig. 2, and the description is omitted here).

4a-1402, generating a fourth mode advertisement message, the fourth mode advertisement message including the multi-active mode and the PWSI. The fourth mode advertisement message is used to advertise PE2, and a portion of the PWS may be used to forward a data stream known to be unicast.

4a-1403, sending the fourth mode advertisement message to PE 2.

The following steps are performed on PE 2:

4a-201, receiving from Interface (Intf) 1, said first mode advertisement message from SPE1-1, said first mode advertisement message comprising said multi-active mode and said PWSI. PE2 identifies the redundant mode of the PWS identified by the PWSI as a multi-active mode.

4a-202, receiving from Intf 2 said second mode advertisement message from SPE1-2, said second mode advertisement message comprising said multi-active mode and said PWSI. PE2 identifies the redundant mode of the PWS identified by the PWSI as a multi-active mode.

4a-203, receiving from Intf 3 said third mode advertisement message from SPE1-3, said third mode advertisement message comprising said multi-active mode and said PWSI. PE2 identifies the redundant mode of the PWS identified by the PWSI as a multi-active mode.

4a-204, receiving from Intf 4 said fourth mode advertisement message from SPE1-4, said fourth mode advertisement message comprising said multi-active mode and said PWSI. PE2 identifies the redundant mode of the PWS identified by the PWSI as a multi-active mode.

Furthermore, SPE1-1 connected to PW1 in the active state and SPE1-2 connected to PW2 in the active state receive a MAC broadcast packet (for MAC learning) sent from CE1, where a source MAC address of the MAC broadcast packet is a MAC address of UE1, and a destination MAC address is a broadcast address. SPE1-1 and SPE1-2 also generate and send MAC routing messages to PE2, respectively, as follows.

SPE1-1 also performs the steps of:

4a-1104, generating a first MAC routing message, where the first MAC routing message includes the MAC address of the UE1 and a next hop network address 1, and the next hop network address 1 is a loopback (english: loopback) address of the SPE 1-1. It should be noted that the loopback address described in the present invention is an IP address configured on a loopback interface of a network device (e.g., a router, a switch, etc.), and is usually used as a network device identifier (e.g., a 32-bit masked IPv4 address: 10.10.1.1/32), as will be understood by those skilled in the art.

4a-1105, sending the first MAC routing message.

SPE1-2 also performs the steps of:

4a-1204, generating a second MAC routing message, the second MAC routing message including the MAC address of the UE1 and the next hop network address 2, the next hop network address 2 being the loopback address of the SPE 1-2.

4a-1205, sending the second MAC routing message.

PE2 also performs the steps of:

4a-204, receives the first MAC routing message from SPE1-1 from Intf 1. The PE2 takes the SPE1-1 as the next hop node to the UE1 according to the next hop network address 1 (which is the loopback address of the SPE 1-1) carried in the first MAC routing message. A control plane of the PE2 (e.g., a control board of the PE2) generates a MAC routing entry (as shown in table 5), where a destination MAC address of the MAC routing entry is a MAC address of the UE1, and a next-hop network address is a loopback address of the SPE 1-1. The control plane of PE2 then generates a MAC forwarding entry from the MAC routing entry (as shown in table 6), and sends the MAC forwarding entry to the forwarding plane of PE2 (e.g., the forwarding board of PE 2). The destination MAC address of the MAC forwarding entry is a destination address in the MAC routing entry (i.e., the MAC address of the UE 1), and the output interface of the MAC forwarding entry is Intf 1. The obtaining method of the Intf1 is consistent with the principle described in the embodiment of the above-mentioned ethernet link access EVPN scenario (as shown in fig. 3a), please refer to this, and for brevity, details are not described here again.

Table 5: MAC routing table

Destination MAC	Next hop network address list
		MAC Address of UE1	Loopback address of SPE1-1

Table 6: MAC forwarding table

Destination MAC	Exit interface list
		MAC Address of UE1	Intf	1

When PE2 receives a known unicast data stream from CE2 destined for UE1 (the destination MAC address carried in the data packet is the MAC address of UE 1), PE2 looks up the MAC forwarding table (as in table 6) to obtain Intf1 as the egress interface. Then, the PE2 forwards the packet in the known unicast data stream from the Intf1, and the packet reaches the SPE1-1 through the tunnel from the PE2 to the SPE1-1, so as to forward the packet to the UE1 by using the PW 1.

4a-205, receives the second MAC routing message from SPE1-2 from Intf 2. As with the 4a-204 steps described above, PE2 has SPE1-2 as the next hop node to UE 1. The control plane of PE2 adds the loopback address of SPE1-2 to the list of next hop network addresses in the MAC routing table (as shown in table 5-1). Then, the control plane of PE2 obtains Intf 2 according to the loopback address of SPE1-2, and adds Intf 2 in the outgoing interface list of the MAC forwarding table entry (as shown in table 6-1).

Table 5-1: MAC routing table

Table 6-1: MAC forwarding table

Destination MAC	Exit interface list
		MAC Address of UE1	Intf	1、Intf 2

At this time, the MAC forwarding table entry is used to instruct the PE2 to forward the data stream (known unicast data stream) destined for the UE1 through Intf1 and Intf 2 load sharing. For example, the packet 1 in the data flow is forwarded from the Intf1, and reaches the SPE1-1 through a tunnel from the PE2 to the SPE1-1, so as to be forwarded to the UE1 by using the PW 1. The packet 2 in the data flow is forwarded from the Intf 2, and reaches the SPE1-2 through the tunnel from the PE2 to the SPE1-2, so as to be forwarded to the UE1 by using the PW 2.

From the above, this method needs PE2 to use all active links to jointly perform load sharing forwarding data flow after receiving the MAC routing messages sent by SPE1-1 and SPE 1-2. When PE2 receives only the first MAC routing message from SPE1-1 and forwards the data stream destined for the UE, PW1 can only be used. When PE2 continues to receive the second MAC routing message from SPE1-2 and forwards the data stream destined for the UE, it can use PW1 and PW2 for forwarding load sharing. That is, the method may delay the implementation of load sharing forwarding.

In one example, the first, second, and third mode advertisement messages are BGP Update messages that contain ethernet auto-discovery routes. The related description is consistent with the description of the multi-active mode advertisement not carrying link status in the above-mentioned portion of fig. 3a, please refer to this description, and will not be described again here. In contrast, the ESI field in the ethernet auto-discovery route shown in fig. 3e is extended to add the PW type. The format of the ESI field is shown in FIG. 4c, and includes a 1-byte T field and a 9-byte ESI Value (English: Value). The T field (i.e., the ESI Type field) is used to specify the format of the ESI Value. In the embodiment of the present invention, the value of the T field is extended, and support for a virtual link (PW) is increased, for example, when the value is 0x06(0x represents hexadecimal), the type of the link or the link segment is indicated as PW. It should be noted that, here, expressing the PWSI by using the value of the extended ESI Type field is only an example, and it is also possible to redefine a PWSI field to identify the PW link segment. These examples are not intended to limit the invention.

Also advertised with the Ethernet auto discovery route is an ESI Label Extended Community attribute. The invention extends the value of Flags in the ESI Label Extended Community attribute, and increases the support for multi-active mode, for example, indicating the multi-active mode when the value is 0x02(0x represents hexadecimal). The related description is consistent with the description of the multi-active mode advertisement not carrying link status in the above-mentioned portion of fig. 3a, please refer to this description, and will not be described again here. It should be noted that the values of Flags herein are merely examples, and other values are possible, and do not limit the present invention.

Through the notification of the PWS multi-active mode and the issuance of the MAC route, it is realized that, in the PW multi-homing scenario, a known unicast data stream to the UE1 is forwarded through PW1 and PW2 load sharing, which provides a larger bandwidth transmission capability. When any one of PW1 and PW2 fails, the protection switch can be switched to any one of PW3 or PW4 for standby protection, so that the reliability of PW load sharing and forwarding is improved.

Fig. 4a illustrates a multi-active mode advertisement and load sharing method without carrying a link state according to an embodiment of the present invention, and based on this, a PW single-active mode advertisement and full-active mode advertisement and load sharing method needs to be described. The PW single active mode and full active mode advertisement are similar to the above multi active mode advertisement method without carrying the link status, but the difference is that the type of the redundant mode carried in the mode advertisement message is the single active mode or the full active mode.

In the case of PW single active mode, as shown in fig. 1b, for example, PW1 is in active state, and PW2, PW3 and PW4 are in inactive state. The redundant modes carried in the first, second, third and fourth mode notification messages sent by SPE1-1, SPE1-2, SPE1-3 and SPE1-4 to PE2 are all single active modes. PE2 determines that the PWS is in single active mode based on the received at least one mode advertisement message. SPE1-1, which is the endpoint of PW1 in the active state, also sends a MAC routing message (carrying the MAC address of UE 1) to PE 2. After receiving the MAC routing message, the PE2 generates an MAC forwarding table according to the single active mode, the MAC address of the UE1, and the next hop network address (the loopback address of the SPE 1-1). The destination address of the MAC forwarding table entry is the MAC address of UE1, and the egress interface is Intf 1. The specific step flow is the same as the flow described in the multi-active mode of fig. 4a, please refer to this, and for brevity, will not be described again. Through the announcement of the single active mode of the PWS and the issuance of the MAC route, the known unicast data stream to the UE1 is forwarded through one PW in the PW multi-homing scenario. When PW1 fails, the fault can be switched to any one of PW2, PW3 or PW4, so that backup protection is realized and reliability is improved.

In the case of PW full active mode, as shown in fig. 1b, for example, PW1, PW2, PW3 and PW4 are all active, and the redundant modes carried in the first, second, third and fourth mode notification messages respectively sent by SPE1-1, SPE1-2, SPE1-3 and SPE1-4 to PE2 are all full active modes. The PE2 determines that the PWS is in the all-active mode based on the received at least one mode advertisement message. SPE1-1, SPE1-2, SPE1-3, and SPE1-4 (which are the endpoints of PW1, PW2, PW3, and PW4, respectively, in an active state) also send first, second, third, and fourth MAC routing messages (carrying the MAC address of UE 1), respectively, to PE 2. After receiving all the four MAC routing messages, the PE2 generates a MAC routing table entry (generated at the control board of the PE2) according to the full active mode, the MAC address of the UE1, and the four next hop network addresses (the loopback address of the SPE1-1, the loopback address of the SPE1-2, the loopback address of the SPE1-3, and the loopback address of the SPE 1-4). The destination MAC address of the MAC routing table entry is the MAC address of the UE1, and the next hop network address list comprises the loopback address of the SPE1-1, the loopback address of the SPE1-2, the loopback address of the SPE1-3 and the loopback address of the SPE 1-4. And acquiring corresponding output interface lists including Intf1, Intf 2, Intf 3 and Intf 4 according to the loopback address of the SPE1-1, the loopback address of the SPE1-2, the loopback address of the SPE1-3 and the loopback address of the SPE 1-4. The control board of the PE2 generates a MAC forwarding table entry, the destination address of which is the MAC address of the UE1, and the egress interfaces are Intf1, Intf 2, Intf 3, and Intf 4. The specific step flow is the same as the flow described in the multi-active mode of fig. 4a, please refer to this, and for brevity, will not be described again. Forwarding of known unicast data streams to the UE1 through PW load sharing is achieved through the announcement of PWS full active mode and the issuance of MAC routes.

Taking the scenario of PW access to EVPN shown in fig. 1b as an example, fig. 4b shows a schematic interaction diagram of a method flow for multi-active mode advertisement and load sharing with a link state provided in an embodiment of the present invention. As shown in FIG. 1b, the UPE is accessed to SPE1-1, SPE1-2, SPE1-3 and SPE1-4 through PW1, PW2, PW3 and PW4 respectively, UE1 is accessed to the EVPN through CE1 and then UPE, and the far-end PE is PE 2. PW1, PW2, PW3, and PW4 constitute one PW segment, and PWSI is an identification of the PW segment. As shown in FIG. 3b, the following steps are performed on SPE 1-1:

4b-1101, obtaining the redundant mode of the PWS is a multi-active mode (the obtaining mode is the same as the obtaining mode of the redundant mode of the ES in the part of fig. 2, and the description is omitted here).

4b-1102, acquiring the state of PW1, which may be active state or inactive state (please refer to fig. 2 in detail), for example, the state of PW1 is active state.

4b-1103, generating a first mode advertisement message comprising the multi-active mode, PW1 as active, and the PWSI. The first mode advertisement message is used for advertising a PE2, a part of PW in the PWS can be used for forwarding data stream of known unicast, wherein PW1 can be used for forwarding data stream of known unicast. It should be noted that, when the state of PW1 is inactive, the first mode advertisement message is used to advertise PE2, and a part of PWs in the PWs may be used to forward a data stream of a known unicast, where PW1 may not be used to forward a data stream of a known unicast, but is used as a backup link.

The first mode advertisement message also includes next hop network address 1, and next hop network address 1 is the loopback address (e.g. IP address: 10.10.1.1/32) of SPE 1-1.

4b-1104, sends the first mode advertisement message to PE 2.

Similar steps to the SPE1-1 are performed on the SPE1-2, the SPE1-3 and the SPE1-4, the principle is the same, please refer to the SPE1-1 in detail, and the description is omitted here for brevity.

SPE1-2, SPE1-3, and SPE1-4 respectively obtain the redundant mode of the PWS as a multi-live mode.

SPE1-2, SPE1-3, and SPE1-4 respectively acquire PW2, PW3, and PW4 states, and may be in an active state or an inactive state, for example, PW2 is active, PW3 is inactive, and PW4 is inactive.

SPE1-2 generates a second mode advertisement message including the multi-active mode, PW2 as an active state, and the PWSI. The second mode advertisement message is used for advertising a PE2, a part of PWS in the PWS can be used for forwarding data stream of known unicast, wherein, PW2 can be used for forwarding data stream of known unicast. The second mode advertisement message also includes next hop network address 2, and next hop network address 2 is a loopback address (e.g., IP address: 20.20.1.1/32) of SPE 1-2. SPE1-2 sends the second mode advertisement message to PE 2.

SPE1-3 generates a third mode advertisement message including the multi-active mode, PW3 is inactive, and the PWSI. The third mode advertisement message is used for advertising PE2, and a part of PWS in the PWS can be used for forwarding data stream of known unicast, wherein PW3 can not be used for forwarding data stream of known unicast. The third mode advertisement message also includes next hop network address 3, and next hop network address 3 is a loopback address (e.g., IP address: 30.30.1.1/32) of SPE 1-3. SPE1-3 sends the third mode advertisement message to PE 2.

SPE1-4 generates a fourth mode advertisement message including the multi-active mode, PW4 is inactive, and the PWSI. The fourth mode advertisement message is used for advertising PE2, and a part of PWS in the PWS can be used for forwarding data stream of known unicast, wherein PW4 can not be used for forwarding data stream of known unicast. The fourth mode advertisement message also includes next hop network address 4, and next hop network address 4 is a loopback address (e.g., IP address: 40.40.1.1/32) of SPE 1-4. SPE1-4 sends the fourth mode advertisement message to PE 2.

The following steps are performed on PE 2:

4b-201, receiving from Intf1 said first mode advertisement message from SPE1-1, said first mode advertisement message comprising said multi-active mode, PW1 active state and said PWSI. PE2 confirms that the PWS' redundant mode identified by the PWSI is a multi-active mode, and PW1 connected to SPE1-1 is a bar-active PW in the PWS, and can be used to forward known unicast data streams.

4b-202, receiving from Intf 2 said second mode advertisement message from SPE1-2, said second mode advertisement message comprising said multi-active mode, PW2 active state and said PWSI. PE2 confirms that the PWS' redundant mode identified by the PWSI is a multi-active mode, and PW2 connected to SPE1-2 is a PW in which a bar is active in the PWS, and can be used to forward a data stream known to be unicast.

4b-203, receiving from Intf 3 said third mode advertisement message from SPE1-3, said third mode advertisement message comprising said multi-active mode, PW3 being inactive and said PWSI. PE2 confirms that the PWS' redundant mode identified by the PWSI is a multi-active mode, and PW3 connected to SPE1-3 is a PW whose bar is inactive in the PWS, and may not be used to forward known unicast data streams, but rather is used as a backup link.

4b-204, receiving from Intf 4 said fourth mode advertisement message from SPE1-4, said fourth mode advertisement message comprising said multi-active mode, PW4 being inactive and said PWSI. PE2 confirms that the PWS' redundant mode identified by the PWSI is a multi-active mode, and PW4 connected to SPE1-4 is a PW whose bar is inactive in the PWS, and may not be used to forward known unicast data streams, but rather is used as a backup link.

4b-205 obtaining PW1, PW2, PW3, and PW4 from said first, second, and third mode advertisement messages have the same said PWSI, and thus belong to the same PWS. PE2 also determines that the redundant mode of the PWS is a multi-active mode based on the multi-active modes carried in the first, second, third, and fourth mode advertisement messages, respectively. PE2 further determines that PW1 and PW2 in the PWS are used as links for load sharing forwarding and PW3 and PW4 are used as backup links according to PW1 and PW2 being in active states and PW3 and PW4 being in inactive states. PE2 further generates and stores a mapping relationship between link information and next hop node information according to the next hop network address information carried in the first, second, third, and fourth mode advertisement messages, for example, as shown in table 7. The mapping relationship is used to indicate an association relationship between information of each link in the link segment and information of a next hop node of a forwarding device (e.g., PE2) forwarding the data packet. The link information may include link segment identification, link status, etc. The next hop node information may include an IP address, typically a loopback address, of the next hop node.

Table 7: mapping relation between link information and next hop network address information

In addition, SPE1-1 and SPE1-2 connected to PW1 and PW2 in an active state receive a MAC broadcast packet (for MAC learning) sent by CE1, where a source MAC address of the MAC broadcast packet is a MAC address of UE1, and a destination MAC address is a broadcast address. SPE1-1 and SPE1-2 also generate and send MAC routing messages to PE2, respectively, as follows.

SPE1-1 also performs the steps of:

4b-1105 generating a first MAC routing message that includes the MAC address of the UE 1.

4b-1106, sending the first MAC routing message.

SPE1-2 also performs the steps of:

4b-1205, generating a second MAC routing message that includes the MAC address of the UE 1.

4b-1206, sending the second MAC routing message.

PE2 also performs the steps of:

4b-206, receiving the first MAC routing message sent by the SPE1-1 from the Intf1, wherein the first MAC routing message carries the MAC address of the UE 1. A control plane (e.g., a control board of PE2) of the PE2 generates a MAC forwarding entry (e.g., table 8) according to the MAC address of the UE1 and the mapping relationship (e.g., table 7) between the link information and the next hop network address information, where a destination MAC address of the MAC forwarding entry is the MAC address of the UE1, and outgoing interfaces of the MAC forwarding entry are Intf1 and Intf 2. Intf1 and Intf 2 are obtained by PE2 according to the loopback address of SPE1-1 and the loopback address of SPE1-2 corresponding to PW1 and PW2 in an active state. For the sake of brevity, please refer to the description of the steps 4a-204, which is similar to the above description.

The MAC forwarding entry is used to direct the PE2 to forward a data flow (known as a unicast data flow) destined for the UE1 through Intf1 and Intf 2 load sharing. For example, the packet 1 in the data flow is forwarded from the Intf1, and reaches the SPE1-1 through a tunnel from the PE2 to the SPE1-1, so as to be forwarded to the UE1 by using the PW 1. The packet 2 in the data flow is forwarded from the Intf 2, and reaches the SPE1-2 through the tunnel from the PE2 to the SPE1-2, so as to be forwarded to the UE1 by using the PW 2.

Table 8: MAC forwarding table

Destination MAC	Outlet interface
		MAC Address of UE1	Intf	1、Intf 2

PE2 then also receives the second MAC routing message from SPE1-2 from Intf 2, but this message has no effect on the generation of the load-sharing MAC forwarding table. It should be understood that this is only an example, and PE2 may also receive the second MAC routing message sent by SPE1-2 from Intf 2, and then receive the first MAC routing message sent by SPE1-1 from Intf 1. When receiving the second MAC routing message sent by SPE1-2 from Intf 2, the MAC forwarding table shown in table 8 is generated in the same manner.

From the above, when PE2 receives a MAC routing message sent by any one of the end point SPE1-1 of the active link PW1 and the end point SPE1-2 of the active link PW2, PE2 may completely generate a MAC forwarding table entry for load sharing (without waiting for receiving MAC routing messages sent by end points corresponding to all active links), and may forward data streams using all links PW1 and PW2 that may be used for load sharing, which may implement load sharing faster, improve the effect of load sharing, and make load sharing more balanced.

In one example, the first, second, third, and fourth mode advertisement messages are BGP Update messages that contain ethernet auto-discovery routes. The related description is consistent with the description of the multi-active mode advertisement that does not carry the link status in the above-mentioned part 3a, and the support of PW or PWs for ESI Type extension is consistent with the description of the above-mentioned part 4a, please refer to this description, and will not be described again here. In contrast, the embodiment of the present invention also announces the status of the PW. In a specific embodiment, the implementation is realized by extending the value of the Flags field in the ESI Label Extended Community. For example, one value is added to indicate the multi-active mode and the link status is active (e.g., Flags is 0x02), and another value is added to indicate the multi-active mode and the link status is inactive (e.g., Flags is 0x 03). It should be noted that the values herein are merely examples, and other values are possible, and do not limit the present invention. In another specific embodiment, a link state field may also be newly added to the ESI Label Extended command or other fields in the message format to carry the active state or the inactive state, which is not limited herein.

Fig. 4b illustrates a method for notifying and load sharing a multi-active mode carrying a link state according to an embodiment of the present invention, and based on this, a method for notifying and load sharing a PW single active mode and a full active mode is further described. The PW single active mode and full active mode advertisement are similar to the above multi active mode advertisement method carrying the link status, but the difference is that the type of the redundant mode carried in the mode advertisement message is the single active mode or the full active mode.

In the case of PW single active mode, as shown in fig. 1b, for example, PW1 is in active state, and PW2, PW3 and PW4 are in inactive state. The redundant modes carried in the first, second, third and fourth mode notification messages sent by SPE1-1, SPE1-2, SPE1-3 and SPE1-4 to PE2 are all single active modes. The first mode notification message also carries an active state (indicating that the state of PW1 is active), and the second, third, and fourth mode notification messages also respectively carry an inactive state (indicating that the states of PW2, PW3, and PW4 are inactive). PE2 determines that the PWS is in the single-hop mode based on the received first, second, third, and fourth mode advertisement messages, and generates a mapping of link information to next-hop network address information (as shown in table 9). SPE1-1, which is the endpoint of PW1 in the active state, also sends a MAC routing message (carrying the MAC address of UE 1) to PE 2. After receiving the MAC routing message, PE2 generates an MAC forwarding table according to the single active mode, the MAC address of UE1, and the mapping relationship between the link information and the next hop network address information shown in table 9. The destination address of the MAC forwarding table entry is the MAC address of UE1, and the egress interface is Intf 1. The specific flow of steps is the same as the flow described in the multi-active mode of fig. 4b, please refer to this, and for brevity, will not be described again. Through the announcement of the single active mode of the PWS and the issuance of the MAC route, the known unicast data stream to the UE1 is forwarded through one PW in the PW multi-homing scenario. When PW1 fails, the fault can be switched to any one of PW2, PW3 or PW4, so that backup protection is realized and reliability is improved.

In one particular embodiment, single active mode and link state advertisement are supported by extending the value of the Flags field in the ESI Label Extended Community. For example, a value is added to indicate the single active mode and the link status is active (e.g., Flags is 0x04), and a value is added to indicate the single active mode and the link status is inactive (e.g., Flags is 0x 05). It should be noted that the values herein are merely examples, and other values are possible, and do not limit the present invention.

In another specific embodiment, a link state field may also be newly added to the ESI Label Extended command or other fields in the message format to carry the active state or the inactive state, which is not limited herein.

Table 9: mapping relation between link information and next hop network address information

In the case of PW full active mode, as shown in fig. 1b, for example, PW1, PW2, PW3 and PW4 are all active, and the redundant modes carried in the first, second, third and fourth mode notification messages respectively sent by SPE1-1, SPE1-2, SPE1-3 and SPE1-4 to PE2 are all full active modes. Wherein, the first, second, third and fourth mode notification messages also carry active states respectively (indicating the states of PW1, PW2, PW3 and PW4 as inactive). PE2 determines that the PWS is in full active mode based on the received first, second, third and fourth mode advertisement messages and generates a mapping of link information to next hop network address information (as shown in table 10). SPE1-1, SPE1-2, SPE1-3, and SPE1-4 (which are the endpoints of PW1, PW2, PW3, and PW4, respectively, in an active state) also send first, second, third, and fourth MAC routing messages (carrying the MAC address of UE 1), respectively, to PE 2. After receiving one of the four MAC routing messages, PE2 may generate a MAC routing table entry (generated at the control board of PE2) according to the full active mode, the MAC address of UE1, and the mapping relationship between the link information and the next hop network address information shown in table 10. The destination MAC address of the MAC routing table entry is the MAC address of the UE1, and the next hop network address list comprises the loopback address of the SPE1-1, the loopback address of the SPE1-2, the loopback address of the SPE1-3 and the loopback address of the SPE 1-4. And acquiring corresponding output interface lists including Intf1, Intf 2, Intf 3 and Intf 4 according to the loopback address of the SPE1-1, the loopback address of the SPE1-2, the loopback address of the SPE1-3 and the loopback address of the SPE 1-4. The control board of the PE2 generates a MAC forwarding table entry, the destination address of which is the MAC address of the UE1, and the egress interfaces are Intf1, Intf 2, Intf 3, and Intf 4. The specific flow of steps is the same as the flow described in the multi-active mode of fig. 4b, please refer to this, and for brevity, will not be described again. The method can completely generate the MAC forwarding table entry for load sharing after receiving one MAC routing message, can forward data streams by using all links PW1, PW2, PW3 and PW4 which can be used for load sharing, can realize load sharing more quickly, improves the effect of load sharing, and enables the load sharing to be more balanced.

In one particular embodiment, the fully active mode and link state advertisement are supported by extending the value of the Flags field in the ESI Label Extended Community. For example, one value is added to indicate the fully active mode and the link status is active (for example, the flag value is 0x046, and another value is added to indicate the fully active mode and the link status is inactive (for example, the flag value is 0x 07).

Table 10: mapping relation between link information and next hop network address information

It should be understood that, as described above for the scenario of fig. 1a and 1b, the first PE device may be a PE device (e.g., PE1-1, PE1-2, PE1-3) in an ethernet link access EVPN networking (fig. 1a) or an SPE device (e.g., SPE1-1, SPE1-2, SPE1-3, SPE1-4) in a PW access EVPN networking (fig. 1b) according to different networking situations. The far-end PE device is a PE device, such as PE2 in fig. 1a and 1 b. In essence, the first PE device and the remote PE device are a network device, such as a router or a switch.

Fig. 5a shows a schematic diagram of a possible structure of the first PE device involved in the above embodiment. The first PE device 500A includes: master control board 510, interface board 530, switch board 520, and interface board 540. The main control board 510 is used to complete functions of system management, device maintenance, protocol processing, and the like. The switch board 520 is used to complete data exchange between interface boards (interface boards are also called line cards or service boards). Interface boards 530 and 540 are used to provide various service interfaces (e.g., POS interface, GE interface, ATM interface, etc.) and to implement forwarding of data packets. The main control board 510, the interface boards 530 and 540, and the switch board 520 are connected to the system backplane through the system bus for communication. The central processor 531 on the interface board 530 is used for controlling and managing the interface board and communicating with the central processor on the main control board.

The central processing unit 511 on the main control board 510 is configured to obtain the redundant mode of the link segment, generate a mode notification message, and send the mode notification message to the interface board 530 or 540. When the redundancy mode of the link segment is a multi-active mode, the mode advertisement message includes information indicating that the redundancy mode of the link segment is the multi-active mode and an identifier of the link segment, where the identifier of the link segment is used for uniquely identifying the link segment, and the multi-active mode indicates that a part of links in the link segment can be used for forwarding a data stream, and the number of the part of links is greater than 1 and less than the maximum number of links in the link segment. The physical interface card 533 on the interface board 530 is configured to send the mode advertisement message to the remote PE device.

Based on that the central processing unit 511 on the main control board 510 is configured to obtain the redundancy mode of the link segment, in a specific embodiment, the central processing unit 511 on the main control board 510 is further configured to generate a MAC routing message, and issue the MAC routing message to the interface board 530 or 540. The MAC routing message includes a destination MAC address and a next hop network address, where the destination MAC address is an MAC address of a terminal device accessing the user side device, and the next hop network address is a network address of the first PE device. The physical interface card 533 on the interface board 530 is further configured to send the MAC routing message to the remote PE device.

Based on that the central processing unit 511 on the main control board 510 is configured to obtain the redundancy mode of the link segment, in another specific embodiment, when the redundancy mode of the link segment is a multi-active mode, the central processing unit 511 is further configured to obtain a state of a first link between the ue and the first PE device, where the state of the first link is active or inactive. Correspondingly, the mode advertisement message further includes a state of the first link and a next hop network address, where the next hop network address is a network address of the first PE device. The central processor 511 is further configured to generate a MAC routing message, and issue the MAC routing message to the interface board 530 or 540. The MAC routing message comprises a destination MAC address, and the destination MAC address is the MAC address of the terminal equipment accessed to the user side equipment.

The physical interface card 533 on the interface board 530 is further configured to send the MAC routing message to the remote PE device.

Based on that the central processing unit 511 on the main control board 510 is further configured to obtain a state of a first link between the user-side device and the first PE device, in a specific embodiment, the link is an ethernet link, the link section is an ethernet section ES, a redundancy mode of the link section is a redundancy mode of the ES, and the first link is a first ethernet link. When the state of the first ethernet link is active, the mode advertisement message is used to advertise the far-end PE device, and a part of ethernet links in the ES may be used to forward data streams, where the first ethernet link may be used to forward data streams. Or, when the state of the first ethernet link is inactive, the mode advertisement message is used to advertise the remote PE device, and a part of ethernet links in the ES may be used to forward a data stream, where the first ethernet link may not be used to forward a data stream.

Based on that the central processing unit 511 on the main control board 510 is further configured to obtain a state of a first link between the user side device and the first PE device, in another specific embodiment, the link is a pseudo wire PW, the link segment is a pseudo wire segment PWs, a redundancy mode of the link segment is a redundancy mode of the PWs, and the first link is a first PW. When the state of the first PW is active, the mode advertisement message is used to advertise the far-end PE device, and a part of PWs in the PWs may be used to forward a data flow, where the first PW may be used to forward a data flow. Or when the state of the first PW is inactive, the mode advertisement message is used to advertise the far-end PE device, and a part of PWs in the PWs may be used to forward a data flow, where the first PW may not be used to forward a data flow.

Based on the central processing unit 511 on the main control board 510 being configured to obtain the redundancy mode of the link segment, in a specific embodiment, the link is a pseudo wire PW, the link segment is a pseudo wire segment PWs, and the redundancy mode of the link segment is the redundancy mode of the PWs. When the redundancy mode of the PWS is a full active mode, the mode advertisement message includes information indicating that the redundancy mode of the PWS is the full active mode and an identification of the PWS, the identification of the PWS being used to uniquely identify the PWS, the full active mode indicating that all PWs in the PWS can be used to forward data streams.

Based on the central processing unit 511 on the main control board 510 being configured to obtain the redundancy mode of the link segment, in another specific embodiment, the link is a pseudo wire PW, the link segment is a pseudo wire segment PWs, and the redundancy mode of the link segment is the redundancy mode of the PWs. When the redundancy mode of the PWS is a single active mode, the mode advertisement message comprises information indicating that the redundancy mode of the PWS is the single active mode and an identification of the PWS, wherein the identification of the PWS is used for uniquely identifying the PWS, and the single active mode represents that only one PW in the PWS can be used for forwarding data streams.

It should be understood that the operations on the interface board 540 in the embodiment of the present invention are the same as the operations on the interface board 530, and therefore, for brevity, detailed descriptions are omitted. It should be understood that the first PE device 500A of this embodiment may correspond to the first PE device in the embodiment corresponding to fig. 1a to 4c, and the main control board 510, the interface board 530 and/or 540 in the first PE device 500A may implement the functions and/or various steps implemented by the first PE device in the embodiment corresponding to fig. 1a to 4c, and are not described herein again for brevity. In addition, in this embodiment and the following embodiments, only the first PE device in the link segment is taken as an example for description, and the second PE device, the third PE device, or more PE devices in the link segment have the same functions as the first PE device, which is not described herein again.

In addition, it should be noted that there may be one or more main control boards, and when there are multiple main control boards, the main control board may include an active main control board and a standby main control board. The interface board may have one or more blocks, and the more interface boards are provided the stronger the data processing capability of the first PE device is. There may also be one or more physical interface cards on an interface board. The exchange network board may not have one or more blocks, and when there are more blocks, the load sharing redundancy backup can be realized together. Under the centralized forwarding architecture, the first PE device may not need a switching network board, and the interface board undertakes the processing function of the service data of the entire system. Under the distributed forwarding architecture, the first PE device may have at least one switching network board, and implement data exchange between multiple interface boards through the switching network board, so as to provide high-capacity data exchange and processing capabilities. Therefore, the data access and processing capabilities of the first PE device in the distributed architecture are greater than those of the centralized architecture. Which architecture is specifically adopted depends on the specific networking deployment scenario, and is not limited herein.

Fig. 5b shows a schematic structural diagram of another implementation manner of the first PE device in the foregoing embodiment. This product form of the first PE device 500B is suitable for Network architectures (e.g., Software Defined Network (SDN)) based on separation of control and forwarding. In an SDN, the main control board 510 of the first PE device 500A shown in fig. 5a is separated from the devices to form a new independent physical device (i.e., the controller 1510 shown in fig. 5 b), and the rest forms another independent physical device (i.e., the first PE forwarding device 1500 shown in fig. 5 b). The controller 1510 interacts with the first PE forwarding device 1500 via a control channel protocol. The control channel Protocol may be an open flow (OpenFlow) Protocol, a Path Computation Communication Protocol (PCEP), a Border Gateway Protocol (BGP), a Routing System Interface (I2 RS), and the like. That is, compared to the embodiment corresponding to the above-mentioned embodiment of fig. 5a, the first PE device 500B in this embodiment includes the separated controller 1510 and the first PE forwarding device 1500, that is, in this embodiment, the first PE device 500B can also be regarded as a system.

The controller 1510 may be implemented based on a general-purpose physical server or a dedicated hardware structure, and in one design example, includes a receiver, a processor, a transmitter, a random access memory, a read only memory, and a bus (not shown). The processor is coupled to the receiver, the transmitter, the random access memory and the read only memory through the bus respectively. When the controller needs to be operated, the basic input and output system solidified in the read-only memory or the bootloader guide system in the embedded system is started, and the guide controller enters a normal operation state. After the controller enters the normal operation state, the application program and the operating system are run in the random access memory, so that the processor performs all the functions and steps of the main control board 510 in fig. 5 a.

First PE forwarding device 1500 may be implemented based on a dedicated hardware structure, whose functions and structures are consistent with those of interface board 530, interface board 540, and switch board 520 in fig. 5a, and execute corresponding functions and steps. Or the virtual first PE forwarding device may be implemented based on a general physical server and Network Function Virtualization (NFV) technology, and the virtual first PE forwarding device is a virtual router. In a scenario of the virtual first PE forwarding device, the first PE forwarding device mentioned in the foregoing embodiment of the entity first PE forwarding device includes an interface board, a switch board, and a processor, which may be considered as an interface resource, a network resource, and a processing resource that are allocated to the virtual first PE forwarding device based on a general physical server in a virtual environment. The following embodiment of fig. 5d may be specifically referred to as the following embodiment, where the function or step of the first forwarding PE device is implemented by using a generic physical server, or the function or step of the first forwarding PE device is implemented by using a generic physical server and using an NFV technology.

It should be understood that, in this embodiment, the controller 1510 and the first PE forwarding device 1500 in the first PE device 500B may implement various functions and steps implemented by the first PE device in the embodiments corresponding to fig. 1a to fig. 4c and fig. 5a, and for brevity, no further description is provided here.

Fig. 5c shows a schematic structural diagram of another implementation manner of the first PE device in the foregoing embodiment. The first PE device 500C includes: a processing unit 504C and a transmitting unit 506C.

The processing unit 504C is configured to acquire the redundant mode of the link segment, and generate a mode notification message. When the redundancy mode of the link segment is a multi-active mode, the mode advertisement message includes information indicating that the redundancy mode of the link segment is the multi-active mode and an identifier of the link segment, where the identifier of the link segment is used for uniquely identifying the link segment, and the multi-active mode indicates that a part of links in the link segment can be used for forwarding a data stream, and the number of the part of links is greater than 1 and less than the maximum number of links in the link segment. The sending unit 506C is configured to send the mode notification message to the far-end PE device.

Based on the processing unit 504C being configured to obtain the redundancy pattern of the link segment, in a specific embodiment, the processing unit 504C is further configured to generate a MAC routing message. The MAC routing message includes a destination MAC address and a next hop network address, where the destination MAC address is an MAC address of a terminal device accessing the user side device, and the next hop network address is a network address of the first PE device. The sending unit 506C is further configured to send the MAC routing message to the remote PE device.

Based on that the processing unit 504C is configured to obtain the redundancy mode of the link segment, in yet another specific embodiment, when the redundancy mode of the link segment is a multi-active mode, the processing unit 504C is further configured to obtain a state of a first link between the user-side device and the first PE device, where the state of the first link is active or inactive. Correspondingly, the mode advertisement message further includes a state of the first link and a next hop network address, where the next hop network address is a network address of the first PE device. Processing unit 504C is also configured to generate a MAC routing message. The MAC routing message comprises a destination MAC address, and the destination MAC address is the MAC address of the terminal equipment accessed to the user side equipment. The sending unit 506C is further configured to send the MAC routing message to the remote PE device.

Based on that the processing unit 504C is further configured to obtain a state of a first link between the user-side device and the first PE device, in a specific implementation manner, the link is an ethernet link, and specifically, the functions and/or various steps implemented by the first PE device in the embodiments corresponding to fig. 1a, fig. 3b, fig. 5a, and fig. 5b may be implemented.

Based on that the processing unit 504C is further configured to obtain a state of a first link between the user-side device and the first PE device, in another specific implementation manner, the link is a pseudo wire PW, and specifically, the functions and/or various steps implemented by the first PE device in the embodiments corresponding to fig. 1b, fig. 4b, fig. 5a, and fig. 5b may be implemented.

When the link is a pseudo wire PW, the link segment is a pseudo wire segment PWs, and the redundancy mode of the link segment is a redundancy mode of the PWs, based on that the processing unit 504C is configured to obtain the redundancy mode of the link segment, the redundancy mode may also be a full-active mode or a single-active mode, and specifically, the functions and/or various steps implemented by the first PE device in the embodiments corresponding to fig. 1b, fig. 4a, fig. 4b, fig. 5a, and fig. 5b may be implemented. .

The first PE device 500C according to the embodiment of the present invention may implement various implementation functions and steps in the first PE device in the embodiments corresponding to fig. 1a to fig. 5b, and for brevity, no further description is given here.

Fig. 5d shows a schematic structural diagram of another implementation manner of the first PE device in the foregoing embodiment. In this embodiment, the functions or operational steps of the first PE device (i.e., the primality functions or operational steps of the first PE described in the above embodiments) are implemented by one or more processors in a general purpose computer or server by executing program code in memory. In this embodiment, the first PE device 500D includes: receiver 510D, processor 520D, transmitter 530D, random access memory 540D, read only memory 550D, and bus 560D. The processor 520D is coupled to the receiver 510D, the transmitter 530D, the random access memory 540D, and the read only memory 550D through the bus 560D. When the first PE device 500D needs to be operated, the first PE device 500D is booted to enter a normal operation state by booting the bootloader boot system in the basic input/output system or the embedded system, which is fixed in the rom 550D. After the first PE device 500D enters the normal operating state, the application and operating system running in the random access memory 540D cause:

processor 520D is configured to obtain a redundancy pattern for the link segment and generate a pattern advertisement message. The transmitter 530D is configured to transmit the mode advertisement message to the remote PE device.

The first PE device 500D according to an embodiment of the present invention may correspond to the first PE device in the embodiment corresponding to fig. 1a to 4c, and the processor 520D, the transmitter 530D, and the like in the first PE device 500D may implement the functions of the first PE device and/or various steps and methods implemented in the embodiment corresponding to fig. 1a to 4 c. The processor 520D is configured to perform all operations of the processing unit 504C of the first PE device illustrated in fig. 5C, and the transmitter 530D is configured to perform all operations of the transmitting unit 506C of the first PE device illustrated in fig. 5C. For brevity, no further description is provided herein.

It should be noted that, in this embodiment, a virtual first PE device implemented by combining a general physical server with a network function virtualization NFV technology may also be based on, where the virtual first PE device is a virtual router, and may virtualize a second, third, and N PE (according to actual needs) PE devices. The Virtual first PE device may be a Virtual Machine (VM) running a program for providing a messaging function, the VM being deployed on a hardware device (e.g., a physical server). A virtual machine refers to a complete computer system with complete hardware system functionality, which is emulated by software, running in a completely isolated environment. Through reading the application, a person skilled in the art can combine the NFV technology to virtually generate a plurality of PE devices with the above functions on a common physical server. And will not be described in detail herein.

Fig. 6a shows a schematic structural diagram of an implementation of the remote PE device involved in the foregoing embodiment. The far-end PE device 600A includes: main control board 610, interface board 630, switch network board 620 and interface board 640. The main control board 610 is used to complete functions such as system management, device maintenance, and protocol processing. The switch network board 620 is used to complete data exchange between interface boards (interface boards are also called line cards or service boards). The interface boards 630 and 640 are used to provide various service interfaces (e.g., POS interface, GE interface, ATM interface, etc.) and implement fast forwarding of data packets. The main control board 610, the interface boards 630 and 640, and the switch board 620 are connected to the system backplane through the system bus for intercommunication. The central processor 631 on the interface board 630 is used for control management of the interface board and communication with the central processor on the main control board.

A physical interface card 633 on the interface board 630 is configured to receive a first mode advertisement message sent from the first PE device, where the first mode advertisement message includes a redundancy mode of the link segment and an identifier of the link segment, and the identifier of the link segment is used to uniquely identify the link segment;

the central processing unit 611 on the main control board 610 is configured to obtain the identifier of the link segment, and when the redundancy mode of the link segment is the multi-active mode, determine, according to the multi-active mode, that a part of links in the link segment identified by the identifier of the link segment can be used for forwarding data streams, where the number of the part of links is greater than 1 and less than the maximum number of links in the link segment.

In a specific embodiment, the at least three PE devices further include a second PE device, where the physical interface card 633 on the interface board 630 is configured to receive the first mode advertisement message sent from the first PE device. When the state of the first link between the first PE device and the user side device and the state of the second link between the second PE device and the user side device are both active, the physical interface card 633 is further configured to receive a first MAC routing message and a second MAC routing message sent by the first PE device and the second PE device, respectively. The first MAC routing message includes a destination MAC address and a first next hop network address, and the second MAC routing message includes the destination MAC address and a second next hop network address. The destination MAC address is an MAC address of a terminal device accessing the user side device, the first next hop network address is a network address of the first PE device, and the second next hop network address is a network address of the second PE device.

The central processor 611 on the main control board 610 is further configured to determine, according to the multi-active mode in the first mode advertisement message, the identifier of the link segment, the first MAC routing message, and the second MAC routing message, that the first link and the second link can forward the data stream destined to the destination MAC in a load sharing manner in the link segment.

Specifically, the central processing unit 611 on the main control board 610 generates an MAC forwarding entry according to the multi-active mode, the identifier of the link segment, the destination MAC address, the first next hop network address, and the second next hop network address in the first mode notification message, where the MAC forwarding entry includes the destination MAC address and an outgoing interface list, the outgoing interface list includes a first outgoing interface of the remote PE device and a second outgoing interface of the remote device, the first outgoing interface is obtained according to the first next hop network address, and the second outgoing interface is obtained according to the second next hop network address. The central processor 611 on the main control board 610 sends the data to the forwarding table entry storage 634 on the interface board 630 for storage through the central processor 631 on the interface board 630. The MAC forwarding table entry includes the destination MAC address and an outgoing interface list, where the outgoing interface list includes a first outgoing interface (for example, one interface located on the physical interface card 633) of the far-end PE device and a second outgoing interface (for example, another interface located on the physical interface card 633) of the far-end PE device, and the network processor 632 on the interface board 630 searches for the MAC forwarding table entry, and forwards the data stream load destined for the destination MAC from the first outgoing interface and the second outgoing interface.

In another specific embodiment, based on that the physical interface card 633 on the interface board 630 is configured to receive the first mode advertisement message sent from the first PE device, the first mode advertisement message further includes a state of the first link, where the state of the first link is active or inactive. The central processor 611 on the main control board 610 is further configured to determine whether the first link between the first PE device and the user-side device can be used for forwarding a data flow according to the multi-active mode, the identifier of the link segment, and the state of the first link.

Further, the at least three PE devices further include a second PE device, and the physical interface card 633 is further configured to receive a second mode advertisement message sent by the second PE device, where the second mode advertisement message includes that the redundancy mode of the link segment is a multi-active mode, the state of the second link is active, and the link segment identifier. The central processor 611 on the main control board 610 is further configured to determine, according to the multi-active mode, the identifier of the link segment, and the state of the second link in the second mode advertisement message, that the second link between the second PE device and the user-side device is available for forwarding a data flow. When the state of the first link is active, the processing unit is further configured to determine that the first link and the second link can forward the data stream in load sharing in the link segment.

The first mode advertisement message further includes a first next hop network address, where the first next hop network address is a network address of the first PE device. The second mode advertisement message further includes a second next hop network address, where the second next hop network address is a network address of the second PE device. When the state of the first link is active, the central processing unit 611 on the main control board 610 is further configured to generate a mapping relationship between link information and next hop network address information according to the multi-active mode, the link segment identifier, the state of the first link is active, the state of the second link is active, the first next hop network address and the second next hop network address, where the mapping relationship indicates a forwarding data stream that can be load-shared by the first link and the second link in the link segment identified by the link segment identifier, the data stream reaches the first link through the first PE device identified by the first next hop network address, and the data stream reaches the second link through the second PE device identified by the second next hop network address.

Correspondingly, the physical interface card 633 is further configured to receive a MAC routing message from the first PE device, where the MAC routing message includes a destination MAC address, and the destination MAC address is a MAC address of an end device accessing the user-side device. The central processor 611 on the main control board 610 is further configured to generate a MAC forwarding table according to the destination MAC address and the mapping relationship, and send the MAC forwarding table to the forwarding table storage 634 on the interface board 630 through the central processor 631 on the interface board 630 for storage. The MAC forwarding table entry includes the destination MAC address and an outgoing interface list, where the outgoing interface list includes a first outgoing interface (for example, one interface located on the physical interface card 633) of the far-end PE device and a second outgoing interface (for example, another interface located on the physical interface card 633) of the far-end PE device, and the network processor 632 on the interface board 630 searches for the MAC forwarding table entry, and forwards the data stream load destined for the destination MAC from the first outgoing interface and the second outgoing interface. The first output interface is obtained according to the first next hop network address, and the second output interface is obtained according to the second next hop network address.

In a specific embodiment, the link is an ethernet link, the link section is an ethernet segment ES, the redundancy mode of the link section is a redundancy mode of the ES, and the first link is a first ethernet link. When the state of the first ethernet link is active, the central processor 611 on the main control board 610 is further configured to determine that the first ethernet link is an active link in the ES, which may be used to forward a data stream. Or, when the state of the first ethernet link is inactive, the central processor 611 on the main control board 610 is further configured to determine that the first ethernet link is an inactive link in the ES and cannot be used to forward a data stream.

In another specific embodiment, the link is a pseudo wire PW, the link segment is a pseudo wire segment PWs, the redundancy mode of the link segment is a redundancy mode of the PWs, and the first link is a first PW. When the state of the first PW is active, the central processor 611 on the main control board 610 is further configured to determine that the first PW is an active link in the PWs, and may be used to forward a data stream. Or, when the state of the first PW is inactive, the central processor 611 on the main control board 610 is further configured to determine that the first PW is not available for forwarding a data stream as an inactive link in the PWs.

Based on that a physical interface card 633 on an interface board 630 is configured to receive a first mode advertisement message sent from the first PE device, the link is a pseudo wire PW, the link segment is a pseudo wire segment PWs, the redundancy mode of the link segment is a redundancy mode of the PWs, and the identifier of the link is an identifier of the PWs.

In a specific embodiment, when the redundancy mode of the PWS is the full active mode, the central processing unit 611 on the main control board 610 is further configured to confirm that all PWS in the PWS identified by the identity of the PWS can be used for forwarding data stream according to the full active mode.

In another specific embodiment, the mode advertisement message further includes status information of the first PW, and when the redundant mode of the PWs is a fully active mode and the status of the first PW is active, the central processor 611 on the main control board 610 is further configured to confirm that all PWs in the PWs identified by the identification of the PWs can be used for forwarding a data stream.

In another specific embodiment, when the redundancy mode of the PWS is the single active mode, the central processing unit 611 on the main control board 610 is further configured to confirm that only one PW of the PWS identified by the identity of the PWS can be used for forwarding a data stream according to the single active mode.

In yet another specific embodiment, the mode advertisement message further includes status information of the first PW, and when the redundant mode of the PWs is single active mode and the status of the first PW is active, the central processor 611 on the main control board 610 is further configured to determine that the first PW identified by the identification of the PWs in the PWs can be used to forward a data stream.

It should be understood that the operations on the interface board 640 in the embodiment of the present invention are the same as the operations of the interface board 630, and therefore, for brevity, detailed descriptions are omitted. It should be understood that the far-end PE device 600A in the embodiment of the present invention may correspond to the far-end PE device in the embodiment corresponding to fig. 1a to 4c, and the main control board 610, the interface board 630, and/or 640 in the far-end PE device 600A may implement the functions and/or various steps implemented by the far-end PE device in the embodiment corresponding to fig. 1a to 4c, which are not described herein again for brevity. In addition, in this embodiment and the following embodiments, only the first PE device in the link segment is taken as an example for description, and the second PE device, the third PE device, or more PE devices in the link segment have the same functions as the first PE device, which is not described herein again.

In addition, it should be noted that there may be one or more main control boards, and when there are multiple main control boards, the main control board may include an active main control board and a standby main control board. The interface board may have one or more blocks, and the more interface boards are provided the stronger the data processing capability of the remote PE device is. There may also be one or more physical interface cards on an interface board. The exchange network board may not have one or more blocks, and when there are more blocks, the load sharing redundancy backup can be realized together. Under the centralized forwarding architecture, the far-end PE device may not need a switching network board, and the interface board undertakes the processing function of the service data of the entire system. Under the distributed forwarding architecture, the far-end PE device may have at least one switching network board, and data exchange between a plurality of interface boards is realized through the switching network board, thereby providing a large capacity of data exchange and processing capability. Therefore, the data access and processing capabilities of the remote PE devices in the distributed architecture are greater than those of the centralized architecture. Which architecture is specifically adopted depends on the specific networking deployment scenario, and is not limited herein.

Fig. 6b shows another possible structure diagram of the remote PE device involved in the above embodiment. This product form of the remote PE device 600B is applicable to a network architecture (e.g., SDN) based on separation of control and forwarding. In an SDN, the main control board 610 of the far-end PE device 600A shown in fig. 6a is separated from the devices to form a new independent physical device (i.e., the controller 1610 shown in fig. 6 b), and the rest forms another independent physical device (i.e., the far-end PE forwarding device 1600 shown in fig. 6 b). The controller 1610 interacts with the remote PE forwarding device 1600 via a control channel protocol. The control channel protocol may be OpenFlow protocol, PCEP, BGP, I2RS, etc. That is, compared to the embodiment corresponding to the above-mentioned embodiment of fig. 5a, the first PE device 500B in this embodiment includes the separated controller 1510 and the first PE forwarding device 1500, that is, in this embodiment, the first PE device 500B can also be regarded as a system.

The controller 1610 may be implemented based on a general-purpose physical server or a dedicated hardware structure, and in one design example includes a receiver, a processor, a transmitter, a random access memory, a read only memory, and a bus (not shown). The processor is coupled to the receiver, the transmitter, the random access memory and the read only memory through the bus respectively. When the controller needs to be operated, the basic input and output system solidified in the read-only memory or the bootloader guide system in the embedded system is started, and the guide controller enters a normal operation state. After the controller enters the normal operation state, the application program and the operating system are run in the random access memory, so that the processor performs all the functions and steps of the main control board 610 in fig. 6 a. In another specific embodiment, the controller shown in fig. 6b may also be a virtual controller implemented based on a general physical server and NFV technology, and the virtual controller may be a VM running a program for providing a function of receiving a message, and the VM is deployed on a hardware device (e.g., a physical server). This virtual controller performs all of the functions and operations of the controller 1610 shown in FIG. 6 b.

The far-end PE forwarding device 1600 may be implemented based on a dedicated hardware structure, and its functions and structures are consistent with those of the interface board 630, the interface board 640 and the switch network board 620 in fig. 6a, and execute corresponding functions and steps. Or the PE device may be a virtual far-end PE device implemented based on a general physical server and NFV technology, where the virtual far-end PE device is a virtual router. In a scenario of a virtual far-end PE device, the far-end PE forwarding device mentioned in the foregoing embodiment of the entity far-end PE forwarding device includes an interface board, a switch board, and a processor, which may be considered as an interface resource, a network resource, and a processing resource that are allocated to the virtual far-end PE forwarding device by a general physical server based on the processor in a virtual environment. The following embodiment of fig. 6d may be specifically referred to as the following embodiment, wherein the functions or steps of the remote PE device are implemented by using a common physical server, or the functions or steps of the remote PE device are implemented by using a common physical server and using an NFV technology.

It should be understood that, in this embodiment, the controller 1610 and the remote PE forwarding device 1600 in the remote PE device 600B may implement various functions and steps implemented by the remote PE device in the embodiments corresponding to fig. 1a to fig. 4c and fig. 6a, and for brevity, no further description is provided here.

Fig. 6c shows a schematic diagram of another possible structure of the remote PE device involved in the above embodiment. The far-end PE device 600C includes: a receiving unit 602C and a processing unit 604C.

The receiving unit 602C is configured to receive a first mode notification message sent by the first PE device, where the first mode notification message includes a redundancy mode of the link segment and an identifier of the link segment, and the identifier of the link segment is used to uniquely identify the link segment;

the processing unit 604C is configured to obtain the identifier of the link segment, and when the redundancy mode of the link segment is the multi-active mode, determine, according to the multi-active mode, that a part of links in the link segment identified by the identifier of the link segment may be used for forwarding a data stream, where the number of the part of links is greater than 1 and less than the maximum number of links in the link segment.

In a specific embodiment, the at least three PE devices further include a second PE device, based on the receiving unit 602C being configured to receive the first mode advertisement message sent by the first PE device. When the state of the first link between the first PE device and the user side device and the state of the second link between the second PE device and the user side device are both active, the receiving unit 602C is further configured to receive a first MAC routing message and a second MAC routing message sent by the first PE device and the second PE device, respectively. The description of the first MAC routing message and the second MAC routing message is consistent with the embodiment shown in fig. 6a, and is not repeated.

The processing unit 604C is further configured to determine, according to the multi-active mode in the first mode advertisement message, the identifier of the link segment, the first MAC routing message, and the second MAC routing message, that the first link and the second link can forward the data flow destined for the destination MAC in a load sharing manner in the link segment.

Optionally, the processing unit 604C generates an MAC forwarding entry according to the multi-active mode, the identifier of the link segment, the destination MAC address, the first next hop network address, and the second next hop network address in the first mode notification message, where the MAC forwarding entry includes the destination MAC address and an outgoing interface list, the outgoing interface list includes a first outgoing interface of the far-end PE device and a second outgoing interface of the far-end device, the first outgoing interface is obtained according to the first next hop network address, and the second outgoing interface is obtained according to the second next hop network address. The far-end PE device further comprises a storage unit, and the storage unit is used for storing the MAC forwarding table entry. The far-end PE device further comprises a sending unit, and the sending unit forwards the data stream load destined to the destination MAC from the first outgoing interface and the second outgoing interface in a sharing manner.

Based on the receiving unit 602C being configured to receive the first mode advertisement message sent from the first PE device, in another specific embodiment, the first mode advertisement message further includes a status of a first link, and the status of the first link is active or inactive. The processing unit 604C is further configured to determine whether the first link between the first PE device and the user-side device is available for forwarding a data flow according to the multi-active mode, the identification of the link segment, and the state of the first link.

Further, the at least three PE devices further include a second PE device, and the receiving unit 602C is further configured to receive a second mode notification message sent by the second PE device, where the second mode notification message includes that the redundancy mode of the link segment is a multi-active mode, the state of the second link is active, and the link segment identifier. The processing unit 604C is further configured to determine that the second link between the second PE device and the user-side device is available for forwarding a data flow according to the multi-active mode, the identification of the link segment, and the state of the second link in the second mode advertisement message. When the state of the first link is active, the processing unit is further configured to determine that the first link and the second link can forward the data stream in load sharing in the link segment.

The first mode advertisement message further includes a first next hop network address, where the first next hop network address is a network address of the first PE device. The second mode advertisement message further includes a second next hop network address, where the second next hop network address is a network address of the second PE device. When the state of the first link is active, the processing unit 604C is further configured to generate a mapping relationship between link information and next hop network address information according to the multi-active mode, the link segment identifier, the state of the first link is active, the state of the second link is active, the first next hop network address and the second next hop network address, where the mapping relationship indicates a forwarding data stream that can be load-shared by the first link and the second link in the link segment identified by the link segment identifier. The data stream reaches the first link through the first PE device identified by the first next hop network address, and the data stream reaches the second link through the second PE device identified by the second next hop network address.

Correspondingly, the receiving unit 602C is further configured to receive a MAC routing message from the first PE device, where the MAC routing message includes a destination MAC address, and the destination MAC address is a MAC address of a terminal device accessing the user-side device. The processing unit 604C is further configured to generate an MAC forwarding entry according to the destination MAC address and the mapping relationship. The far-end PE device 600C further includes a storage unit, and the storage unit is configured to store the MAC forwarding table entry. The MAC forwarding table entry includes the destination MAC address and an outgoing interface list, where the outgoing interface list includes a first outgoing interface of the remote PE device and a second outgoing interface of the remote PE device, and the processing unit 604C is further configured to search the MAC forwarding table entry to obtain the first outgoing interface and the second outgoing interface. The far-end PE device further includes a sending unit, configured to forward the data stream load destined for the destination MAC to the first outgoing interface and the second outgoing interface. The first output interface is obtained according to the first next hop network address, and the second output interface is obtained according to the second next hop network address.

In a specific embodiment, the link is an ethernet link, the link section is an ethernet segment ES, the redundancy mode of the link section is a redundancy mode of the ES, and the first link is a first ethernet link. When the state of the first ethernet link is active, the processing unit 604C is further configured to determine that the first ethernet link is an active link in the ES and may be used to forward a data stream. Or, when the state of the first ethernet link is inactive, the processing unit 604C is further configured to determine that the first ethernet link is an inactive link in the ES and cannot be used to forward a data stream.

In another specific embodiment, the link is a pseudo wire PW, the link segment is a pseudo wire segment PWs, the redundancy mode of the link segment is a redundancy mode of the PWs, and the first link is a first PW. When the state of the first PW is active, processing unit 604C is further configured to determine the first PW to be an active link in the PWs, which can be used for forwarding a data flow. Alternatively, when the state of the first PW is inactive, processing unit 604C is further configured to determine that the first PW is not available for forwarding a data stream as an inactive link in the PWs.

Based on the receiving unit 602C, configured to receive a first mode advertisement message sent from the first PE device, where the link is a pseudo wire PW, the link segment is a pseudo wire segment PWs, a redundancy mode of the link segment is a redundancy mode of the PWs, and an identifier of the link is an identifier of the PWs.

In a specific embodiment, when the redundant mode of the PWS is an all-active mode, the processing unit 604C is further configured to confirm that all PWS in the PWS identified by the identification of the PWS are available for forwarding data streams according to the all-active mode.

In another specific embodiment, the mode advertisement message further includes status information of the first PW, and when the redundant mode of the PWs is an all-active mode and the status of the first PW is active, the processing unit 604C is further configured to confirm that all PWs in the PWs identified by the identification of the PWs can be used for forwarding a data stream.

In yet another specific embodiment, when the redundancy mode of the PWS is a single active mode, the processing unit 604C is further configured to confirm that only one PW of the PWS identified by the identification of the PWS is available for forwarding a data stream according to the single active mode.

In yet another specific embodiment, the mode advertisement message further includes status information of a first PW, and when the redundant mode of the PWs is a single active mode and the status of the first PW is active, the processing unit 604C is further configured to determine that the first PW of the PWs identified by the identification of the PWs can be used to forward a data flow.

The far-end PE device 600C in the embodiment of the present invention may implement various implementation functions and steps in the far-end PE device in the embodiments corresponding to fig. 1a to fig. 6b, and for brevity, no further description is given here.

Fig. 6d shows a schematic diagram of still another possible structure of the remote PE device involved in the above embodiment. In this embodiment, the functions or operational steps of the first PE device (i.e., the primality functions or operational steps of the first PE described in the above embodiments) are implemented by one or more processors in a general purpose computer or server by executing program code in memory. In this embodiment, the remote PE device 600D includes: receiver 610D, processor 620D, transmitter 630D, random access memory 640D, read only memory 650D, and bus 660D. The processor 620D is coupled to the receiver 610D, the transmitter 630D, the random access memory 640D and the read only memory 650D through the bus 660D. When the remote PE device 600D needs to be operated, the remote PE device 600D is booted to enter a normal operation state by starting a basic input/output system solidified in the rom 650D or a bootloader boot system in the embedded system. After the far-end PE device 600D enters a normal operating state, the application and operating system running in the random access memory 640D causes:

the receiver 610D is configured to receive a first mode advertisement message sent from the first PE device.

The processor 620D is configured to obtain the identifier of the link segment, and when the redundancy mode of the link segment is the multi-active mode, determine, according to the multi-active mode and the identifier of the link segment, that a part of links in the link segment identified by the identifier of the link segment can be used for forwarding a data stream, where the number of the part of links is greater than 1 and less than the maximum number of links in the link segment.

The remote PE device 600D according to the embodiment of the present invention may implement various functions and steps implemented by the remote PE device in the embodiments corresponding to fig. 1a to fig. 4 c. The processor 620D, the transmitter 630D, and the like in the far-end PE device 600D may implement the functions of and/or various steps and methods implemented by the far-end PE device in the embodiments corresponding to fig. 1a to 4 c. The processor 620D is configured to perform all operations of the processing unit 604C of the remote PE device in fig. 6C, and the receiver 610D is configured to perform all operations of the receiving unit 602C of the remote PE device in fig. 6C. For brevity, no further description is provided herein.

It should be noted that, in this embodiment, a virtual far-end PE device implemented by combining a general physical server with a network function virtualization NFV technology may also be based on, where the virtual far-end PE device is a virtual router, and a second, a third, and N (according to actual needs) far-end PE devices may be virtualized. The Virtual remote PE device may be a Virtual Machine (VM) running a program for providing a function of receiving messages, and the VM is deployed on a hardware device (e.g., a physical server). A virtual machine refers to a complete computer system with complete hardware system functionality, which is emulated by software, running in a completely isolated environment. Through reading the application, a person skilled in the art can combine the NFV technology to virtually generate a plurality of PE devices with the above functions on a common physical server. And will not be described in detail herein.

Fig. 7 is a schematic diagram of a system for sending and receiving messages according to an embodiment of the present invention. As shown in fig. 7, the system 700 includes a first PE device 710 and a remote PE device 720. The first PE device 710 is any one of the first PE devices or the virtual first PE device described in fig. 5a, 5b, 5c, and 5d, and the remote PE device 720 is any one of the remote PE devices or the virtual remote PE device described in fig. 6a, 6b, 6c, and 6 d. For a detailed description of each device in the system, please refer to the relevant sections of fig. 5a to 5d and fig. 6a to 6d, which are not described herein again.

It should be understood that fig. 6a, 6b, 6c and 6d only show a simplified design of the first PE device and the controller. Fig. 6a, 6b, 6c and 6d only show a simplified design of the remote PE device and controller. In practical applications, the first PE device and the remote PE device may respectively include any number of receivers, transmitters, processors, memories, main control boards, interface boards, switch boards, physical interface cards, etc., and all the first PE devices and the remote PE devices that can implement the present invention are within the protection scope of the present invention.

It should be understood that, on reading the present application, those skilled in the art may make various combinations of optional features, steps or methods described in the embodiments of the present application without making any special details, all of which belong to the embodiments disclosed in the present application, and only because the descriptions or the texts are simple, the descriptions of the various combinations are not repeated.

It should be understood that the term "and/or" herein is merely one type of association relationship that describes an associated object, meaning that three relationships may exist, e.g., a and/or B may mean: a exists alone, A and B exist simultaneously, and B exists alone. In addition, the character "/" herein generally indicates that the former and latter related objects are in an "or" relationship.

It should be understood that, in various embodiments of the present invention, the sequence numbers of the above-mentioned processes do not mean the execution sequence, and the execution sequence of each process should be determined by its function and inherent logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

Those of ordinary skill in the art will appreciate that the various illustrative elements and algorithm steps described in connection with the embodiments disclosed herein may be implemented as electronic hardware or combinations of computer software and electronic hardware. Whether such functionality is implemented as hardware or software depends upon the particular application and design constraints imposed on the implementation. Skilled artisans may implement the described functionality in varying ways for each particular application, but such implementation decisions should not be interpreted as causing a departure from the scope of the present invention.

It is clear to those skilled in the art that, for convenience and brevity of description, the specific working processes of the above-described systems, apparatuses and units may refer to the corresponding processes in the foregoing method embodiments, and are not described herein again.

In the several embodiments provided in the present application, it should be understood that the disclosed system, apparatus and method may be implemented in other ways. For example, the above-described apparatus embodiments are merely illustrative, and for example, the division of the units is only one logical division, and other divisions may be realized in practice, for example, a plurality of units or components may be combined or integrated into another system, or some features may be omitted, or not executed. In addition, the shown or discussed mutual coupling or direct coupling or communication connection may be an indirect coupling or communication connection through some interfaces, devices or units, and may be in an electrical, mechanical or other form.

The units described as separate parts may or may not be physically separate, and parts displayed as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of the embodiment.

In addition, functional units in the embodiments of the present invention may be integrated into one processing unit, or each unit may exist alone physically, or two or more units are integrated into one unit.

The functions, if implemented in the form of software functional units and sold or used as a stand-alone product, may be stored in a computer readable storage medium. Based on such understanding, the technical solution of the present invention may be embodied in the form of a software product, which is stored in a storage medium and includes instructions for causing a computer device (which may be a personal computer, a server, or a network device) to execute all or part of the steps of the method according to the embodiments of the present invention. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk or an optical disk, and other various media capable of storing program codes.

The above description is only for the specific embodiments of the present invention, but the scope of the present invention is not limited thereto, and any person skilled in the art can easily conceive of the changes or substitutions within the technical scope of the present invention, and all the changes or substitutions should be covered within the scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the protection scope of the appended claims.

Claims

1. A method for sending a message is applied to an Ethernet Virtual Private Network (EVPN), a user side device is respectively accessed to a scene of at least three Provider Edge (PE) devices through at least three links, the at least three links form a group of links, the group of links are link sections, the at least three PE devices include a first PE device, and the method comprises the following steps:

the first PE equipment acquires a redundancy mode of the link section;

when the redundancy mode of the link segment is a multi-active mode, the first PE device sends a mode advertisement message to a remote PE device, where the mode advertisement message includes the multi-active mode and an identifier of the link segment, where the identifier of the link segment is used to uniquely identify the link segment, and the multi-active mode indicates that a part of links in the link segment may be used to forward a data stream, and the number of the part of links is greater than 1 and less than the maximum number of links in the link segment.

2. The method of claim 1, further comprising:

the first PE device sends an MAC routing message to the far-end PE device, wherein the MAC routing message comprises a destination MAC address and a next hop network address, the destination MAC address is an MAC address of a terminal device accessed to the user side device, and the next hop network address is a network address of the first PE device.

3. The method of claim 1, further comprising:

when the redundancy mode of the link section is a multi-active mode, the first PE device obtains a state of a first link between the user side device and the first PE device, where the state of the first link is active or inactive;

correspondingly, the mode advertisement message further includes a state of the first link and a next hop network address, where the next hop network address is a network address of the first PE device;

the first PE device further sends an MAC routing message to the far-end PE device, wherein the MAC routing message comprises a destination MAC address, and the destination MAC address is an MAC address of a terminal device accessed to the user-side device.

4. The method of claim 3, wherein the link is an Ethernet link, the link segment is an Ethernet segment ES, the redundancy mode of the link segment is the redundancy mode of the ES, and the first link is a first Ethernet link; accordingly, the number of the first and second electrodes,

when the state of the first ethernet link is active, the mode advertisement message is used to advertise the far-end PE device, and a part of ethernet links in the ES may be used to forward data streams, where the first ethernet link may be used to forward data streams; or

When the state of the first ethernet link is inactive, the mode advertisement message is used to advertise the remote PE device, and a part of ethernet links in the ES may be used to forward data streams, where the first ethernet link may not be used to forward data streams.

5. The method of claim 3, wherein the link is a pseudowire PW, the link segment is a pseudowire segment PWS, the redundancy mode of the link segment is a redundancy mode of the PWS, and the first link is a first PW; accordingly, the number of the first and second electrodes,

when the state of the first PW is active, the mode advertisement message is used to advertise the far-end PE device, and a part of PWs in the PWs may be used to forward a data flow, where the first PW may be used to forward a data flow; or

When the state of the first PW is inactive, the mode advertisement message is used to advertise the far-end PE device, and a part of PWs in the PWs may be used to forward a data flow, where the first PW may not be used to forward a data flow.

6. The method of claim 1, wherein the link is a pseudowire PW, wherein the link segment is a pseudowire segment PWs, and wherein the redundancy mode of the link segment is a redundancy mode of the PWs;

when the redundancy mode of the PWS is a full-active mode, the first PE device sends a mode notification message to a remote PE device, wherein the mode notification message comprises the full-active mode and an identifier of the link segment, the identifier of the link segment is used for uniquely identifying the link segment, and the full-active mode indicates that all PWs in the PWS can be used for forwarding data streams.

7. The method of claim 1, wherein the link is a pseudowire PW, wherein the link segment is a pseudowire segment PWs, and wherein the redundancy mode of the link segment is a redundancy mode of the PWs;

when the redundancy mode of the PWS is a single active mode, the first PE device sends a mode notification message to a remote PE device, wherein the mode notification message comprises the single active mode and an identifier of the link segment, the identifier of the link segment is used for uniquely identifying the link segment, and the single active mode indicates that only one PW in the PWS can be used for forwarding data streams.

8. A method for receiving a message is applied to an Ethernet Virtual Private Network (EVPN), a user side device is respectively accessed to a scene of at least three Provider Edge (PE) devices through at least three links, the at least three links form a group of links, the group of links are link sections, the at least three PE devices include a first PE device, and the method comprises the following steps:

a far-end PE device receives a first mode notification message sent by the first PE device, wherein the first mode notification message comprises a redundant mode of the link segment and an identifier of the link segment, and the identifier of the link segment is used for uniquely identifying the link segment;

and the far-end PE equipment acquires the identifier of the link section, and when the redundancy mode of the link section is a multi-active mode, the far-end PE equipment confirms that part of links in the link section identified by the identifier of the link section can be used for forwarding data streams according to the multi-active mode, and the number of the part of links is more than 1 and less than the maximum number of links in the link section.

9. The method of claim 8, wherein the at least three PE devices further comprise a second PE device, the method further comprising:

when the state of a first link between the first PE device and the user side device and the state of a second link between the second PE device and the user side device are both active, the remote PE device receives a first MAC routing message and a second MAC routing message sent from the first PE device and the second PE device, respectively, where the first MAC routing message includes a destination MAC address and a first next hop network address, and the second MAC routing message includes the destination MAC address and a second next hop network address; the destination MAC address is an MAC address of a terminal device accessing the user side device, the first next hop network address is a network address of the first PE device, and the second next hop network address is a network address of the second PE device;

the far-end PE device determines, according to the multi-active mode in the first mode advertisement message, the identifier of the link segment, the first MAC routing message, and the second MAC routing message, that the first link and the second link can forward the data stream destined for the destination MAC in a load-sharing manner in the link segment.

10. The method of claim 8, further comprising:

the first mode advertisement message further comprises a state of a first link, the state of the first link being active or inactive;

the far-end PE device determines whether the first link between the first PE device and the user side device can be used for forwarding data flow according to the multi-active mode, the identification of the link section and the state of the first link.

11. The method of claim 10, wherein the at least three PE devices further comprise a second PE device, the method further comprising:

the far-end PE equipment receives a second mode notification message sent by the second PE equipment, wherein the second mode notification message comprises that the redundant mode of the link section is a multi-active mode, the state of a second link is active and the link section identifier;

the far-end PE device determines that the second link between the second PE device and the user-side device can be used for forwarding data flow according to the multi-active mode, the identification of the link segment and the state of the second link in the second mode notification message;

when the state of the first link is active, the far-end PE device determines that the first link and the second link can carry out load sharing forwarding of data flow in the link section.

12. The method of claim 11, wherein when the state of the first link is active, the determining, by the remote PE device, that the first link and the second link can forward data streams in load sharing in the link segment comprises:

the first mode advertisement message further includes a first next hop network address, where the first next hop network address is a network address of the first PE device; the second mode advertisement message further includes a second next hop network address, where the second next hop network address is a network address of the second PE device;

the far-end PE device generates a mapping relationship between link information and next hop network address information according to the multi-active mode, the link segment identifier, the state of the first link being active, the state of the second link being active, the first next hop network address and the second next hop network address, where the mapping relationship indicates that the first link and the second link in the link segment identified by the link segment identifier can share load and forward a data stream, where the data stream reaches the first link through the first PE device identified by the first next hop network address, and the data stream reaches the second link through the second PE device identified by the second next hop network address;

correspondingly, the remote PE device further receives an MAC routing message from the first PE device, where the MAC routing message includes a destination MAC address, and the destination MAC address is an MAC address of a terminal device accessing the user side device;

the far-end PE device further generates an MAC forwarding table item according to the destination MAC address and the mapping relation, wherein the MAC forwarding table item comprises the destination MAC address and an outgoing interface list, the outgoing interface list comprises a first outgoing interface of the far-end PE device and a second outgoing interface of the far-end PE device, and the far-end PE device forwards a data stream to the destination MAC in a load sharing manner from the first outgoing interface and the second outgoing interface; the first output interface is obtained according to the first next hop network address, and the second output interface is obtained according to the second next hop network address.

13. The method according to any one of claims 10 to 12, wherein the link is an ethernet link, the link segment is an ethernet segment ES, the redundancy mode of the link segment is a redundancy mode of the ES, and the first link is a first ethernet link; accordingly, the determining, by the far-end PE device, whether the first link between the first PE device and the user-side device can be used for forwarding a data flow according to the multi-active mode, the identifier of the link segment, and the state of the first link includes:

when the state of the first ethernet link is active, the far-end PE device determines that the first ethernet link is an active link in the ES, and may be used to forward a data stream; or

When the state of the first ethernet link is inactive, the far-end PE device determines that the first ethernet link is not available for forwarding a data stream as an inactive link in the ES.

14. The method according to any one of claims 10 to 12, wherein the link is a pseudowire PW, the link segment is a pseudowire segment PWs, the redundancy mode of the link segment is a redundancy mode of the PWs, and the first link is a first PW; accordingly, the determining, by the far-end PE device, whether the first link between the first PE device and the user-side device can be used for forwarding a data flow according to the multi-active mode, the identifier of the link segment, and the state of the first link includes:

when the state of the first PW is active, the far-end PE device determines that the first PW is an active link in the PWs and can be used to forward a data stream; or

When the state of the first PW is inactive, the far-end PE device determines that the first PW cannot be used to forward a data stream as an inactive link in the PWs.

15. The method of claim 8, wherein the link is a pseudowire PW, wherein the link segment is a pseudowire segment PWs, wherein the redundancy mode of the link segment is a redundancy mode of the PWs, and wherein the identity of the link is an identity of the PWs;

and when the redundancy mode of the link section is a full-active mode, the far-end PE confirms that all PWs in the PWS identified by the PWS identification can be used for forwarding data streams according to the full-active mode and the PWS identification.

16. The method of claim 8, wherein the link is a pseudowire PW, wherein the link segment is a pseudowire segment PWs, wherein the redundancy mode of the link segment is a redundancy mode of the PWs, and wherein the identity of the link is an identity of the PWs;

and when the redundancy mode of the link section is a single-active mode, the far-end PE confirms that only one PW in the PWS identified by the PWS identification can be used for forwarding data streams according to the single-active mode and the PWS identification.

17. The first provider edge PE device is applied to an ethernet virtual private network EVPN, and a user side device is respectively accessed to a scene of at least three provider edge PE devices through at least three links, where the at least three links form a group of links, the group of links is a link segment, the at least three PE devices include the first PE device, and the first PE device includes:

the processing unit is used for acquiring the redundant mode of the link section and generating a mode notification message; when the redundancy mode of the link segment is a multi-active mode, the mode advertisement message includes information indicating that the redundancy mode of the link segment is the multi-active mode and an identifier of the link segment, the identifier of the link segment is used for uniquely identifying the link segment, the multi-active mode indicates that a part of links in the link segment can be used for forwarding data streams, and the number of the part of links is greater than 1 and less than the maximum number of links in the link segment;

and the sending unit is used for sending the mode notification message to the far-end PE equipment.

18. The first PE device of claim 17, wherein the sending unit is further configured to:

and sending an MAC routing message to the far-end PE device, wherein the MAC routing message comprises a destination MAC address and a next hop network address, the destination MAC address is an MAC address of a terminal device accessed to the user side device, and the next hop network address is a network address of the first PE device.

19. The first PE device of claim 17, wherein the processing unit is further configured to:

when the redundancy mode of the link section is a multi-active mode, acquiring the state of a first link between the user side equipment and the first PE equipment, wherein the state of the first link is active or inactive;

the sending unit is further configured to send an MAC routing message to the remote PE device, where the MAC routing message includes a destination MAC address, and the destination MAC address is an MAC address of a terminal device accessing the user side device.

20. The first PE device of claim 19, wherein the link is an ethernet link, the link segment is an ethernet segment ES, the redundancy mode of the link segment is a redundancy mode of the ES, and the first link is a first ethernet link; accordingly, the number of the first and second electrodes,

21. The first PE device of claim 19, wherein the link is a pseudowire PW, the link segment is a pseudowire segment PWs, the redundancy mode of the link segment is a redundancy mode of the PWs, and the first link is a first PW; accordingly, the number of the first and second electrodes,

22. The first PE device of claim 17, wherein the link is a pseudowire PW, the link segment is a pseudowire segment PWs, and the redundancy mode of the link segment is a redundancy mode of the PWs;

when the redundancy mode of the PWS is a full active mode, the mode advertisement message includes information indicating that the redundancy mode of the PWS is the full active mode and an identification of the PWS, the identification of the PWS being used to uniquely identify the PWS, the full active mode indicating that all PWs in the PWS can be used to forward data streams.

23. The first PE device of claim 17, wherein the link is a pseudowire PW, the link segment is a pseudowire segment PWs, and the redundancy mode of the link segment is a redundancy mode of the PWs;

when the redundancy mode of the PWS is a single active mode, the mode advertisement message comprises information indicating that the redundancy mode of the PWS is the single active mode and an identification of the PWS, wherein the identification of the PWS is used for uniquely identifying the PWS, and the single active mode represents that only one PW in the PWS can be used for forwarding data streams.

24. The utility model provides a far-end PE equipment, is characterized in that, is applied to ethernet virtual private network EVPN, and user side's equipment inserts respectively in the scene of at least three operator edge PE equipment through at least three links, at least three links form a set of link, a set of link is the link section, at least three PE equipment includes first PE equipment, far-end PE equipment includes:

a receiving unit, configured to receive a first mode notification message sent by the first PE device, where the first mode notification message includes a redundancy mode of the link segment and an identifier of the link segment, and the identifier of the link segment is used to uniquely identify the link segment;

and the processing unit is used for acquiring the identifier of the link section, and when the redundancy mode of the link section is a multi-active mode, determining that part of links in the link section identified by the identifier of the link section can be used for forwarding data streams according to the multi-active mode, wherein the number of the part of links is greater than 1 and less than the maximum number of links in the link section.

25. The remote PE device of claim 24, wherein the at least three PE devices further include a second PE device, and wherein the receiving unit is further configured to:

when the state of a first link between the first PE device and the user side device and the state of a second link between the second PE device and the user side device are both active, receiving a first MAC routing message and a second MAC routing message sent by the first PE device and the second PE device respectively, wherein the first MAC routing message comprises a destination MAC address and a first next hop network address, and the second MAC routing message comprises the destination MAC address and a second next hop network address; the destination MAC address is an MAC address of a terminal device accessing the user side device, the first next hop network address is a network address of the first PE device, and the second next hop network address is a network address of the second PE device;

the processing unit is further configured to determine, according to the multi-active mode in the first mode advertisement message, the identifier of the link segment, the first MAC routing message, and the second MAC routing message, that the first link and the second link can forward the data flow destined for the destination MAC in a load sharing manner in the link segment.

26. The remote PE device of claim 24, wherein the first mode advertisement message further includes a state of the first link, the state of the first link being active or inactive;

the processing unit is further configured to determine whether the first link between the first PE device and the user-side device is available for forwarding a data flow according to the multi-active mode, the identification of the link segment, and the state of the first link.

27. The remote PE device of claim 26, wherein the at least three PE devices further include a second PE device, and wherein the receiving unit is further configured to:

receiving a second mode notification message sent by the second PE device, wherein the second mode notification message comprises that the redundancy mode of the link section is a multi-active mode, the state of a second link is active, and the link section identifier;

the processing unit is further configured to determine, according to the multi-active mode, the identifier of the link segment, and the state of the second link in the second mode advertisement message, that the second link between the second PE device and the user-side device is available for forwarding a data flow;

when the state of the first link is active, the processing unit is further configured to determine that the first link and the second link can forward the data stream in load sharing in the link segment.

28. The remote PE device of claim 27, wherein the processing unit, when the state of the first link is active, further configured to determine that the first link and the second link can forward data flows in load sharing over the link segment comprises:

the processing unit generates a mapping relationship between link information and next hop network address information according to the multi-active mode, the link section identifier, the state of the first link being active, the state of the second link being active, the first next hop network address and the second next hop network address, where the mapping relationship indicates a forwarding data stream that can be load-shared by the first link and the second link in the link section identified by the link section identifier, the data stream reaches the first link through the first PE device identified by the first next hop network address, and the data stream reaches the second link through the second PE device identified by the second next hop network address;

correspondingly, the receiving unit is further configured to receive a MAC routing message from the first PE device, where the MAC routing message includes a destination MAC address, and the destination MAC address is a MAC address of a terminal device accessing the user side device;

the processing unit further generates an MAC forwarding table according to the destination MAC address and the mapping relationship, where the MAC forwarding table includes the destination MAC address and an outgoing interface list, the outgoing interface list includes a first outgoing interface of the remote PE device and a second outgoing interface of the remote PE device, the first outgoing interface is obtained according to the first next hop network address, and the second outgoing interface is obtained according to the second next hop network address;

the far-end PE device further comprises a sending unit, wherein the sending unit is used for forwarding the data stream destined for the destination MAC in a load sharing manner from the first outgoing interface and the second outgoing interface;

the far-end PE device further comprises a storage unit, and the storage unit is used for storing the mapping relation between the link information and the next hop network address information and the MAC forwarding table entry.

29. The remote PE device as recited in any of claims 26 to 28 wherein the link is an ethernet link, the link segment is an ethernet segment ES, the redundancy mode of the link segment is a redundancy mode of the ES, and the first link is a first ethernet link; accordingly, the processing unit is further configured to confirm whether the first link between the first PE device and the user-side device can be used for forwarding a data flow according to the multi-active mode, the identification of the link segment, and the state of the first link, including:

when the state of the first ethernet link is active, the processing unit is further configured to determine that the first ethernet link is an active link in the ES and may be used to forward a data stream; or

When the state of the first ethernet link is inactive, the processing unit is further configured to determine that the first ethernet link is not available for forwarding a data stream as an inactive link in the ES.

30. The remote PE device of any of claims 26 to 28, wherein the link is a pseudowire PW, the link segment is a pseudowire segment PWs, the redundancy mode of the link segment is a redundancy mode of the PWs, and the first link is a first PW; accordingly, the processing unit being further configured to determine whether the first link between the first PE device and the user-side device can be used to forward a data flow according to the multi-active mode, the identification of the link segment, and the state of the first link comprises:

when the state of the first PW is active, the processing unit is further configured to determine that the first PW is available for forwarding a data flow as an active link in the PWS; or

When the state of the first PW is inactive, the processing unit is further configured to determine that the first PW is not available for forwarding a data stream as an inactive link in the PWs.

31. The remote PE device of claim 24, wherein the link is a pseudowire PW, wherein the link segment is a pseudowire segment PWs, wherein the redundancy mode of the link segment is a redundancy mode of the PWs, and wherein the identification of the link is an identification of the PWs;

when the redundant mode of the PWS is a fully active mode, the processing unit is further configured to confirm that all PWs in the PWS identified by the identification of the PWS are available for forwarding a data stream according to the fully active mode and the identification of the PWS.

32. The remote PE device of claim 24, wherein the link is a pseudowire PW, wherein the link segment is a pseudowire segment PWs, wherein the redundancy mode of the link segment is a redundancy mode of the PWs, and wherein the identification of the link is an identification of the PWs;

when the redundant mode of the PWS is a single active mode, the processing unit is further configured to confirm that only one PW in the PWS identified by the identification of the PWS is available for forwarding a data stream according to the single active mode and the identification of the PWS.

33. A system for sending and receiving messages, the system comprising a first PE device according to any of claims 17 to 23 and a remote PE device according to any of claims 24 to 32.