A method, device and system for processing routing
Patent Information
- Application Number
- CN202110182721.9
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2017-03-14
- Publication Date
- 2024-11-15
- Estimated Expiration
- 2037-03-14
Smart Images

Figure CN112929274B_ABST
Abstract
Description
[0001] This application is a divisional application of a Chinese patent application filed with the State Intellectual Property Office on March 14, 2017, with application number CN 201710151473.5 and invention name “A method, device and system for processing routing”, all contents of which are incorporated by reference in this application. Technical Field
[0002] The present application relates to the field of communication technology, and more particularly to a method, device and system for processing routing. More specifically, the present application relates to a virtual extended local area network (VXLAN) technology with an Ethernet virtual private network (EVPN) control plane. Background Art
[0003] VXLAN is a technology that encapsulates Layer 2 messages with Layer 3 protocols. VXLAN technology involves messages in MAC-in-UDP format. Specifically, Ethernet frames based on the Media Access Control (MAC) protocol are encapsulated in User Datagram Protocol (UDP) messages. Furthermore, UDP messages are encapsulated in Internet Protocol (IP) messages, and IP messages can be transmitted in a Layer 3 network. Therefore, Ethernet frames are transmitted in a Layer 3 network. VXLAN technology uses VXLAN Network Identifier (VNI) to identify VXLAN segments. Different VXLAN segments correspond to different VNIs. Different VXLAN segments are isolated from each other. Two virtual machines (VMs) in the same VNI do not need to communicate through a VXLAN Layer 3 gateway (VXLAN L3 Gateway). The communication between two VMs in different VNIs needs to go through the VXLAN Layer 3 gateway. The VNI field contains 24 bits. A management domain can contain up to 2 16VXLAN segments. A virtual extended LAN tunnel endpoint (VTEP) can be integrated into a network virtualization edge (NVE) device as an edge device in a VXLAN. NVE devices transmit VXLAN traffic through VXLAN tunnels. A VXLAN tunnel is a point-to-point logical tunnel between two NVE devices.
[0004] EVPN is a Layer 2 virtual private network (VPN) technology. EVPN connects customer sites in different regions through a cross-IP / Multiprotocol Label Switching (MPLS) bearer network, which is equivalent to these customer sites being located in the same local area network (LAN).
[0005] In the application scenario, VXLAN has no control plane. VXLAN uses multicast flooding to perform VTEP discovery and remote host address learning through the data plane, which results in a lot of flooding traffic in the data center network. EVPN based on Border Gateway Protocol (BGP) and MPLS can be used to implement the control plane of VXLAN. In this article, this VXLAN scenario with EVPN control plane is named EVPN VXLAN.
[0006] In the EVPN VXLAN application scenario, two NVE devices are directly connected through a physical link. For example, two NVE devices are connected through a PeerLink link. However, the direct connection between two NVE devices through a physical link limits the application scenarios of EVPN and has problems such as low reliability and complex deployment. Summary of the invention
[0007] In view of this, the embodiments of the present application provide a method, device and system for processing routing, which are applied in EVPN VXLAN application scenarios. The application scenarios of EVPN do not limit the constraints that the links between NVE devices must use physical direct links, which helps to expand the application scenarios of EVPN. In addition, it helps to improve reliability and reduce deployment complexity.
[0008] The technical solutions provided in the embodiments of the present application are as follows.
[0009] In a first aspect, a method for processing routing is provided. The method is applied in a VXLAN, and the VXLAN includes a first NVE device and a second NVE device. The first NVE device receives a second integrated multicast Ethernet tag (IMET) route from the second NVE device. The Originating Router's IP Address domain in the second IMET route includes a public VTEP address, and the second IMET route also includes a second VTEP address. The public VTEP address included in the Originating Router's IPAddress domain in the second IMET route is different from the second VTEP address. Then, the first NVE device determines whether the public VTEP address included in the Originating Router's IP Address domain in the second IMET route is the same as the public VTEP address stored by the first NVE device. When the first NVE device determines that the public VTEP address included in the Originating Router's IP Address field in the second IMET route is the same as the public VTEP address stored by the first NVE device, the first NVE device establishes a VXLAN tunnel from the first NVE device to the second NVE device according to the second VTEP address in the second IMET route.
[0010] Based on the solution provided in the embodiment, the IMET route is extended to carry a VTEP address different from the public VTEP address, so that after the NVE device receives the extended IMET route, a VXLAN tunnel from the NVE device to the BGPpeer of the NVE device is established according to the VTEP address. The application scenario of EVPN is not limited to the constraint that the link between NVE devices must adopt a physical direct link, which helps to expand the application scenario of EVPN. In addition, it helps to improve reliability and reduce deployment complexity.
[0011] In a possible implementation of the first aspect, the method further includes the first NVE device sending a first IMET route to the second NVE device. The Originating Router's IPAddress field of the first IMET route carries a public VTEP address stored by the first NVE device, and the first IMET route also includes a first VTEP address. The first VTEP address is used by the second NVE device to establish a VXLAN tunnel from the second NVE device to the first NVE device. The public VTEP address stored by the first NVE device is different from the first VTEP address.
[0012] In another possible implementation of the first aspect, the VXLAN also includes a provider edge (English: Provider Edge, abbreviated: PE) device. The method also includes, the first NVE device sending the first IMET route to the PE device. The first NVE device receives the IMET route from the PE device, and the Originating Router's IP Address field in the IMET route from the PE device carries the VTEP address of the PE device. The first NVE device establishes a VXLAN tunnel from the first NVE device to the PE device according to the VTEP address of the PE device in the IMET route from the PE device, and the VTEP address of the PE device is not equal to the public VTEP address stored by the first NVE device.
[0013] In another possible implementation of the first aspect, the VXLAN further includes a third NVE device, and the method further includes, the first NVE device receiving a third IMET route from the third NVE device. The Originating Router's IP Address field in the third IMET route includes a public VTEP address, and the third IMET route also includes a third VTEP address. The public VTEP address included in the Originating Router's IPAddress field in the third IMET route is different from the third VTEP address. The first NVE device determines whether the public VTEP address included in the Originating Router's IP Address field in the third IMET route is the same as the public VTEP address stored by the first NVE device. When the first NVE device determines that the public VTEP address included in the Originating Router's IP Address field in the third IMET route is the same as the public VTEP address stored by the first NVE device, the first NVE device establishes a VXLAN tunnel from the first NVE device to the third NVE device according to the third VTEP address in the third IMET route.
[0014] Optionally, a customer edge (CE) device is multi-homed to the first NVE device, the second NVE device, and the third NVE device via multiple Ethernet links, and the multiple Ethernet links constitute an Ethernet segment (ES). The method also includes the first NVE device determining the first NVE device as a designated forwarder (DF) in the ES according to ES routes from the second NVE device and the third NVE device.
[0015] Based on the above implementation, the scenario where CE devices are multi-homed to more than two NVE devices can be realized, and the multi-active redundancy mode of more NVE devices can be supported, which helps to increase the reliability of EVPN VXLAN and increase the flexibility of network deployment.
[0016] Optionally, the method further includes, the first NVE device receiving broadcast, unknown unicast and multicast BUM traffic sent by the second NVE device via a VXLAN tunnel from the second NVE device to the first NVE device. The first NVE device avoids forwarding the BUM traffic to the third NVE device via a VXLAN tunnel from the first NVE device to the third NVE device based on the fact that the BUM traffic is received by the first NVE device via the VXLAN tunnel from the second NVE device to the first NVE device.
[0017] Based on the above implementation method, using the VXLAN tunnel between NVE devices can not only realize the normal forwarding of BUM traffic, but also avoid forwarding of BUM traffic to the VXLAN tunnels between other NVE devices, thereby avoiding double sending of BUM traffic.
[0018] In another possible implementation of the first aspect, the CE device is dual-homed to the first NVE device and the second NVE device via multiple Ethernet links, and the multiple Ethernet links constitute an ES. The method also includes, the first NVE device determining the first NVE device as the DF in the ES according to the ES route from the second NVE device. The first NVE device sends an Ethernet AD per ES route to the second NVE device, and the Ethernet A-D per ES route carries an Ethernet segment identifier (English: Ethernet Segment Identifier, abbreviated: ESI) label allocated by the first NVE device to the Ethernet link between the CE device and the first NVE device. The first NVE device receives a VXLAN message sent by the second NVE device via a VXLAN tunnel from the second NVE device to the first NVE device, the original Ethernet payload in the VXLAN message includes BUM traffic, and the VXLAN message carries the ESI label, wherein the BUM traffic comes from the CE device. The first NVE device carries the ESI label based on the VXLAN message to avoid forwarding the BUM traffic to the CE device via the ES between the CE device and the first NVE device.
[0019] Optionally, the ESI label is encapsulated between the VXLAN header and the original Ethernet payload in the VXLAN message, and the VXLAN header carries indication information, where the indication information is used to indicate that the VXLAN message carries the ESI label.
[0020] Based on the above implementation, the format of the VXLAN message is extended so that the VXLAN message can carry the ESI label, thereby solving the loop problem generated in the process of forwarding BUM traffic.
[0021] In another possible implementation of the first aspect, the CE device is dual-homed to the first NVE device and the second NVE device via multiple Ethernet links, and the multiple Ethernet links constitute an ES. The method also includes, the first NVE device receiving a MAC / IP Advertisement route from the second NVE device, the MAC / IP Advertisement route including a MAC-VLAN ID. The MAC-VLAN ID is used to indicate the VLAN to which the MAC address carried by the MAC / IP Advertisement route belongs. The MAC address carried by the MAC / IP Advertisement route is the MAC address of the CE device, or the MAC address carried by the MAC / IP Advertisement route is the MAC address of a host in an Ethernet virtual private network site EVPN site governed by the CE device. The first NVE device determines the local interface of the first NVE that connects the Ethernet link between the CE device and the first NVE device according to the ESI carried by the MAC / IP Advertisement route and the MAC-VLAN ID. The ESI carried by the MAC / IP Advertisement route is used to indicate the ES between the CE device and the second NVE device. The control plane of the first NVE device sends a first MAC table entry to the forwarding plane of the first NVE device, wherein the outgoing interface in the first MAC table entry is the local interface of the first NVE device, wherein the first MAC table entry includes the MAC address carried by the MAC / IP Advertisement route and the MAC-VLAN ID.
[0022] Optionally, the first NVE device receives unicast traffic from the second NVE device via a VXLAN tunnel from the second NVE device to the first NVE device, where the unicast traffic is sent by the second NVE device based on a failure of an Ethernet link between the second NVE device and the CE device. The first NVE device sends the unicast traffic to the CE device via the Ethernet link between the first NVE device and the CE device according to the first MAC table entry.
[0023] Based on the above implementation, the forwarding path can be redirected through the VXLAN tunnel between the MAC / IP Advertisement routing and the NVE device. When a link failure occurs on an Ethernet link, the normal forwarding of unicast traffic can still be ensured.
[0024] Optionally, the MAC / IP Advertisement route is carried in the MP_REACH_NLRI, and the next hop field in the MP_REACH_NLRI includes a public VTEP address. The MAC / IP Advertisement route also includes the second VTEP address. The public VTEP address included in the next hop field in the MP_REACH_NLRI is the same as the public VTEP address included in the Originating Router's IP Address field in the second IMET route. The method also includes the control plane of the first NVE device sending a second MAC table entry to the forwarding plane of the first NVE device based on the public VTEP address included in the next hop field in the MP_REACH_NLRI and the second VTEP address included in the MAC / IP Advertisement route. The outgoing interface in the second MAC table entry is the local interface of the first NVE device that connects the VXLAN tunnel from the first NVE device to the second NVE device. Among them, the second MAC table entry includes the MAC address carried by the MAC / IP Advertisement route.
[0025] Optionally, the first NVE device receives unicast traffic sent from the PE device to the CE device. The first NVE device sends the unicast traffic via the VXLAN tunnel from the first NVE device to the second NVE device according to the second MAC table entry, so that the second NVE device forwards the unicast traffic to the CE device via the Ethernet link between the second NVE device and the CE device according to the MAC table entry stored in the second NVE device.
[0026] Based on the above implementation, load sharing of unicast traffic can be further achieved, which helps to improve the transmission bandwidth of unicast traffic.
[0027] In another possible implementation of the first aspect, the CE device is connected to the second NVE device. The method also includes, the first NVE device receiving a MAC / IP Advertisement route from the second NVE device, the MAC / IP Advertisement route being carried in the MP_REACH_NLRI, the next hop domain in the MP_REACH_NLRI including a public VTEP address, the MAC / IP Advertisement route including a MAC-VLAN ID and the second VTEP address, wherein the MAC-VLAN ID is used to indicate the VLAN to which the MAC address carried by the MAC / IP Advertisement route belongs, the MAC address carried by the MAC / IP Advertisement route is the MAC address of the CE device, or the MAC address carried by the MAC / IP Advertisement route is the MAC address of a host in an EVPN site governed by the CE device. The first NVE device determines that there is no local interface in the first NVE device that connects the Ethernet link between the CE device and the first NVE device based on the ESI and the MAC-VLAN ID carried by the MAC / IP Advertisement route, wherein the ESI carried by the MAC / IP Advertisement route is used to indicate the ES between the CE device and the second NVE device. The control plane of the first NVE device sends a MAC table entry to the forwarding plane of the first NVE device based on the public VTEP address included in the next hop domain in the MP_REACH_NLRI and the second VTEP address included in the MAC / IP Advertisement route, wherein the outbound interface of the MAC table entry is the local interface of the first NVE device that connects the VXLAN tunnel from the first NVE device to the second NVE device, wherein the MAC table entry includes the MAC address carried by the MAC / IP Advertisement route.
[0028] Optionally, the first NVE device receives unicast traffic sent from the PE device to the CE device. The first NVE device sends the unicast traffic via a VXLAN tunnel from the first NVE device to the second NVE device according to the MAC table entry, so that the second NVE device forwards the unicast traffic to the CE device via an Ethernet link between the second NVE device and the CE device according to the MAC table entry stored in the second NVE device.
[0029] Based on the above implementation manner, it can be ensured that the CE device that is singly-homed and connected to the NVE device can normally receive unicast traffic.
[0030] In the first aspect, optionally, the VXLAN attribute in the first IMET route carries the first VTEP address.
[0031] In the first aspect, optionally, the VXLAN attribute in the second IMET route carries the second VTEP address.
[0032] In the first aspect, optionally, the first IMET route includes a VTEP VNI attribute, the VTEP VNI attribute carries a VTEP VNI, the VTEP VNI is bound to the first VTEP address, and the VTEP VNI is used to indicate the BD to which the first NVE device belongs.
[0033] In the first aspect, optionally, the second IMET route includes a VTEP VNI attribute, the VTEP VNI attribute carries a VTEP VNI, the VTEP VNI is bound to the second VTEP address, and the VTEP VNI is used to indicate the BD to which the second NVE device belongs.
[0034] In a second aspect, a first NVE device is provided, wherein the first NVE device has a function of implementing the behavior of the first NVE device in the above method. The function can be implemented based on hardware, or can be implemented based on hardware executing corresponding software. The hardware or software includes one or more modules corresponding to the above functions.
[0035] In one possible design, the structure of the first NVE device includes a processor and an interface, and the processor is configured to support the first NVE device to perform corresponding functions in the above method. The interface is used to support communication between the first NVE device and the BGP peer of the first NVE device, send information or instructions involved in the above method to the BGP peer of the first NVE device, or receive information or instructions involved in the above method from the BGP peer of the first NVE device. The first NVE device may also include a memory, which is coupled to the processor and stores necessary program instructions and data for the first NVE device.
[0036] In another possible design, the first NVE device includes: a processor, a transmitter, a random access memory, a read-only memory, and a bus. The processor is respectively coupled to the transmitter, the random access memory, and the read-only memory through the bus. When the first NVE device needs to be run, it is started by the basic input / output system solidified in the read-only memory or the bootloader boot system in the embedded system to guide the first NVE device into a normal operating state. After the first NVE device enters the normal operating state, the application program and the operating system are run in the random access memory, so that the processor executes the method in the first aspect or any possible implementation of the first aspect.
[0037] In a third aspect, a first NVE device is provided, wherein the first NVE device includes: a main control board and an interface board, and further, may also include a switching network board. The first NVE device is used to execute the method in the first aspect or any possible implementation of the first aspect. Specifically, the first NVE device includes a module for executing the method in the first aspect or any possible implementation of the first aspect.
[0038] In a fourth aspect, a first NVE device is provided, wherein the first NVE device includes a controller and a first NVE forwarding device. The first NVE forwarding device includes: an interface board, and further, may also include a switching network board. The first NVE device is used to perform the function of the interface board in the third aspect, and further, may also perform the function of the switching network board in the third aspect. The controller includes a receiver, a processor, a transmitter, a random access memory, a read-only memory, and a bus. The processor is respectively coupled to the receiver, the transmitter, the random access memory, and the read-only memory through the bus. When the controller needs to be run, it is started by the basic input / output system solidified in the read-only memory or the bootloader boot system in the embedded system, and the controller is guided to enter a normal operating state. After the controller enters the normal operating state, the application program and the operating system are run in the random access memory, so that the processor performs the function of the main control board in the third aspect.
[0039] In a fifth aspect, a network system for processing routing is provided, the network system comprising at least two NVE devices, each of the at least two NVE devices being the first NVE device in the aforementioned second aspect, third aspect or fourth aspect.
[0040] In a sixth aspect, a computer storage medium is provided for storing programs, codes or instructions used for the above-mentioned first NVE device. When a processor or hardware device executes these programs, codes or instructions, the functions or steps of the first NVE device in the above aspect can be completed.
[0041] Through the above scheme, the method, device and system for processing routing provided by the embodiment of the present application, in the EVPNVXLAN application scenario, the NVE device set consisting of at least two NVE devices includes a public VTEP address, the public VTEP address is used to identify the public VTEP, and the public VTEP is deployed on each NVE device in the NVE device set. Moreover, each NVE device in the NVE device set also includes a VTEP address, and the VTEP address is used to identify the VTEP included in the corresponding NVE device. The NVE devices in the NVE device set establish a VXLAN tunnel from the NVE device to the BGP peer of the NVE device through IMET routing based on the VTEP address and the public VTEP address. Therefore, in the EVPNVXLAN application scenario, the application scenario of EVPN does not limit the constraint that the link between NVE devices must adopt a physical direct link, which helps to expand the application scenario of EVPN. In addition, it helps to improve reliability and reduce deployment complexity. BRIEF DESCRIPTION OF THE DRAWINGS
[0042] Figure 1 A schematic diagram of an EVPN VXLAN structure according to an embodiment of the present application;
[0043] Figure 2 A flow chart of a method for processing routing according to an embodiment of the present application;
[0044] Figure 3 This is another schematic diagram of the EVPN VXLAN structure of an embodiment of the present application;
[0045] Figure 4 This is another schematic diagram of the EVPN VXLAN structure of an embodiment of the present application;
[0046] Figure 5 Schematic diagram of the format of the VXLAN message of the embodiment of the present application
[0047] Figure 6 A schematic diagram of the format of the VXLAN attributes of an embodiment of the present application;
[0048] Figure 7 This is a schematic diagram of the format of the VTEP VNI attribute of an embodiment of the present application;
[0049] Figure 8 A schematic diagram of the format of the MAC-VLAN attribute of an embodiment of the present application;
[0050] Fig. 9 This is another schematic diagram of the EVPN VXLAN structure of an embodiment of the present application;
[0051] Fig.10This is another schematic diagram of the EVPN VXLAN structure of an embodiment of the present application;
[0052] Fig.11 This is a schematic diagram of the structure of a first NVE device according to an embodiment of the present application;
[0053] Fig.12 This is a schematic diagram of the hardware structure of a first NVE device according to an embodiment of the present application;
[0054] Fig.13 A schematic diagram of the hardware structure of another first NVE device according to an embodiment of the present application;
[0055] Fig.14 Schematic diagram of the hardware structure of yet another first NVE device according to an embodiment of the present application. DETAILED DESCRIPTION
[0056] The embodiments of the present application provide a method, device and system for processing routing, which are applied in EVPN VXLAN application scenarios. The application scenarios of EVPN do not limit the constraints that the links between NVE devices must use physical direct links, which helps to expand the application scenarios of EVPN. In addition, it helps to improve reliability and reduce deployment complexity.
[0057] The following describes the embodiments in detail.
[0058] The EVPN technology involved in this application can refer to the description of Internet Engineering Task Force (English: Internet Engineering Task Force, abbreviated: IETF) Request For Comments (English: Request For Comments, abbreviated: RFC) 7432, and the RFC7432 is incorporated into this application by reference in its entirety.
[0059] In the embodiments of the present application, the expression "the device includes a VTEP address" or the expression "the VTEP address in the device" means that the VTEP address is stored in the device. The expression "the device includes a public VTEP address" or the expression "the public VTEP address in the device" means that the public VTEP address is stored in the device. The device may be a NVE device, a provider edge (English: Provider Edge, abbreviated: PE) device or a VTEP.
[0060] Figure 1 Schematic diagram of an EVPN VXLAN structure of an embodiment of the present application. EVPN VXLAN refers to a VXLAN with an EVPN control plane. The EVPN VXLAN includes a set of NVE devices, and the NVE device set includes at least two NVE devices. Figure 1As shown, the NVE device set includes a first NVE device, a second NVE device, and a third NVE device. It should be understood that in the present application, the number of NVE devices included in the NVE device set is not limited, and can be 2 or 3 or more. Any two NVE devices in the NVE device set are a pair of BGP peers (English: BGP peer). In the EVPN scenario, BGP peer can also be called EVPN peer. Figure 1 For example, the first NVE device and the second NVE device are a pair of BGP peers, the first NVE device and the third NVE device are a pair of BGP peers, and the second NVE device and the third NVE device are a pair of BGP peers. Among them, "a pair of BGP peers" can be understood as: one device is a BGP peer of another device. For example, the first NVE device and the second NVE device are a pair of BGP peers, which can be understood as the first NVE device is the BGP peer of the second NVE device, or the second NVE device is the BGP peer of the first NVE device. The BGP peer can also be called a BGP neighbor; correspondingly, the EVPNpeer can also be called an EVPN neighbor. In this application, for the convenience of explanation, BGPpeer is uniformly used in subsequent embodiments. The BGP peer is established through the OPEN message specified in BGP, and the established BGPpeer is maintained through the KEEPALIVE message. For the implementation of the OPEN message and the KEEPALIVE message, refer to the relevant description of IETF RFC2858 and IETF RFC1771. In addition, route reflectors (English: Route Reflector, abbreviation: RR) can be deployed in the devices at both ends of establishing BGP peer, so as to complete the establishment of BGP peer by using RR. The NVE devices included in the NVE device set can also be called PE devices. In this application, NVE devices are equivalent to PE devices unless otherwise specified. In the description of this application, the devices included in the NVE device set are named NVE devices. In fact, the NVE devices included in the NVE device set can also be called PE devices. Accordingly, the NVE device set can also be called a PE device set.
[0061] The EVPN VXLAN also includes PE devices, such as Figure 1 As shown, the PE device is located outside the NVE device set. The PE device can form a BGP peer with each NVE device in the NVE device set. Figure 1For example, the PE devices are BGP peers of the first NVE device, the second NVE device, and the third NVE device. The PE device can be a physical device or a logical device composed of multiple physical devices. Moreover, the PE device can also be connected to other network devices, for example, the PE device is connected to one or more customer edge (English: Customer Edge, abbreviated: CE) devices.
[0062] The EVPN VXLAN also includes at least one CE device. Figure 1 Three CE devices are shown as an example, namely, a first CE device, a second CE device and a third CE device. Among them, the first CE device is multi-owned to the first NVE device, the second NVE device and the third NVE device, that is, the first CE device communicates with the first NVE device, the second NVE device and the third NVE device respectively. The second CE device is dual-owned to the first NVE device and the second NVE device, that is, the second CE device communicates with the first NVE device and the second NVE device respectively. The third CE device is singly owned to the second NVE device, that is, the third CE device communicates with the second NVE device. In addition, it should be understood that multi-owned can be dual owned. In order to distinguish the number of NVE devices connected to the CE device, the present application uses two expressions to coexist, as shown above.
[0063] In the EVPN VXLAN, the CE device and the NVE device are connected via an Ethernet link. Moreover, all Ethernet links connected to the same CE device constitute an Ethernet segment (English: Ethernet segment, abbreviated: ES). Figure 1 For example, the first CE device is multi-homed to the first NVE device, the second NVE device, and the third NVE device via three Ethernet links, and the three Ethernet links constitute an ES. Figure 1 ES1 is used in the Ethernet Segment Identifier (ESI) to identify the corresponding ES. Figure 1For example, the ESI value of ES1 of the three Ethernet links connecting the first CE device to the first NVE device, the second NVE device and the third NVE device is the same value. The ESI value of ES2 of the two Ethernet links connecting the second CE device to the first NVE device and the second NVE device is the same value. Among them, the ESI value of ES1 is not equal to the ESI value of ES2, and both are non-zero values; the ESI value of ES3 is zero. ESI includes a type (English: Type) domain and an ESI value domain, wherein the Type domain is used to indicate the generation method of ESI. The two commonly used generation methods are Type0 and Type1, wherein Type0 indicates generation through manual configuration, and Type1 indicates generation through the Link Aggregation Control Protocol (English: Link Aggregation Control Protocol, abbreviated: LACP) running between NVE and CE, and the value range of the ESI value domain is 0 to 0xFF, wherein "0x" indicates hexadecimal. For the generation and setting of ES and ESI, please refer to the description in Chapter 5 of RFC7432.
[0064] exist Figure 1 In the EVPN VXLAN shown, the NVE device and the PE device can be a router or a layer 3 switch. The CE device can be a router or a switch or a host. The NVE device, the PE device, and the CE device in the embodiment of the present application are corresponding devices defined in RFC7432. When the CE device is a router or a switch, one or more hosts can be connected. Among them, the host can be a physical device or a VM. Figure 1The EVPN VXLAN shown can be applied to a variety of scenarios. For example, the EVPN VXLAN is applied to a mobile bearer network (English: Mobile Bearer Network), and a typical mobile bearer network is an Internet Protocol Radio Access Network (English: Internet Protocol Radio Access Network, abbreviated: IP RAN). In the mobile bearer network, the first CE device, the second CE device, and the third CE device can be a base station (English: Base Transceiver Station, abbreviated: BTS), the PE device can be connected to a base station controller (English: Base Station Controller, abbreviated: BSC) or a radio network controller (English: Radio Network Controller, abbreviated: RNC), the first NVE device, the second NVE device, and the third NVE device can be a base station side gateway (English: Cell Site Gateway, abbreviated: CSG), and the PE device can be a base station controller side gateway (English: Radio Network Controller Site Gateway, abbreviated: RSG). For another example, the EVPN VXLAN is applied to a fixed network (English: Fixed Network). In a fixed network, the first CE device, the second CE device and the third CE device may be user-side sites, the first NVE device, the second NVE device and the third NVE device may be digital subscriber line access multiplexers (English: Digital Subscriber Line Access Multiplexer, abbreviated: DSLAM), and the PE device may be a broadband access server (English: Broadband Access Server, abbreviated: BAS).
[0065] exist Figure 1 In the EVPN VXLAN shown in FIG. 1 , a VXLAN tunnel is established between the PE device and the NVE device set. For the PE device, the NVE device set is like a single NVE device, that is, all NVE devices in the NVE device set are virtualized into a single NVE device. In a possible implementation scenario, the actual traffic forwarded from the PE device to the first NVE device, the second NVE device, and the third NVE device is carried on multiple physical links from the PE device to the first NVE device, the second NVE device, and the third NVE device (e.g., Figure 1In the figure, the PE device is respectively connected to the first NVE device, the second NVE device and the third NVE device as shown by the dotted lines). In another possible implementation scenario, a router is included between the PE device and the NVE device set. The router performs IP forwarding to realize the function of traffic diversion. The PE device is connected to the router via a physical link. The router connects the first NVE device, the second NVE device and the third NVE device via multiple physical links. In the above two possible implementation scenarios, the physical links from the PE device to the multiple NVE devices in the NVE device set are virtualized into a single VXLAN tunnel. Therefore, when the PE device sends a message to the NVE device set, it regards the NVE device set as a single NVE device. In order to achieve the above purpose, the NVE device set includes a public VTEP. The public VTEP includes a public VTEP address. The public VTEP address is used to identify the public VTEP. The public VTEP address can be an IP address. In addition, the public VTEP is deployed on each NVE device in the NVE device set. In this way, each NVE device in the NVE device set includes the public VTEP. The NVE devices in the NVE device set include the same VTEP, and the NVE devices in the NVE device set include the same VTEP address. Therefore, in the present application, the set of NVE devices including the same VTEP is referred to as an NVE device set. After receiving the integrated multicast Ethernet tag (English: Inclusive Multicast Ethernet Tag, abbreviated: IMET) route sent by each NVE device in the NVE device set, the PE device determines that there is only one VTEP address (public VTEP address) at the sending end of the IMET route. Therefore, for the PE device, the NVE device set is just like one NVE device. The PE device establishes the VXLAN tunnel using the public VTEP as the endpoint of the VXLAN tunnel. Figure 1For example, the value of the public VTEP address of the NVE device set is 9.9.9.9. The public VTEP including the public VTEP address is deployed on the first NVE device, the second NVE device and the third NVE device, respectively. The PE device can establish a VXLAN tunnel from the PE device to the NVE device set based on the IMET route received from the NVE device. For the specific implementation method of establishing the VXLAN tunnel, please refer to the description of the subsequent embodiments of this application. In addition, from the perspective of the PE device, the NVE device set is a virtual NVE device. The use of "set" in the technical term "NVE device set" does not constitute a limitation on the present application. For example, "group" can be used instead of "set" and express the same meaning, that is, "NVE device group".
[0066] In the NVE device set, a VXLAN tunnel is established between NVE devices. That is, a VXLAN tunnel is established between each pair of BGP peers in the NVE device set. Figure 1 For example, a first VXLAN tunnel is established between the first NVE device and the second NVE device. A second VXLAN tunnel is established between the second NVE device and the third NVE device. A third VXLAN tunnel is established between the first NVE device and the third NVE device. In order to achieve the above-mentioned purpose, each NVE device in the NVE device set also includes a VTEP, and the VTEP includes a VTEP address. Moreover, the VTEP address is unique in the EVPN VXLAN. The VTEP address is used to identify the VTEP in the corresponding NVE device. The VTEP address can be an IP address. Furthermore, the public VTEP address of the NVE device set is different from the endpoint (English end point) address of the VXLAN tunnel between any two NVE devices in the NVE device set. Figure 1For example, the first NVE device includes the aforementioned public VTEP, the value of the public VTEP address included in the public VTEP is 9.9.9.9, the first NVE device also includes a first VTEP, and the value of the first VTEP address included in the first VTEP is 1.1.1.1. The second NVE device includes the aforementioned public VTEP, the value of the public VTEP address is 9.9.9.9, the second NVE device also includes a second VTEP, and the value of the second VTEP address included in the second VTEP is 2.2.2.2. The third NVE device includes the aforementioned public VTEP, the value of the public VTEP address is 9.9.9.9, the third NVE device also includes a third VTEP, and the value of the third VTEP address included in the third VTEP is 3.3.3.3. The NVE devices in the NVE device set can establish a VXLAN tunnel from the NVE device to the BGP peer of the NVE device through IMET routing based on the VTEP address and the public VTEP address. For the specific implementation method of establishing a VXLAN tunnel, please refer to the description of the subsequent embodiments of this application.
[0067] Through the above implementation, in the EVPN VXLAN application scenario, the NVE device set consisting of at least two NVE devices includes a public VTEP address, the public VTEP address is used to identify the public VTEP, and the public VTEP is deployed on each NVE device in the NVE device set. Moreover, each NVE device in the NVE device set also includes a VTEP address, and the VTEP address is used to identify the VTEP included in the corresponding NVE device. The NVE devices in the NVE device set establish a VXLAN tunnel from the NVE device to the BGP peer of the NVE device through IMET routing according to the VTEP address and the public VTEP address. Therefore, in the EVPN VXLAN application scenario, the NVE devices in the NVE device set are interconnected through the VXLAN tunnel, so that the NVE devices in the NVE device set are not restricted by the direct physical link. The interconnection of NVE devices is realized by using the VXLAN tunnel, and even if there are switching nodes between the NVE devices in the NVE device set, it will not affect the implementation of EVPN VXLAN. When deploying the NVE devices in the NVE device set, it will not be affected by the deployment region and distance, which improves the flexibility of deployment. Furthermore, the links between NVE devices in the NVE device set are not limited to direct physical links, which helps to improve deployment flexibility.
[0068] Figure 2 A flow chart of a method for processing routing according to an embodiment of the present application. Figure 2 The method shown can be applied to Figure 1 In the EVPN VXLAN shown in Figure 1, Figure 2 The method shown can be applied to Figure 1 In the VXLAN with EVPN control plane shown. The EVPN VXLAN includes a first NVE device and a second NVE device. The first NVE device and the second NVE device are a pair of BGP peers. The implementation method of establishing BGP peers can refer to the description of the aforementioned embodiment, which will not be repeated here. Figure 2 An NVE device set is shown, in which the first NVE device and the second NVE device are included. It should be understood that in actual application scenarios, from the perspective of PE devices, the NVE device set is a single virtual NVE device. The first NVE device and the second NVE device both include a public VTEP. For ease of expression, the expression of the NVE device set is used in the embodiments of the present application. Figure 2 Two NVE devices are shown. It should be understood that in actual application scenarios, the number of NVE devices in the NVE device set may be more than 2. When the number of NVE devices in the NVE device set may be more than 2, any two NVE devices in the NVE device set may implement Figure 2 The method for processing routing is shown. Figure 2 The method shown includes S101 to S104, and may further include S105.
[0069] S101. The second NVE device sends a second IMET route to the first NVE device.
[0070] S102. The first NVE device receives a second IMET route from the second NVE device, wherein the originating router Internet Protocol address (English: Originating Router's IP Address) field in the second IMET route includes a public VTEP address, wherein the second IMET route also includes a second VTEP address, and the public VTEP address included in the Originating Router's IP Address field in the second IMET route is different from the second VTEP address.
[0071] According to the above embodiments, Figure 1 In the description of the scenario shown, the NVE device set includes a public VTEP, the public VTEP includes a public VTEP address, and the VTEP address can be an IP address. The public VTEP is deployed on each NVE device in the NVE device set. Figure 2, the public VTEP is deployed on the first NVE device and the second NVE device. Optionally, the public VTEP address can be deployed in the public VTEP on each NVE device in the NVE device set by static configuration. In this way, based on the public VTEP address, from the perspective of the PE device, the NVE device set can be virtualized as a single NVE device.
[0072] The NVE device in the NVE device set further includes a VTEP, and the VTEP includes a VTEP address, and the VTEP address may be an IP address. Figure 2 , the first NVE device also includes a first VTEP, and the first VTEP includes a first VTEP address; the second NVE device also includes a second VTEP, and the second VTEP includes a second VTEP address. The first VTEP address and the second VTEP address are unique in the EVPN VXLAN, and the value of the first VTEP address and the value of the second VTEP address are not equal to the value of the public VTEP address. Based on the above, the NVE devices in the NVE device set include both public VTEP addresses and their own VTEP addresses. As an example, the public VTEP address and the VTEP address are stored in the memory of the NVE device for access by the processor of the NVE device.
[0073] The second NVE device sends a second IMET route to the first NVE device. The IMET route used in the embodiment of the present application utilizes the IMET route mentioned in RFC7432, and the embodiment of the present application extends the IMET route mentioned in RFC7432. The format and function of the IMET route can be found in the description of RFC7432. Specifically, the second NVE device sending the second IMET route to the first NVE device can be implemented in the following manner. The second NVE device can send an update message (English: UPDATE message) to the first NVE device, and the UPDATE message carries a multi-protocol reachable network layer reachability information (English: Multiprotocol Reachable NLRI, abbreviation: MP_REACH_NLRI) attribute (English: attribute), and the full name of NLRI is Network Layer Reachability Information. The MP_REACH_NLRI attribute includes an EVPN NLRI field and a next hop (English: next hop) field. The EVPN NLRI field is used to carry the second IMET route. For definitions of the UPDATE message and the MP_REACH_NLRI attribute, refer to the description in RFC4760, and for definition of the EVPN NLRI field, refer to the description in Chapter 7 of RFC7432.
[0074] The second IMET route includes an Originating Router's IP Address field. In the embodiment of the present application, the value of the next hop field and the value of the Originating Router's IP Address in the second IMET route are both set to the public VTEP address 9.9.9.9. Figure 1 As shown), after the second NVE device is released, it is ensured that the second NVE device can receive the message transmitted by the PE device and the other NVE devices.
[0075] In the embodiment of the present application, the IMET route mentioned in RFC7432 is also extended so that the IMET route can carry the VTEP address of the NVE device that sends the IMET route. For example, the management IP address (or device IP address) of the NVE device can be used as the VTEP address. In this way, the second IMET route carries both the public VTEP address 9.9.9.9 and the second VTEP address 1.1.1.1.
[0076] As a preferred solution, the embodiment of the present application extends a new attribute to carry the VTEP address. Figure 6 As shown, the embodiment of the present application extends the VXLAN attribute to carry the VTEP address. Among them, the type field and the subtype field are used to indicate the type of the VXLAN attribute, and the specific value can be set according to the requirements of the standard organization. The flag field can be used to indicate whether the VXLAN attribute is allowed to be read. For example, the flag field is 0 to indicate that the VTEP address in the VXLAN attribute is not allowed to be read, and the flag field is 1 to indicate that the VTEP address in the VXLAN attribute is allowed to be read. It should be understood that the function of the above-mentioned flag field is an example, and the function of the flag field can be transformed accordingly according to the usage scenario. The VTEP address field is used to carry the value of the VTEP address. The reserved field is for further expansion. The VXLAN attribute can be carried in the IMET route as an attribute to implement the solution of the embodiment of the present application. Taking the second IMET route as an example, the second IMET route carries a VXLAN attribute, and the VXLAN attribute includes a second VTEP address.
[0077] The second IMET route may also carry a VNI, and the VNI is used to indicate the BD to which the second NVE device belongs. More specifically, the VNI is used to indicate the BD to which the port of the second NVE device sending the second IMET route belongs. In the present application, a BD may be referred to as a broadcast domain (English: Broadcast Domain, abbreviated: BD) or a bridge domain (English: Bridge Domain, abbreviated: BD). The VNI may be carried in the provider multicast service interface (English: Provider Multicast Service Interface, abbreviated: PMSI) tunnel attribute of the second IMET route. The PMSI tunnel attribute includes an MPLS label field for carrying the VNI.
[0078] In the embodiment of the present application, the VNI can be configured in two forms, namely, a public VTEP VNI and a VTE VNI. The public VTEP VNI is used to bind to a public VTEP address, and the VTEP VNI is used to bind to a VTEP address. Figure 1, the endpoint address of the VXLAN tunnel established between the PE device and the NVE device set uses the public VTEP address, therefore, the VXLAN tunnel established between the PE device and the NVE device set corresponds to the public VTEP VNI. The endpoint address of the VXLAN tunnel established between the NVE devices in the NVE device set uses the VTEP address, therefore, the VXLAN tunnel established between the NVE devices corresponds to the VTEP VNI. In the embodiment of the present application, the public VTEP VNI and the VTEP VNI can be set to be the same or different.
[0079] If the public VTEP VNI and the VTEP VNI are set to be the same, the VXLAN tunnel established between the PE device and the NVE device set and the VXLAN tunnel established between the NVE devices in the NVE device set correspond to the same BD. In the case where the public VTEP VNI and the VTEP VNI are set to be the same, the VNI information can be carried according to the above implementation method. That is, the VNI is carried in the MPLS label field included in the PMSI tunnel attribute, and the VNI is used to indicate both the public VTEP VNI and the VTEP VNI. In the process of establishing the VXLAN tunnel between the PE device and the NVE device set and establishing the VXLAN tunnel between the NVE devices in the NVE device set, the VNI carried in the MPLS label field included in the PMSI tunnel attribute is used.
[0080] If the public VTEP VNI and the VTEP VNI are set to be different, the VXLAN tunnel established between the PE device and the NVE device set, and the VXLAN tunnel established between the NVE devices in the NVE device set correspond to different BDs. For example, the VXLAN tunnel established between the PE device and the NVE device set corresponds to BD1, and the VXLAN tunnel established between the NVE devices in the NVE device set corresponds to BD2. The above-mentioned "setting the public VTEP VNI and the VTEP VNI to be different" means that the VXLAN tunnels established between the NVE devices in the NVE device set all use the same VTEPVNI, and the VTEP VNI is different from the public VTEP VNI. In the case where the public VTEP VNI and the VTEP VNI are set to be different, the VNI information can be carried in the following implementation method. The public VTEP VNI is carried in the MPLS label field included in the PMSI tunnel attribute. Moreover, the embodiment of the present application also extends a new attribute to carry the VTEP VNI. As Figure 7As shown, the embodiment of the present application extends the VTEP VNI attribute to carry the VTEP VNI. Among them, the type field and the subtype field are used to indicate the type of the VTEP VNI attribute, and the specific value can be set according to the requirements of the standard organization. The flag field can be used to indicate whether the VTEP VNI attribute is allowed to be read. For example, the flag field is 0 to indicate that the VTEP VNI in the VTEP VNI attribute is not allowed to be read, and the flag field is 1 to indicate that the VTEPVNI in the VTEP VNI attribute is allowed to be read. It should be understood that the function of the above-mentioned flag field is an example, and the function of the flag field can be transformed accordingly according to the usage scenario. The VTEP VNI field is used to carry the value of the VTEP VNI. The reserved field is for further expansion. The VTEP VNI attribute can be carried as an attribute in the IMET route. When the PE device receives the IMET route issued by the NVE device in the NVE device set, it establishes a VXLAN tunnel according to the public VTEP VNI, and does not parse the VTEP VNI attribute, thereby ignoring the VTEP VNI. Accordingly, after receiving the IMET route issued by other NVE devices, the NVE devices in the NVE device set preferentially establish the VXLAN tunnel according to the VTEP VNI in the VTEP VNI attribute. If the NVE devices in the NVE device set determine that the VTEP VNI attribute does not exist in the received IMET route, the public VTEP VNI carried in the MPLS label field included in the PMSI tunnel attribute is used to establish the corresponding VXLAN tunnel.
[0081] Furthermore, the VXLAN tunnels established between the NVE devices in the NVE device set may also use different VTEP VNIs. Figure 1 In the embodiment, the first VXLAN tunnel uses VTEP VNI1, corresponding to BD21, the second VXLAN tunnel uses VTEP VNI2, corresponding to BD22, and the third VXLAN tunnel uses VTEP VNI3, corresponding to BD23. In this way, the NVE device is allowed to configure multiple VNIs so that the NVE device can establish VXLAN tunnels corresponding to different BDs. In the embodiment of the present application, in order to make the technical solution easier to understand, the public VTEP VNI and the VTEP VNI are set to be the same, and VNI can be used in the subsequent description to represent the public VTEP VNI and the VTEP VNI. It should be understood that the embodiment of the present application is not limited to setting the public VTEP VNI and the VTEP VNI to be the same. The public VTEP VNI and the VTEP VNI can be set to be different. Further, different VTEP VNIs can be set. The embodiment of the present application is implemented in combination with the above description.
[0082] The Route Distinguisher (RD) field in the second IMET route is used to indicate the EVPN instance (EVI) corresponding to the second NVE device. The Ethernet tag (Ethernet tag) field in the second IMET route can be set to the Virtual Local Area Network Identifier (VLAN ID or VID) corresponding to the MAC address included in the CE device connected to the second NVE device or the BD corresponding to the second NVE device. The IP Address Length (IP Address Length) field in the second IMET route is used to indicate the length of the Originating Router's IP Address.
[0083] The PMSI tunnel attributes of the second IMET route also include a tunnel type (English: Tunnle Type) field, and the tunnel type field is used to indicate the type of tunnel to be established. In the embodiment of the present application, the value of the tunnel type field is set to 6, which is used to identify the type of tunnel to be established as ingress replication. The second IMET route also includes VXLAN tunnel attributes, and the format of the VXLAN tunnel attributes is the same as Figure 6 or Figure 7 The properties shown in the table are similar in format, except that: Figure 6 VTEP address field in or Figure 7 The VTEP VNI field in is replaced with the VXLAN tunnel type field. The VXLAN tunnel type field is used to indicate that the established tunnel is a VXLAN tunnel. The value of the VXLAN tunnel type field is specifically determined by the standard organization, for example, the value of the VXLAN tunnel type field is 8. For the implementation method of the first NVE device establishing VXLAN using the second IMET route, please refer to the subsequent description of this embodiment. The PMSI tunnel attribute of the second IMET route also includes a tunnel identifier (English: Tunnle Identifier) field. In this embodiment of the present application, the tunnel identifier field can be filled with the public VTEP address.
[0084] S103: The first NVE device determines whether the public VTEP address included in the Originating Router's IPAddress field in the second IMET route is the same as the public VTEP address stored in the first NVE device.
[0085] The first NVE device can send an IMET route carrying the public VTEP address stored by the first NVE device in the OriginatingRouter's IP Address field to the BGP peer of the first NVE device. The public VTEP address included in the Originating Router's IP Address field in the second IMET route is the same as the public VTEP address stored by the first NVE device.
[0086] Since the second NVE device publishes the second IMET route externally, and the first NVE device and the second NVE device are a pair of BGP peers, the first NVE device can receive the second IMET route.
[0087] After receiving the second IMET route, the first NVE device parses the second IMET route and obtains the public VTEP address included in the Originating Router's IP Address field in the second IMET route. The first NVE device determines whether the public VTEP address included in the Originating Router's IP Address field in the second IMET route is the same as the public VTEP address stored in the first NVE device. According to an embodiment of the present application, both the first NVE device and the second NVE device include a public VTEP, so the public VTEP address included in the Originating Router's IP Address field in the second IMET route is the same as the public VTEP address stored in the first NVE device. Among them, when the first NVE device publishes the IMET route externally, the public VTEP address stored in the first NVE device is carried in the Originating Router's IP Address field of the IMET route. That is, the public VTEP address stored in the first NVE device is the public VTEP address in the public VTEP included in the first VTEP device. If the first NVE device determines that the public VTEP address included in the Originating Router's IPAddress field in the second IMET route is the same as the public VTEP address stored by the first NVE device, the first NVE device can determine that the first NVE device and the second NVE device belong to the same NVE device set.
[0088] S104. When the first NVE device determines that the public VTEP address included in the Originating Router's IPAddress field in the second IMET route is the same as the public VTEP address stored by the first NVE device, the first NVE device establishes a VXLAN tunnel from the first NVE device to the second NVE device according to the second VTEP address in the second IMET route.
[0089] Based on the result of the above judgment, the first NVE device can establish a VXLAN tunnel from the first NVE device to the second NVE device according to the second VTEP address. Specifically, the first NVE device continues to parse the second IMET route, obtains the second VTEP address in the second IMET route, and establishes a VXLAN tunnel from the first NVE device to the second NVE device using the second VTEP address as the destination endpoint address of the VXLAN tunnel.
[0090] In a possible implementation, in combination with the description of S101 and S102, the second VTEP address is carried in the VXLAN attribute. Based on the result of the above judgment, that is, after the first NVE device determines that the public VTEP address included in the Originating Router's IP Address field in the second IMET route is the same as the public VTEP address stored by the first NVE device, the first NVE device checks whether the second IMET route carries the VXLAN attribute. The first NVE device determines that the second IMET route carries the VXLAN attribute, obtains the second VTEP address in the VXLAN attribute, and establishes a VXLAN tunnel from the first NVE device to the second NVE device using the second VTEP address as the destination endpoint address of the VXLAN tunnel.
[0091] For example, the first NVE device determines that the type of tunnel to be established is head-end replication based on the value of the tunnel type field in the PMSI tunnel attribute of the second IMET route being 6. The first NVE device parses the VXLAN tunnel attribute of the second IMET route, obtains the value of the VXLAN tunnel type field in the VXLAN tunnel attribute, and thereby determines that the established tunnel is a VXLAN tunnel. After determining that the established tunnel is a VXLAN tunnel, the first NVE device generates a VXLAN tunnel table and a VXLAN tunnel head-end replication table. The VXLAN tunnel table includes a tunnel identifier, a source endpoint address, and a destination endpoint address. The tunnel identifier is used to identify the VXLAN tunnel. The source endpoint address is the first VTEP address included in the first NVE device, such as 1.1.1.1. The destination endpoint address is the second VTEP address obtained from the second IMET route, such as 2.2.2.2. The VXLAN tunnel head-end replication table includes a VNI and a BGP peer address. The VNI can be obtained from the MPLS label field included in the PMSI tunnel attribute or from the VTEP VNI attribute according to the above description. The BGP peer address is the second VTEP address obtained from the second IMET route, for example, 2.2.2.2. Thus, the first NVE device establishes a VXLAN tunnel from the first NVE device to the second NVE device.
[0092] If the first NVE device determines that the VTEP address included in the Originating Router's IPAddress field in the received IMET route is not the same as the public VTEP address stored by the first NVE device, the first NVE device can determine that the first NVE device and the device sending the IMET route do not belong to the same NVE device set. The first NVE device will establish a VXLAN tunnel from the first NVE device to the device sending the IMET route based on the VTEP address included in the Originating Router's IP Address field in the received IMET route, rather than establishing a VXLAN tunnel based on the VTEP address in the VXLAN attributes. Similarly, please refer to the process of establishing a VXLAN tunnel between a PE device and a NVE device set described in the subsequent embodiments.
[0093] If the first NVE device determines that the public VTEP address included in the Originating Router's IPAddress field in the second IMET route is the same as the public VTEP address stored by the first NVE device, but cannot further find the second VTEP address or the VXLAN attribute, the first NVE device will discard the second IMET route and return an error message to the second NVE device. Further optionally, the first NVE device can also feedback the error message to the network management device of EVPNVXLAN. Such a setting helps to improve the reliability of VXLAN tunnel establishment.
[0094] Optionally, S105, the first NVE device sends a first IMET route to the second NVE device, the Originating Router's IP Address domain of the first IMET route carries the public VTEP address stored by the first NVE device, wherein the first IMET route also includes a first VTEP address, and the first VTEP address is used by the second NVE device to establish a VXLAN tunnel from the second NVE device to the first NVE device, and the public VTEP address stored by the first NVE device is different from the first VTEP address.
[0095] The implementation process of the first NVE device sending the first IMET route to the second NVE device and the implementation method of the first IMET route can refer to the description of S101 and S102 in the present embodiment, which will not be repeated here.
[0096] After receiving the first IMET route, the second NVE device may establish a VXLAN tunnel from the second NVE device to the first NVE device according to the first IMET route. The implementation process of the second NVE device establishing a VXLAN tunnel from the second NVE device to the first NVE device according to the first IMET route can refer to the description of S103 and S104 of this embodiment, which will not be repeated here.
[0097] It should be understood that the above embodiment does not limit the execution order of S105 and S101-S104. For example, S105 may be executed before S101, or S105 and S101 may be triggered at the same time, or S105 may be executed after S101.
[0098] Through the above process, a VXLAN tunnel can be established between the first NVE device and the second NVE device, and the endpoint addresses at both ends of the VXLAN tunnel are the first VTEP address and the second VTEP address.
[0099] Through the above implementation, in the application scenario of EVPN VXLAN, the NVE device set consisting of at least two NVE devices includes a public VTEP address, which is used to identify the public VTEP and is deployed on each NVE device in the NVE device set. Moreover, each NVE device in the NVE device set also includes a VTEP address, which is used to identify the VTEP included in the corresponding NVE device. The NVE devices in the NVE device set establish VXLAN tunnels between the NVE devices in the NVE device set through IMET routing based on the VTEP address and the public VTEP address. Therefore, in the application scenario of EVPN VXLAN, the NVE devices in the NVE device set are interconnected through VXLAN tunnels, so that the NVE devices in the NVE device set are not restricted by directly connected physical links. The interconnection is achieved using VXLAN tunnels, and even if there are switching nodes between the NVE devices in the NVE device set, it will not affect the implementation of EVPN VXLAN. Moreover, when deploying the NVE devices in the NVE device set, it will not be affected by the deployment region and distance, reducing the complexity of deployment. Furthermore, the interconnection lines of the NVE devices in the NVE device set are no longer affected by the direct physical links, which helps to improve the reliability of communication.
[0100] According to the above embodiments, Figure 1 Optionally, the EVPN VXLAN further includes a PE device. The PE device and the first NVE device are a pair of BGP peers, the PE device and the second NVE device are a pair of BGP peers, and the PE device and the third NVE device are a pair of BGP peers. Figure 1 Taking the example of FIG. 1 as an example, the process of establishing a VXLAN tunnel between the PE device and the NVE device set is described. The method for processing routing also includes:
[0101] S201. The first NVE device sends the first IMET route to the PE device.
[0102] S202: The second NVE device sends the second IMET route to the PE device.
[0103] S203: The third NVE device sends the third IMET route to the PE device.
[0104] According to the above embodiments, Figure 2According to the description, the first NVE device, the second NVE device and the third NVE device as NVE devices in the NVE device set all publish IMET routes to their respective BGP peers. That is, the first NVE device publishes the first IMET route to the BGP peer of the first NVE device; the second NVE device publishes the second IMET route to the BGP peer of the second NVE device; the third NVE device publishes the third IMET route to the BGP peer of the third NVE device. The implementation of the first IMET route, the second IMET route and the third IMET route refers to the description of the aforementioned embodiment, which will not be repeated here.
[0105] S204. The PE device receives the first IMET route, the second IMET route, and the third IMET route.
[0106] S205. The PE device establishes a VXLAN tunnel from the PE device to the NVE device set according to the first IMET route, the second IMET route, and the third IMET route.
[0107] Since the PE device is the BGP peer of each NVE device in the NVE device set, the PE device can receive the first IMET route, the second IMET route and the third IMET route. The above S204 and S205 can be implemented in two possible ways.
[0108] In a possible implementation, the PE device waits to receive IMET routes sent by all BGP peers before establishing a VXLAN tunnel. After receiving the above three IMET routes, the PE device parses the above three IMET routes respectively, and obtains the public VTEP addresses included in the Originating Router's IP Address field in the above three IMET routes respectively. The PE device determines that the public VTEP addresses included in the Originating Router's IP Address field in the above three IMET routes are the same, and can determine that the NVE devices sending the above three IMET routes belong to the same NVE device set. The PE device selects any one IMET route from the three IMET routes, and the PE device uses the public VTEP address included in the selected IMET route as the destination endpoint address of the VXLAN tunnel to establish a VXLAN tunnel from the PE device to the public VTEP including the public VTEP address.
[0109] In another possible implementation, after the PE device receives any one of the above three IMET routes, it establishes a VXLAN tunnel. For example, the PE device receives the first IMET route sent by the first NVE device, parses the first IMET route, and obtains the public VTEP address included in the Originating Router's IPAddress field in the first IMET route. The PE device uses the public VTEP address included in the first IMET route as the destination endpoint address of the VXLAN tunnel to establish a VXLAN tunnel from the PE device to the public VTEP including the public VTEP address. After the VXLAN tunnel is established, when the PE device receives the second IMET route sent by the second NVE device or the third IMET route sent by the third NVE device, it determines that the public VTEP address included in the Originating Router's IP Address field in the second IMET route or the third IMET route is the same as the public VTEP address used to establish the VXLAN tunnel, and no longer repeatedly establishes the VXLAN tunnel from the PE device to the public VTEP including the public VTEP address.
[0110] In the above two implementations, the specific implementation method of establishing the VXLAN tunnel can refer to the description of the aforementioned embodiment, which will not be repeated here. In the above two implementations, the public VTEP addresses included in the Originating Router's IP Address field in the above three IMET routes are the same. After the PE device receives the IMET route with the same public VTEP address, the PE device believes that there is only one VTEP corresponding to the public VTEP address on the peer end of the PE device. Therefore, the PE device only establishes a VXLAN tunnel once for the public VTEP address, instead of establishing a VXLAN tunnel for the public VTEP address stored in each NVE device in the NVE device set. In this way, the VXLAN tunnel established by the PE device from the PE device to the public VTEP including the public VTEP address using the public VTEP address as the destination endpoint address of the VXLAN tunnel can be called a VXLAN tunnel from the PE device to the NVE device set.
[0111] Optionally, the above three IMET routes all include a VNI, and for establishing a VXLAN tunnel, the VNI is the aforementioned public VTEP VNI. After determining that the above three IMET routes include the same public VTEP address, the PE device further determines that the above three IMET routes include the same public VTEP VNI, thereby establishing the VXLAN tunnel from the PE device to the NVE device set. Among them, the explanation of the public VTEP VNI can refer to the corresponding description of the aforementioned embodiment, which will not be repeated here.
[0112] S206. The PE device sends an IMET route to the first NVE device, the second NVE device and the third NVE device respectively. The next hop domain and the Originating Router's IP Address domain in the IMET route both include the VTEP address of the PE device. The VTEP address of the PE device is used to identify the VTEP included in the PE device.
[0113] The PE device may publish the IMET route to the BGP peer of the PE device. The format and basic usage of the IMET route may refer to the description of RFC7432. As an example, in the embodiment of the present application, the value of the VTEP address included in the VTEP of the PE device is set to 4.4.4.4. The following is an example of the first NVE device receiving the IMET route sent by the PE device. It should be understood that the second NVE device and the third NVE device may perform the same processing.
[0114] S207: The first NVE device receives the IMET route from the PE device.
[0115] S208. The first NVE device establishes a VXLAN tunnel from the first NVE device to the PE device according to the VTEP address from the PE device, and the VTEP address of the PE device is not equal to the public VTEP address stored by the first NVE device.
[0116] After receiving the IMET route sent by the PE device, the first NVE device parses the IMET route, obtains the VTEP address of the PE device carried in the Originating Router's IP Address field in the IMET route, and uses the VTEP address as the destination endpoint address of the VXLAN tunnel to establish a VXLAN tunnel from the first NVE device to the PE device. Optionally, the IMET route also includes a public VTEP VNI, and the first NVE device obtains the public VTEP VNI and establishes a VXLAN tunnel from the first NVE device to the PE device corresponding to the public VTEP VNI.
[0117] Optionally, after obtaining the VTEP address of the PE device, the first NVE device may perform a determination process similar to the aforementioned S103 and S104, that is, the first NVE device determines that the VTEP address of the PE device is different from the public VTEP address saved by the first NVE device, and the first NVE device may determine that the first NVE device and the PE device do not belong to the same NVE device set. The first NVE device will use the public VTEP address as the source endpoint address to establish a VXLAN tunnel from the first NVE device to the PE device, instead of using the first VTEP address as the source endpoint address to establish a VXLAN tunnel from the first NVE device to the PE device.
[0118] According to S207 and S208, the second NVE device establishes a VXLAN tunnel from the second NVE device to the PE device, and the third NVE device establishes a VXLAN tunnel from the third NVE device to the PE device. In combination with the above, the three VXLAN tunnels established by the first NVE device, the second NVE device and the third NVE device to the PE device are the same, that is, the source endpoint addresses of the three VXLAN tunnels are all public VTEP addresses, the destination endpoint addresses are all VTEP addresses of the PE device, and the three VXLAN tunnels all correspond to the public VTEP VNI. Correspondingly, the PE device establishes a VXLAN tunnel from the PE device to the NVE device set, the source endpoint addresses are all VTEP addresses of the PE device, and the destination endpoint addresses are all public VTEP addresses. In this way, the PE device will think that there is only one NVE device on the opposite end, but in fact, the NVE device is the NVE device set. Moreover, the first NVE device, the second NVE device and the third NVE device can still send traffic through the VXLAN tunnels established with the PE device.
[0119] Based on the above embodiments, the embodiments of the present application can be applied to a network scenario where more than two NVE devices form a NVE device set. Figure 1 For example, the NVE device set includes 3 NVE devices. Figure 2 In the method shown, a first VXLAN tunnel is established between the first NVE device and the second NVE device, a second VXLAN tunnel is established between the second NVE device and the third NVE device, and a third VXLAN tunnel is established between the first NVE device and the third NVE device. In this way, the technical solution of the present application can realize a network scenario in which a CE device is connected to three or more NVE devices, that is, the CE device is multi-owned by three or more NVE devices. Figure 1 For example, the first CE device is multi-owned to the first NVE device, the second NVE device and the third NVE device. The above implementation method is an all-active redundancy mode (English: All-Active Redundancy Mode) of EVPN VXLAN, and multiple NVE devices are connected to at least one CE device through ES, wherein multiple NVE devices form a device set of all-active redundancy mode, all of which are used to forward traffic. In comparison, the existing all-active redundancy mode of EVPN VXLAN adopts a dual-active redundancy mode, and the CE device can only be dual-owned to two NVE devices at most; and the two NVE devices are directly connected through a physical link, for example, the two NVE devices are connected through a PeerLink link. Therefore, a further beneficial effect of the embodiments of the present application is that it supports the all-active redundancy mode of more NVE devices, which helps to increase the reliability of EVPN VXLAN and increase the flexibility of network deployment.
[0120] Figure 3 This is another schematic diagram of the EVPN VXLAN structure of an embodiment of the present application. Figure 3 It is shown that after the VXLAN tunnel between the NVE devices in the NVE device set is established by using the technical solution provided by the aforementioned embodiment, the designated forwarder (English: Designated Forwarder, abbreviation: DF) election and the forwarding of broadcast, unknown unicast and multicast (English: Broadcast, Unknown unicast, and Multicast, BUM) traffic are realized. Figure 3Three CE devices are shown as an example, namely, a first CE device, a second CE device and a third CE device. The first CE device is multi-owned by the first NVE device, the second NVE device and the third NVE device, that is, the first CE device communicates with the first NVE device, the second NVE device and the third NVE device respectively. The second CE device is dual-owned by the first NVE device and the second NVE device, that is, the second CE device communicates with the first NVE device and the second NVE device respectively. The third CE device is singly owned by the second NVE device, that is, the third CE device communicates with the second NVE device. Figure 3 The following description is given by taking the case where the BUM traffic is forwarded from the PE device to the first CE device, the second CE device and the third CE device as an example.
[0121] The first CE device is multi-homed to the first NVE device, the second NVE device and the third NVE device via multiple Ethernet links. Therefore, the multiple Ethernet links constitute an ES. Figure 3 As shown in ES1. The ESI value of ES1 of the three Ethernet links connecting the first CE device to the first NVE device, the second NVE device and the third NVE device is the same value. The first NVE device, the second NVE device and the third NVE device respectively send Ethernet segment (English: Ethernet Segment, abbreviation: ES) routes to their respective BGP peers, and the ES routes are used for DF election. The ES route includes the RD field, ESI field, IP address length field and Originating Router's IP Address field. The definitions of the above-mentioned field fields are similar to the aforementioned IMET route and are not repeated here. In addition, the format and basic usage of ES routes can also be found in the descriptions of Chapter 7.4 and Chapter 8.5 of RFC7432. Figure 3 For example, the first CE device in the example receives ES routes from the second NVE device and the third NVE device. The first NVE device determines the first NVE device as the DF in the ES composed of the multiple Ethernet links based on the ES routes received from the second NVE device and the third NVE device. Similarly, the second NVE device is determined as a non-designated forwarder (NDF) in the ES composed of the multiple Ethernet links; the third NVE device is determined as an NDF in the ES composed of the multiple Ethernet links. Figure 3As shown, the ES1 between the first CE device and the first NVE device is DF, the ES1 between the first CE device and the second NVE device is NDF, and the ES1 between the first CE device and the third NVE device is NDF. The second CE device is dual-homed to the first NVE device and the second NVE device via two Ethernet links, and the two Ethernet links constitute an ES, such as Figure 3 As shown in ES2. The ESI value of ES2 of the two Ethernet links connecting the second CE device to the first NVE device and the second NVE device is the same value. Similarly, the second NVE device is determined as the DF in the ES composed of the two Ethernet links; the first NVE device is determined as the NDF in the ES composed of the two Ethernet links. The third CE device is solely affiliated to the second NVE device, so there is no DF election process.
[0122] The NVE device may send an ES route to the BGP peer of the NVE device in the following manner: The NVE device may send an UPDATE message to the BGP peer of the NVE device, and the UPDATE message may carry an MP_REACH_NLRI attribute. The MP_REACH_NLRI attribute may include an EVPN NLRI domain and a next hop domain. The EVPN NLRI domain may be used to carry an ES route.
[0123] The PE device sends BUM traffic to the NVE device set based on the VXLAN tunnel. Specifically, the PE device sends a VXLAN message to the NVE device set according to a hash rule, and the VXLAN message carries the BUM traffic. Optionally, the VXLAN message also carries a BUM tag to identify that the traffic carried by the VXLAN message is BUM traffic. In a possible implementation, the BUM tag can be carried in the VXLAN header of the VXLAN message.
[0124] The VXLAN message is hashed to the second NVE device as an example for explanation. In this embodiment and subsequent embodiments, in order to distinguish the VXLAN message transmitted between the PE device and the NVE device set from the VXLAN message transmitted between the NVE devices in the NVE device set, the VXLAN message transmitted between the PE device and the NVE device set is named as a PE VXLAN message. The second NVE device receives the PE VXLAN message from the PE device, parses the PE VXLAN message, and according to the BUM mark carried in the PE VXLAN message, it can be determined that the traffic carried by the PE VXLAN message is BUM traffic. The second NVE device forwards the BUM traffic to the CE device that has a connection relationship with the second NVE device, and the second NVE device also forwards the BUM traffic to other NVE devices in the NVE device set.
[0125] The second NVE device forwards the BUM traffic to the CE device connected to the second NVE device: the second NVE device determines that ES2 between the second NVE device and the second CE device is in the DF state, and the second NVE device transmits the BUM traffic to the output interface connecting the second NVE device and ES2, and sends it to the second CE device via ES2. The second NVE device determines that ES1 between the second NVE device and the first CE device is in the NDF state, and the second NVE device does not transmit the BUM traffic to the output interface connecting the second NVE device and ES1, so that the second NVE device avoids sending the BUM traffic to the first CE device via ES1. In addition, the second NVE device determines that ES3 between the second NVE device and the third CE device does not exist in the DF or NDF state, and directly sends the BUM traffic to the third CE device via ES3.
[0126] The second NVE device forwards the BUM traffic to other NVE devices in the NVE device set: the second NVE device parses the PE VXLAN message from the PE device to obtain the BUM traffic. The second NVE device determines that the BUM traffic comes from the PE device, encapsulates the BUM traffic into a VXLAN message, sends it to the first NVE device via the first VXLAN tunnel, and sends it to the second NVE device via the second VXLAN tunnel.
[0127] After receiving the VXLAN message via the first VXLAN tunnel, the first NVE device can determine that the BUM traffic in the VXLAN message is received via the first VXLAN tunnel, and no longer forward the BUM traffic to the third VXLAN tunnel connected to the first NVE device. Figure 3 As shown, after the first NVE device receives the VXLAN message via the first VXLAN tunnel, it parses the VXLAN message, obtains BUM traffic, and no longer forwards the BUM traffic to the third NVE device via the third VXLAN tunnel. In other words, the first NVE device avoids forwarding the BUM traffic to the third NVE device via the VXLAN tunnel from the first NVE device to the third NVE device based on the fact that the BUM traffic is received via the first VXLAN tunnel. The beneficial effect based on the above is that double sending of BUM traffic is avoided.
[0128] After receiving the VXLAN message via the first VXLAN tunnel, the first NVE device parses the VXLAN message, obtains BUM traffic, and sends the BUM traffic to the CE device connected to the first NVE device. Figure 3 As shown, the first NVE device determines that ES1 between the first NVE device and the first CE device is in the DF state, and the first NVE device transmits the BUM traffic to the outbound interface connected to ES1 of the first NVE device, and sends it to the first CE device via ES1. The first NVE device determines that ES2 between the first NVE device and the second CE device is in the NDF state, and the first NVE device does not transmit the BUM traffic to the outbound interface connected to ES2 of the first NVE device, so that the first NVE device avoids sending the BUM traffic to the second CE device via ES2.
[0129] After receiving the VXLAN message via the second VXLAN tunnel, the third NVE device parses the VXLAN message and obtains the BUM traffic. On the one hand, the third NVE device avoids forwarding the BUM traffic to the first NVE device via the third VXLAN tunnel. On the other hand, the third NVE device determines that ES1 between the third NVE device and the first CE device is in the NDF state, and the third NVE device does not transmit the BUM traffic to the output interface connecting the third NVE device and ES1, so that the third NVE device avoids sending the BUM traffic to the first CE device via ES1.
[0130] Through the above implementation, using the VXLAN tunnel between NVE devices can not only realize the normal forwarding of BUM traffic, but also avoid forwarding of BUM traffic to the VXLAN tunnel between other NVE devices, thereby avoiding double sending of BUM traffic.
[0131] Similarly, forwarding of BUM traffic from CE devices to PE devices and other CE devices is the reverse process of the above implementation, so the forwarding of BUM traffic can be completed with reference to the above method. For example, BUM traffic is forwarded from the second CE device to the PE device, the first CE device, and the third CE device. BUM reaches the second NVE device via ES2 between the second NVE device and the second CE device. The second NVE device encapsulates the BUM traffic as a PE VXLAN message, and forwards it to the PE device via the VXLAN tunnel between the second NVE device and the PE device. The second NVE device also encapsulates the BUM traffic as a VXLAN message, and forwards it to the first NVE device and the third NVE device via the first VXLAN tunnel and the second VXLAN tunnel, respectively. The second NVE device also forwards the BUM traffic directly to the third CE device via ES3, but avoids forwarding the BUM traffic to the first CE device via ES1 between the second NVE device and the first CE device. After receiving the BUM traffic from the second NVE device, the first NVE device avoids forwarding the BUM traffic to the third NVE device via the third VXLAN tunnel, and avoids forwarding the BUM traffic to the PE device via the VXLAN tunnel. The first NVE device forwards the BUM traffic to the first CE device via ES1, but avoids forwarding the BUM traffic to the second CE device via ES2. After receiving the BUM traffic from the second NVE device, the third NVE device avoids forwarding the BUM traffic to the first NVE device via the third VXLAN tunnel, and avoids forwarding the BUM traffic to the first CE device via ES1, and further avoids forwarding the BUM traffic to the PE device via the VXLAN tunnel.
[0132] In the above-mentioned example, in which the BUM traffic is forwarded from the second CE device to the PE device, the first CE device, and the third CE device, the BUM sent by the second CE device reaches the second NVE device via the ES2 between the second NVE device and the second CE device. However, in an actual scenario, the BUM sent by the second CE device may reach the first NVE device via the ES2 between the first NVE device and the second CE device, because the second CE device does not care about the DF or NDF status of the communication link. Figure 4 As shown in FIG. 1 , when the second CE device forwards the BUM to the first NVE device via ES2. The first NVE device encapsulates the BUM traffic into a VXLAN message and forwards it to the second NVE device via the first VXLAN tunnel. Since ES2 between the second NVE device and the second CE device is in the DF state, the second NVE device forwards the BUM traffic to the second CE device via ES2. Figure 4 As shown by the dotted arrow between the second NVE device and the second CE device in FIG. 1 , the second CE device, the first NVE device and the second NVE device form a BUM traffic loop. It should be understood that in order to present the loop problem more clearly, Figure 4 The BUM traffic forwarding line shown in the figure is only the part where the loop occurs. Figure 4 The entire forwarding path of BUM traffic in EVPN VXLAN is not shown.
[0133] Combine the following Figure 4 A method for solving the BUM traffic loop problem formed by the second CE device, the first NVE device and the second NVE device is specifically described. The second CE device is dual-homed to the first NVE device and the second NVE device via two Ethernet links, and the two Ethernet links constitute an ES, such as Figure 4 The ESI value of ES2 of the two Ethernet links connecting the second CE device to the first NVE device and the second NVE device is the same value. The second NVE device determines the second NVE device as the DF in the ES formed by the two Ethernet links according to the ES route received from the first NVE device.
[0134] S301. The second NVE device sends an Ethernet AD per ES route to the first NVE device. The Ethernet AD per ES route carries an ESI label allocated by the second NVE device to the Ethernet link between the second CE device and the second NVE device.
[0135] S302: The first NVE device receives the Ethernet AD per ES route.
[0136] Since the first NVE device and the second NVE device are a pair of BGP peers, the second NVE device can send an Ethernet AD per ES route to the first NVE device. The Ethernet AD perES route belongs to an Ethernet Auto-discovery route. The format and basic usage of the Ethernet Auto-discovery route can also be found in the descriptions of Chapter 7.1 and Chapter 8.2 of RFC7432. The Ethernet AD per ES route can carry an ESI Label Extended Community attribute, and the ESI Label Extended Community attribute can carry an ESI label. After the first NVE device receives the Ethernet AD per ES route, it can determine that the second CE device is not only connected to the first NVE device, but also to the second NVE device based on the ESI included in the Ethernet AD per ES route. Moreover, the first NVE device may determine, based on the ESI label carried by the Ethernet AD per ES route, a value of the ESI label allocated by the second NVE device to the Ethernet link between the second CE device and the second NVE device.
[0137] The NVE device may send the Ethernet AD per ES route to the BGP peer of the NVE device in the following manner. The NVE device may send an UPDATE message to the BGP peer of the NVE device, and the UPDATE message carries an MP_REACH_NLRI attribute. The MP_REACH_NLRI attribute includes an EVPN NLRI domain and a next hop domain. The EVPN NLRI domain is used to carry the Ethernet AD per ES route.
[0138] S303. The first NVE device sends a VXLAN message to the second NVE device via the first VXLAN tunnel, the original Ethernet payload in the VXLAN message includes BUM traffic, and the VXLAN message carries the ESI label, wherein the BUM traffic comes from the second CE device.
[0139] S304. The second NVE device receives the VXLAN message sent by the first NVE device via the first VXLAN tunnel.
[0140] like Figure 4 As shown, when the second CE device forwards the BUM to the first NVE device via ES2. After the first NVE device receives the BUM traffic, it determines the ESI of ES1 between the second CE device and the first NVE device, and further determines that the second CE device is not only connected to the first NVE device, but also connected to the second NVE device. Therefore, the first NVE device encapsulates the BUM traffic into a VXLAN message, wherein the original Ethernet payload in the VXLAN message includes the BUM traffic, and the VXLAN message carries the ESI label. In other words, the VXLAN message sent by the first NVE device to the second NVE device via the first Bypass VXLAN tunnel includes not only the BUM traffic, but also carries the ESI label assigned by the second NVE device to the Ethernet link between the second CE device and the second NVE device. For the definition of the original Ethernet payload, see the description of RFC7348.
[0141] Optionally, the VXLAN message carrying the ESI label may be carried according to Figure 5 The format shown is implemented. Figure 5 As shown, the VXLAN message includes a VXLAN tunnel part and an original Ethernet payload, the VXLAN tunnel part includes an outer Eth header (Eth header: Ethernet header), an outer IP header, an outer UDP header and a VXLAN header, and the original Ethernet payload carries the BUM traffic. The VXLAN message also includes an ESI label, and the ESI label is encapsulated between the VXLAN header and the original Ethernet payload in the VXLAN message. Moreover, the VXLAN header carries indication information, and the indication information is used to indicate that the VXLAN message carries the ESI label.
[0142] S305. The second NVE device carries the ESI label according to the VXLAN message to avoid forwarding the BUM traffic to the CE device via the ES between the second CE device and the second NVE device.
[0143] The second NVE device parses the VXLAN message, obtains the ESI label carried by the VXLAN message, and determines, according to the value of the ESI label, that the ESI label is the ESI label assigned by the second NVE device to the Ethernet link between the second CE device and the second NVE device. The second NVE device no longer transmits the BUM traffic to the outbound interface connected to the second NVE device and ES2, thereby avoiding forwarding the BUM traffic to the CE device via the ES between the second CE device and the second NVE device.
[0144] The above method is based on Figure 4 The loop shown is explained as an example, and it should be understood that the above method is not limited to solving the BUM traffic loop problem formed by the second CE device, the first NVE device and the second NVE device. The above method can be applied to solve the loop problem generated in the application scenario of EVPN VXLAN.
[0145] Through the above implementation, the format of the VXLAN message is extended so that the VXLAN message can carry the ESI label, thereby solving the loop problem generated in the process of forwarding BUM traffic.
[0146] The EVPN VXLAN of the embodiment of the present application can also be used for MAC routing processing, so as to Figure 2 After the VXLAN tunnel is established by the method shown in the figure, the EVPN VXLAN can be used to guide the forwarding of unicast traffic. Figure 1 The method for processing EVPN VXLAN for MAC routing is specifically described. The second CE device is dual-homed to the first NVE device and the second NVE device via two Ethernet links, and the two Ethernet links constitute an ES. The ESI value of ES2 of the two Ethernet links connecting the second CE device to the first NVE device and the second NVE device is the same value.
[0147] S401. The first NVE device receives a media access control / Internet protocol advertisement (MAC / IP Advertisement) route from the second NVE device, where the MAC / IP Advertisement route carries a media access control-virtual local area network identifier (MAC-VLAN ID), wherein the MAC-VLAN ID is used to indicate the VLAN to which the MAC address carried by the MAC / IP Advertisement route belongs, and the MAC address carried by the MAC / IP Advertisement route is the MAC address of the CE device, or the MAC address carried by the MAC / IP Advertisement route is the MAC address of a host in an Ethernet virtual private network site (EVPN site) governed by the CE device.
[0148] Since the first NVE device and the second NVE device are a pair of BGP peers, the second NVE device can send a MAC / IP Advertisement route to the first NVE device. Accordingly, the first NVE device receives the MAC / IP Advertisement route from the second NVE device. The format and basic usage of the MAC / IP Advertisement route can also be found in the descriptions of Chapter 7.2, Chapter 9 and Chapter 14 of RFC7432. The embodiment of the present application extends the MAC / IP Advertisement route described in RFC7432 so that the MAC / IP Advertisement route can carry a MAC-VLAN ID. The MAC-VLAN ID is used to indicate the VLAN to which the MAC address carried by the MAC / IP Advertisement route belongs. The MAC address is used to identify the second CE device, or the MAC address is used to identify the host to which the second CE device is connected. The host is a physical device or a VM.
[0149] Optionally, the embodiment of the present application extends a new attribute to carry the MAC-VLAN ID. Figure 8As shown, the embodiment of the present application extends the MAC-VLAN attribute to carry the MAC-VLAN ID. Among them, the type field and the subtype field are used to indicate the type to which the MAC-VLAN attribute belongs, and the specific value can be set according to the requirements of the standard organization. The flag field can be used to indicate whether the MAC-VLAN attribute is allowed to be read. For example, the flag field is 0 to indicate that the MAC-VLAN ID in the MAC-VLAN attribute is not allowed to be read, and the flag field is 1 to indicate that the MAC-VLAN ID in the MAC-VLAN attribute is allowed to be read. It should be understood that the function of the above-mentioned flag field is an example, and the function of the flag field can be changed accordingly according to the usage scenario. The MAC-VLAN ID field is used to carry the value of the MAC-VLAN ID. The reserved field is for further expansion. The MAC-VLAN attribute can be carried in the MAC / IP Advertisement routing as an attribute to implement the solution of the embodiment of the present application.
[0150] S402. The first NVE device determines the local interface of the first NVE that connects the Ethernet link between the second CE device and the first NVE device according to the ESI carried by the MAC / IP Advertisement route and the MAC-VLAN ID, wherein the ESI carried by the MAC / IP Advertisement route is used to indicate the ES between the second CE device and the second NVE device.
[0151] The MAC / IP Advertisement route sent by the second NVE device to the first NVE device includes an ESI field, and the ESI field is used to carry the ESI indicating the Ethernet link between the second CE device and the second NVE device. The value of the ESI of ES2 between the second CE device and the first NVE device and the second NVE device is equal, so the first NVE device can determine the local interface of the first NVE that connects the Ethernet link between the second CE device and the first NVE device according to the ESI carried by the MAC / IP Advertisement route and the MAC-VLAN ID. In other words, the first NVE device can determine the local interface on the first NVE device according to the ESI and the MAC-VLAN ID, and the local interface is used to communicate with the device identified by the MAC address carried by the MAC / IP Advertisement route.
[0152] S403. The control plane of the first NVE device sends a first MAC table entry to the forwarding plane of the first NVE device, where the outgoing interface in the first MAC table entry is the local interface of the first NVE device, wherein the first MAC table entry includes the MAC address carried by the MAC / IP Advertisement route and the MAC-VLAN ID.
[0153] The first NVE device sends the first MAC table entry including the MAC address and the MAC-VLAN ID from the control plane of the first NVE device to the forwarding plane of the first NVE device, so that the first NVE device can guide the forwarding of unicast traffic according to the first MAC table entry. The outgoing interface in the first MAC table entry is the local interface of the first NVE of the Ethernet link between the second CE device and the first NVE device.
[0154] For example, Figure 1 As shown, the PE device sends unicast traffic to the second CE device. Assume that the PE device hashes the unicast traffic to the second NVE device. The second NVE device receives a VXLAN message from the PE device, and the VXLAN message carries the unicast traffic. The second NVE device obtains the unicast traffic in the VXLAN message, and forwards the unicast traffic to the second CE device based on the MAC table entry stored in the second NVE device.
[0155] like Fig. 9 As shown, if the ES2 between the second NVE device and the second CE device fails, the unicast traffic will not be able to reach the second CE device via the ES2 between the second NVE device and the second CE device. Fig. 9 As shown by the dotted arrow. The first NVE device and the second NVE device have executed the above S401-S403. Therefore, the first NVE device has sent the first MAC table entry including the MAC address and MAC-VLAN ID from the control plane of the first NVE device to the forwarding plane of the first NVE device. The second NVE device sends a VXLAN message carrying the unicast traffic to the first NVE device via the first VXLAN tunnel. After receiving the VXLAN message, the first NVE device obtains the unicast traffic, and according to the first MAC table entry, forwards the unicast traffic to the second CE device via ES2 between the first NVE device and the second CE device.
[0156] Through the above implementation, the forwarding path can be redirected through the VXLAN tunnel between the MAC / IP Advertisement routing and the NVE device, and the normal forwarding of unicast traffic can still be ensured when a link failure occurs on a certain Ethernet link. The above method is described by taking the CE device dual-homed to two NVE devices as an example. It should be understood that the above method can also be applied to the scenario where the CE device is multi-homed to more than two NVE devices.
[0157] Optionally, the method for processing the MAC routing further includes:
[0158] S404. The MAC / IP Advertisement route is carried in the MP_REACH_NLRI, the next hop field in the MP_REACH_NLRI includes a public VTEP address, the MAC / IP Advertisement route also includes the second VTEP address, and the public VTEP address included in the next hop field in the MP_REACH_NLRI is the same as the public VTEP address included in the Originating Router's IP Address field in the second IMET route; the method also includes: the control plane of the first NVE device sends a second MAC table entry to the forwarding plane of the first NVE device according to the public VTEP address included in the next hop field in the MP_REACH_NLRI and the second VTEP address included in the MAC / IP Advertisement route, the outgoing interface in the second MAC table entry is the local interface of the first NVE device of the VXLAN tunnel connecting the first NVE device to the second NVE device, wherein the second MAC table entry includes the MAC address carried by the MAC / IP Advertisement route.
[0159] The second NVE device sending the MAC / IP Advertisement route to the first NVE device can be implemented in the following manner. The second NVE device can send an UPDATE message to the first NVE device, and the UPDATE message carries the MP_REACH_NLRI attribute. The MP_REACH_NLRI attribute includes an EVPN NLRI domain and a next hop domain. The EVPN NLRI domain is used to carry the MAC / IP Advertisement route. The next hop domain includes a public VTEP address, and the MAC / IP Advertisement route also includes the second VTEP address. The second NVE device can determine that the first NVE device and the second NVE device belong to the same NVE device set based on the public VTEP address. The second NVE device can determine that the MAC / IP Advertisement route is issued by the second NVE device based on the second VTEP address. The first NVE device sends a second MAC table entry to the forwarding plane of the first NVE device, wherein the outgoing interface in the second MAC table entry is the local interface of the first NVE device that connects the VXLAN tunnel from the first NVE device to the second NVE device, wherein the second MAC table entry includes the MAC address carried by the MAC / IPAdvertisement route.
[0160] For example, Figure 1 As shown, the PE device sends unicast traffic to the second CE device. Assume that the PE device hashes the unicast traffic to the first NVE device. The first NVE device receives a PE VXLAN message from the PE device, and the PE VXLAN message carries the unicast traffic. The first NVE device obtains the unicast traffic in the PE VXLAN message, and forwards the unicast traffic to the second CE device via ES2 between the first NVE device and the second CE device based on the MAC table entry stored by the first NVE device. Moreover, the first NVE device has sent a second MAC table entry to the forwarding plane of the first NVE device, and the outgoing interface in the second MAC table entry is the local interface of the first NVE device that connects the VXLAN tunnel of the first NVE device to the second NVE device, wherein the second MAC table entry includes the MAC address. Therefore, the first NVE device can forward the unicast traffic to the second NVE device via the first VXLAN tunnel. The second NVE device forwards the unicast traffic to the second CE device via ES2 between the second NVE device and the second CE device based on the MAC table entry stored by the second NVE device.
[0161] Through the above implementation, load sharing of unicast traffic can be further achieved, which helps to improve the transmission bandwidth of unicast traffic.
[0162] The above MAC routing processing is applicable to the scenario where the CE device is dual-homed to two NVE devices or the scenario where the CE device is multi-homed to more than two NVE devices. Figure 1 This section describes in detail how to process MAC routing in the scenario where EVPN VXLAN is used for CE devices to be homed to a single NVE device. Figure 1 As shown, the third CE device is solely affiliated with the second NVE device.
[0163] S501. The first NVE device receives a MAC / IP Advertisement route from the second NVE device, where the MAC / IP Advertisement route is carried in the MP_REACH_NLRI, where the next hop domain in the MP_REACH_NLRI includes a public VTEP address, and the MAC / IP Advertisement route includes a MAC-VLAN ID and the second VTEP address, where the MAC-VLAN ID is used to indicate the VLAN to which the MAC address carried by the MAC / IP Advertisement route belongs, and the MAC address carried by the MAC / IP Advertisement route is the MAC address of the CE device, or the MAC address carried by the MAC / IP Advertisement route is the MAC address of a host in an EVPN site governed by the CE device.
[0164] The execution process in S501 can refer to the corresponding explanation of S401 in the above-mentioned embodiment, which will not be described in detail here.
[0165] S502. The first NVE device determines that there is no local interface connecting the Ethernet link between the CE device and the first NVE device based on the ESI carried by the MAC / IP Advertisement route and the MAC-VLAN ID, wherein the ESI carried by the MAC / IP Advertisement route is used to indicate the ES between the CE device and the second NVE device.
[0166] Combined with the corresponding explanation of S402 in the above-mentioned embodiment, since the third CE device is single-home connected to the second NVE device, the value of the ESI of ES3 between the second NVE device and the third CE device is 0 or an invalid value. The first NVE device receives the MAC / IP Advertisement route from the second NVE device, obtains the ESI carried by the MAC / IP Advertisement route, and finds that the value of the ESI is 0 or an invalid value, then it is considered that there is no ESI. Therefore, the first NVE device determines that there is no local interface connecting the Ethernet link between the CE device and the first NVE device based on the ESI carried by the MAC / IP Advertisement route and the MAC-VLAN ID.
[0167] S503. The control plane of the first NVE device sends a MAC table entry to the forwarding plane of the first NVE device according to the public VTEP address included in the next hop domain in the MP_REACH_NLRI and the second VTEP address included in the MAC / IP Advertisement route. The outbound interface of the MAC table entry is the local interface of the first NVE device that connects the VXLAN tunnel from the first NVE device to the second NVE device. The MAC table entry includes the MAC address carried by the MAC / IP Advertisement route.
[0168] After determining that there is no local interface connecting the ES between the CE device and the first NVE device, the first NVE device executes S503. The execution process in S503 can refer to the corresponding explanation of S404 in the above embodiment, which will not be repeated here.
[0169] For example, Fig.10As shown, the PE device sends unicast traffic to the third CE device. Assume that the PE device hashes the unicast traffic to the first NVE device. The first NVE device receives a PE VXLAN message from the PE device, and the PE VXLAN message carries the unicast traffic. The first NVE device obtains the unicast traffic in the PE VXLAN message, and the first NVE device does not find the MAC table entry from the first NVE device to the third CE device locally. The first NVE device has executed the above S501-S053, so the first NVE device can forward the unicast traffic to the second NVE device via the first VXLAN tunnel. The second NVE device forwards the unicast traffic to the third CE device via ES3 between the second NVE device and the third CE device based on the MAC table entry stored by the second NVE device.
[0170] Through the above implementation, it can be ensured that the CE device that is singly-homed and connected to the NVE device can normally receive unicast traffic.
[0171] The EVPN VXLAN of the embodiment of the present application can also be applied in the scenario of Layer 3 forwarding. Figure 1 As shown, virtual routing and forwarding (English: Virtual Routing and Forwarding, abbreviated: VRF) instances are deployed in PE devices, first NVE devices, second NVE devices and third NVE devices respectively. In this embodiment, for the convenience of explanation, the VRF instances deployed in PE devices, first NVE devices, second NVE devices and third NVE devices are set to be the same VRF instance. It should be understood that each of the above devices can set multiple VRF instances, for example, different VRF instances correspond to different services. The first NVE device, the second NVE device and the third NVE device publish Internet Protocol prefix (English: IP Prefix) routes to their respective BGP peers. The IP Prefix route includes an RD field, an ESI field, an Ethernet tag ID field, an IP Prefix length field, an IP Prefix field, a gateway Internet Protocol address (English: GW IP Address) field and an MPLSLabel field. The definitions of the above-mentioned domains are similar to those in the aforementioned embodiments of the present application, wherein the IP Prefix domain is used to indicate the IP address of the CE device or the host to which the CE device is connected; the MPLS Label domain is used to indicate the three-layer VNI, and the three-layer VNI corresponds to the VRF instance; the GW IP Address domain is set to 0, which is an invalid value.
[0172] The NVE device may send an IP Prefix route to the BGP peer of the NVE device in the following manner. The NVE device may send an UPDATE message to the BGP peer of the NVE device, and the UPDATE message carries an MP_REACH_NLRI attribute. The MP_REACH_NLRI attribute includes an EVPN NLRI field and a next hop field. The EVPN NLRI field is used to carry an IP Prefix route.
[0173] In this embodiment, the next hop domain in the MP_REACH_NLRI attribute carrying the IP Prefix route published externally by the first NVE device, the second NVE device, and the third NVE device includes the public VTEP address of the aforementioned embodiment. As an extension of the IP Prefix route, the IP Prefix route also carries the VTEP address of the NVE device that sends the IP Prefix route. Further optionally, the IP Prefix route can carry the VXLAN attribute described in the aforementioned embodiment, and the VXLAN attribute carries the VTEP address. Through the processing of IP Prefix routing, it is possible to learn IP routing and send IP routing table entries to the forwarding plane, thereby guiding the three-layer forwarding of data traffic. The processing process of IP Prefix routing and the three-layer traffic forwarding process of this embodiment are similar to the MAC routing processing method in the aforementioned embodiment, and will not be repeated here.
[0174] Fig.11 Schematic diagram of the structure of the first NVE device 1000 according to an embodiment of the present application. Fig.11 The first NVE device 1000 shown in the figure can execute the corresponding steps executed by the first NVE device in the method of the above embodiment. Fig.11 As shown, the first NVE device 1000 includes a receiving unit 1002 and a processing unit 1004 .
[0175] The receiving unit 1002 is configured to receive a second IMET route from the second NVE device, wherein the Originating Router's IP Address field in the second IMET route includes a public VTEP address, wherein the second IMET route also includes a second VTEP address, and the public VTEP address included in the Originating Router's IPAddress field in the second IMET route is different from the second VTEP address;
[0176] The processing unit 1004 is used to determine whether the public VTEP address included in the Originating Router's IPAddress field in the second IMET route is the same as the public VTEP address stored in the first NVE device. The first NVE device can send an IMET route carrying the public VTEP address stored in the first NVE device in the Originating Router's IPAddress field to the BGP peer of the first NVE device. The public VTEP address included in the Originating Router's IP Address field in the second IMET route is the same as the public VTEP address stored in the first NVE device.
[0177] When the processing unit 1004 determines that the public VTEP address included in the Originating Router's IPAddress field in the second IMET route is the same as the public VTEP address stored in the first NVE device, the processing unit 1004 is further configured to establish a VXLAN tunnel from the first NVE device to the second NVE device according to the second VTEP address in the second IMET route. Further optionally, the VXLAN attribute in the second IMET route carries the second VTEP address.
[0178] Optionally, the first NVE device also includes a sending unit 1006. The sending unit 1006 is used to send a first IMET route to the second NVE device. The Originating Router's IP Address field of the first IMET route carries the public VTEP address saved by the first NVE device. Among them, the first IMET route also includes a first VTEP address, and the first VTEP address is used by the second NVE device to establish a VXLAN tunnel from the second NVE device to the first NVE device. The public VTEP address saved by the first NVE device is different from the first VTEP address. Optionally, the VXLAN attribute in the first IMET route carries the first VTEP address.
[0179] Optionally, the sending unit 1006 is further used to send the first IMET route to the PE device. The receiving unit 1002 is further used to receive the IMET route from the PE device. The Originating Router's IP Address field in the IMET route from the PE device carries the VTEP address of the PE device. The processing unit 1004 is further used to establish a VXLAN tunnel from the first NVE device to the PE device according to the VTEP address of the PE device in the IMET route from the PE device. The VTEP address of the PE device is not equal to the public VTEP address stored by the first NVE device.
[0180] Optionally, the receiving unit 1002 is further configured to receive a third IMET route from the third NVE device. The Originating Router's IP Address field in the third IMET route includes a public VTEP address. The third IMET route also includes a third VTEP address. The public VTEP address included in the Originating Router's IPAddress field in the third IMET route is different from the third VTEP address. The processing unit 1004 is further used to determine whether the public VTEP address included in the Originating Router's IP Address field in the third IMET route is the same as the public VTEP address stored in the first NVE device, wherein the public VTEP address included in the Originating Router's IP Address field in the third IMET route is the same as the public VTEP address stored in the first NVE device; when the processing unit 1004 determines that the public VTEP address included in the Originating Router's IP Address field in the third IMET route is the same as the public VTEP address stored in the first NVE device, the processing unit 1004 is further used to establish a VXLAN tunnel from the first NVE device to the third NVE device according to the third VTEP address in the third IMET route. Optionally, the VXLAN attribute in the third IMET route carries the third VTEP address.
[0181] Optionally, the CE device is multi-attributed to the first NVE device, the second NVE device and the third NVE device via multiple Ethernet links, and the multiple Ethernet links constitute an ES; the processing unit 1004 is also used to determine the first NVE device as the DF in the ES based on the ES routing from the second NVE device and the third NVE device.
[0182] Optionally, the receiving unit 1002 is also used to receive BUM traffic sent by the second NVE device via the VXLAN tunnel from the second NVE device to the first NVE device; the processing unit 1004 is also used to avoid forwarding the BUM traffic to the third NVE device via the VXLAN tunnel from the first NVE device to the third NVE device based on the fact that the BUM traffic is received by the first NVE device via the VXLAN tunnel from the second NVE device to the first NVE device.
[0183] Optionally, the CE device is dual-owned to the first NVE device and the second NVE device via multiple Ethernet links, and the multiple Ethernet links constitute an ES; the processing unit 1004 is also used to determine the first NVE device as the DF in the ES based on the ES route from the second NVE device; the sending unit 1006 is also used to send an Ethernet AD per ES route to the second NVE device, and the Ethernet AD per ES route carries the ESI label assigned by the first NVE device to the Ethernet link between the CE device and the first NVE device; the receiving unit 1002 is also used to receive a VXLAN message sent by the second NVE device via a VXLAN tunnel from the second NVE device to the first NVE device, the original Ethernet payload in the VXLAN message includes BUM traffic, and the VXLAN message carries the ESI label, wherein the BUM traffic comes from the CE device; the processing unit 1004 is also used to avoid forwarding the BUM traffic to the CE device via the ES between the CE device and the first NVE device based on the VXLAN message carrying the ESI label. Further optionally, the ESI label is encapsulated between the VXLAN header and the original Ethernet payload in the VXLAN message, and the VXLAN header carries indication information, and the indication information is used to indicate that the VXLAN message carries the ESI label.
[0184] Optionally, the CE device is dual-owned to the first NVE device and the second NVE device via multiple Ethernet links, and the multiple Ethernet links constitute an ES. The receiving unit 1002 is also used to receive a MAC / IP Advertisement route from the second NVE device, and the MAC / IP Advertisement route includes a MAC-VLAN ID, wherein the MAC-VLAN ID is used to indicate the VLAN to which the MAC address carried by the MAC / IP Advertisement route belongs. The MAC address carried by the MAC / IP Advertisement route is the MAC address of the CE device, or the MAC address carried by the MAC / IP Advertisement route is the MAC address of a host in an Ethernet virtual private network site EVPNsite governed by the CE device. The processing unit 1004 is also used to determine the local interface of the first NVE that connects the Ethernet link between the CE device and the first NVE device according to the ESI and the MAC-VLAN ID carried by the MAC / IP Advertisement route, wherein the ESI carried by the MAC / IP Advertisement route is used to indicate the ES between the CE device and the second NVE device; the processing unit 1004 is also used to trigger sending a first MAC table entry to the forwarding plane of the first NVE device, the outgoing interface in the first MAC table entry being the local interface of the first NVE device, wherein the first MAC table entry includes the MAC address and the MAC-VLAN ID carried by the MAC / IP Advertisement route.
[0185] Optionally, the MAC / IP Advertisement route is carried in the MP_REACH_NLRI. The next hop field in the MP_REACH_NLRI includes a public VTEP address. The MAC / IP Advertisement route also includes the second VTEP address, and the public VTEP address included in the next hop field in the MP_REACH_NLRI is the same as the public VTEP address included in the Originating Router's IP Address field in the second IMET route. The processing unit 1004 is also used to trigger the sending of a second MAC table entry to the forwarding plane of the first NVE device according to the public VTEP address included in the next hop field in the MP_REACH_NLRI and the second VTEP address included in the MAC / IPAdvertisement route, the outgoing interface in the second MAC table entry being the local interface of the first NVE device of the VXLAN tunnel connecting the first NVE device to the second NVE device, wherein the second MAC table entry includes the MAC address carried by the MAC / IPAdvertisement route.
[0186] Optionally, the CE device is connected to the second NVE device. The receiving unit 1002 is also used to receive a MAC / IP Advertisement route from the second NVE device, the MAC / IP Advertisement route is carried in the MP_REACH_NLRI, the next hop domain in the MP_REACH_NLRI includes a public VTEP address, the MAC / IP Advertisement route includes a MAC-VLAN ID and the second VTEP address, wherein the MAC-VLAN ID is used to indicate the VLAN to which the MAC address carried by the MAC / IP Advertisement route belongs, the MAC address carried by the MAC / IP Advertisement route is the MAC address of the CE device, or the MAC address carried by the MAC / IP Advertisement route is the MAC address of a host in the EVPN site governed by the CE device; the processing unit 1004 is also used to determine that there is no local interface in the first NVE device that connects the Ethernet link between the CE device and the first NVE device according to the ESI carried by the MAC / IP Advertisement route and the MAC-VLANID. The ESI carried by the MAC / IP Advertisement route is used to indicate the ES between the CE device and the second NVE device. The processing unit 1004 is also used to send a MAC table entry to the forwarding plane of the first NVE device according to the public VTEP address included in the next hop domain in the MP_REACH_NLRI and the second VTEP address included in the MAC / IP Advertisement route. The outbound interface of the MAC table entry is the local interface of the first NVE device that connects the VXLAN tunnel from the first NVE device to the second NVE device. The MAC table entry includes the MAC address carried by the MAC / IP Advertisement route.
[0187] Fig.11 The first NVE device shown can execute the corresponding steps executed by the first NVE device in the method of the above embodiment. Ensuring that the application scenario of EVPN is not limited to the constraint that the link between NVE devices must adopt a physical direct link helps to expand the application scenario of EVPN. In addition, it helps to improve reliability and reduce deployment complexity. It should be understood that the Fig.11 The same structure applies to Figure 1 The second NVE device and the third NVE device in.
[0188] Fig.12Schematic diagram of the hardware structure of the first NVE device 1100 according to an embodiment of the present application. Fig.12 The first NVE device 1100 shown can execute the corresponding steps executed by the first NVE device in the method of the above embodiment.
[0189] like Fig.12 As shown, the first NVE device 1100 includes a processor 1101, a memory 1102, an interface 1103 and a bus 1104. The interface 1103 can be implemented wirelessly or wired, and can be a network card. The processor 1101, the memory 1102 and the interface 1103 are connected via the bus 1104.
[0190] The interface 1103 may specifically include a transmitter and a receiver, which are used to send and receive information between the first NVE device and the second NVE device and the third NVE device in the above embodiment; or to send and receive information between the first NVE device and the PE device; or to send and receive information between the first NVE device and the CE device connected to the first NVE device. As an example, the interface 1103 is used to support Figure 2 Processes S102 and S105 in the above embodiment. The processor 1101 is used to execute the processing performed by the first NVE device in the above embodiment. For example, the processor 1101 is used to compare whether the public VTEP address included in the OriginatingRouter's IP Address field in the received IMET route is the same as the public VTEP address stored in the memory 1102; it is also used to establish a VXLAN tunnel according to the VTEP address in the received IMET route based on the fact that the public VTEP address included in the OriginatingRouter's IP Address field in the received IMET route is the same as the public VTEP address stored in the memory 1102; it is also used to process received ES routes, MAC / IP Advertisement routes, or Ethernet A-D per ES routes; it is also used to process BUM traffic or unicast traffic; and / or other processes for the technology described in this document. As an example, the processor 1101 is used to support Figure 2The memory 1102 includes an operating system 11021 and an application 11022, which are used to store programs, codes or instructions. When the processor or hardware device executes these programs, codes or instructions, the processing process of the first NVE device in the method embodiment can be completed. Optionally, the memory 1102 may include a read-only memory (ROM) and a random access memory (RAM). The ROM includes a basic input / output system (BIOS) or an embedded system; the RAM includes an application and an operating system. When the first NVE device 1100 needs to be run, it is started by the BIOS solidified in the ROM or the bootloader boot system in the embedded system to guide the first NVE device 1100 into a normal operating state. After the first NVE device 1100 enters a normal operating state, the application and operating system in the RAM are run, thereby completing the processing process of the first NVE device in the method embodiment.
[0191] Understandably, Fig.12 Only a simplified design of the first NVE device is shown. In practical applications, the first NVE device may include any number of interfaces, processors or memories. In addition, this embodiment is only described by taking the first NVE device as an example. It should be understood that the second NVE device, the third NVE device, or more NVE devices have the same functions as the first NVE device, and they will not be described one by one here.
[0192] Fig.13 Schematic diagram of the hardware structure of another first NVE device 1200 according to an embodiment of the present application. Fig.13 The first NVE device 1200 shown can execute the corresponding steps executed by the first NVE device in the method of the above embodiment.
[0193] like Fig.13The first NVE device 1200 includes: a main control board 1210, an interface board 1230, a switching network board 1220 and an interface board 1240. The main control board 1210 is used to complete functions such as system management, equipment maintenance, and protocol processing. The switching network board 1220 is used to complete data exchange between various interface boards (interface boards are also called line cards or service boards). The interface boards 1230 and 1240 are used to provide various service interfaces (for example, POS interfaces, GE interfaces, ATM interfaces, etc.) and realize data packet forwarding. The main control board 1210, the interface boards 1230 and 1240, and the switching network board 1220 are connected to the system backplane through a system bus to achieve intercommunication. The central processor 1231 on the interface board 1230 is used to control and manage the interface board and communicate with the central processor on the main control board.
[0194] The physical interface card 1233 on the interface board 1230 receives the second IMET route from the second NVE device, and sends the second IMET route to the central processor 1211 on the main control board 1210 via the central processor 1231 on the interface board 1230 .
[0195] The central processor 1211 on the main control board 1210 is used to obtain the second IMET route. The central processor 1211 is also used to determine whether the public VTEP address included in the Originating Router's IP Address field in the second IMET route is the same as the public VTEP address saved by the first NVE device. When the central processor 1211 determines that the public VTEP address included in the Originating Router's IP Address field in the second IMET route is the same as the public VTEP address saved by the first NVE device, the central processor 1211 establishes a virtual extended local area network VXLAN tunnel from the first NVE device to the second NVE device according to the second VTEP address in the second IMET route. The public VTEP address included in the Originating Router's IP Address field in the second IMET route is different from the second VTEP address.
[0196] The central processor 1211 on the main control board 1210 is also used to generate a first IMET route. The Originating Router's IP Address field of the first IMET route carries the public VTEP address saved by the first NVE device. Among them, the first IMET route also includes the first VTEP address. The first VTEP address is used by the second NVE device to establish a VXLAN tunnel from the second NVE device to the first NVE device. The public VTEP address saved by the first NVE device is different from the first VTEP address. The central processor 1211 on the main control board 1210 sends the generated first IMET route to the physical interface card 1233 via the central processor 1231 on the interface board 1230. The physical interface card 1233 on the interface board 1230 sends the first IMET route to the BGP peer of the first NVE device.
[0197] The central processor 1211 on the main control board 1210 is further configured to obtain ES routes from the second NVE device and the third NVE device from the physical interface card 1233 on the interface board 1230. The central processor 1211 is further configured to determine the first NVE device as a DF or NDF in the ES according to the ES routes.
[0198] The central processor 1211 on the main control board 1210 is also used to control the interface board 1230 to complete the forwarding and processing of BUM traffic or unicast traffic.
[0199] The central processor 1211 on the main control board 1210 is further configured to generate an Ethernet AD per ES route, and send the Ethernet A-D per ES route to the BGP peer of the first NVE device through the physical interface card 1233 on the interface board 1230 .
[0200] The central processor 1211 on the main control board 1210 is further used to obtain the MAC / IP Advertisement route from the second NVE device from the physical interface card 1233 on the interface board 1230, and generate a MAC table entry according to the ESI and the MAC-VLAN ID carried by the MAC / IP Advertisement route. The central processor 1211 on the main control board 1210 transmits the MAC table entry to the forwarding table entry memory 1234 on the interface board 1230 via the central processor 1231 on the interface board 1230.
[0201] The forwarding entry memory 1234 on the interface board 1230 is used to store the MAC entry. The central processor 1231 on the interface board 1230 is used to control the network memory 1232 to obtain the MAC entry in the forwarding entry memory 1234, and the central processor 1231 is used to control the network memory 1232 to complete the reception and transmission of traffic via the physical interface card 1233.
[0202] It should be understood that the operation on the interface board 1240 in the embodiment of the present invention is consistent with the operation of the interface board 1230, and for the sake of brevity, it will not be repeated. It should be understood that the first NVE device 1200 of this embodiment can correspond to the functions and / or various steps implemented in the above method embodiment. For the sake of brevity, it will not be repeated here. In addition, this embodiment is only described by taking the first NVE device as an example. It should be understood that the second NVE device, the third NVE device, or more NVE devices have the same functions as the first NVE device, and they will not be repeated here one by one.
[0203] In addition, it should be noted that there may be one or more main control boards, and when there are multiple boards, they may include a primary main control board and a backup main control board. There may be one or more interface boards. The stronger the data processing capability of the first NVE device, the more interface boards are provided. There may also be one or more physical interface cards on the interface board. There may be no switching network board, or there may be one or more switching network boards. When there are multiple switching network boards, load sharing and redundant backup can be achieved together. Under the centralized forwarding architecture, the first NVE device may not need a switching network board, and the interface board is responsible for the processing function of the business data of the entire system. Under the distributed forwarding architecture, the first NVE device may have at least one switching network board, and the data exchange between multiple interface boards is realized through the switching network board, providing large-capacity data exchange and processing capabilities. Therefore, the data access and processing capabilities of the first NVE device of the distributed architecture are greater than those of the device of the centralized architecture. Which architecture to adopt depends on the specific networking deployment scenario, and no limitation is made here.
[0204] Fig.14 Schematic diagram of the hardware structure of yet another first NVE device 1300 according to an embodiment of the present application. Fig.14 The first NVE device 1300 shown can execute the corresponding steps executed by the first NVE device in the method of the above embodiment.
[0205] This product form of the first NVE device 1300 is suitable for a network architecture based on separation of control and forwarding (for example, a software defined network (English: Software Defined Network, abbreviated: SDN)). In SDN, Fig.13 The main control board 1210 of the first NVE device 1200 is separated from the device to form a new independent physical device (ie, Fig.14 The controller 1210A shown in FIG. 12A ) and the rest form another independent physical device (i.e., Fig.14 The controller 1210A interacts with the first NVE forwarding device 1200A through a control channel protocol. The control channel protocol may be an OpenFlow protocol, a Path Computation Element Communication Protocol (PCEP), BGP, an Interface to the Routing System (I2RS), etc. Fig.13 Compared with the corresponding embodiment, the first NVE device 1300 in this implementation includes a separated controller 1210A and a first NVE forwarding device 1200A, that is, in this embodiment, the first NVE device 1300 can also be regarded as a system.
[0206] The controller 1210A can be implemented based on a general physical server or a dedicated hardware structure. In a design example, the controller includes a receiver, a processor, a transmitter, a RAM, a ROM, and a bus (not shown in the figure). The processor is coupled to the receiver, the transmitter, the RAM, and the ROM through the bus. When the controller needs to be run, it is started by the BIOS solidified in the ROM or the bootloader boot system in the embedded system to guide the controller into a normal operating state. After the controller enters the normal operating state, the application program and the operating system are run in the RAM, so that the processor executes the above Fig.13 All functions and steps of the main control board 1210.
[0207] The first NVE forwarding device 1200A may be implemented based on a dedicated hardware structure, and its function and structure are similar to those described above. Fig.13The functions and structures of the interface board 1230, the interface board 1240 and the switching network board 1220 are consistent, and the corresponding functions and steps are performed. It can also be a virtual first NVE forwarding device implemented based on a general physical server and network function virtualization (English: Netwrk FunctionVirtulization, abbreviation: NFV) technology, and the virtual first NVE forwarding device is a virtual router. In the scenario of the virtual first NVE forwarding device, the first NVE forwarding device mentioned in the above-mentioned entity first NVE forwarding device embodiment includes an interface board, a switching network board and a processor. In a virtual environment, it can be considered that the general physical server on which it is based allocates the interface resources, network resources and processing resources used by the virtual first NVE forwarding device. The use of a general physical server to implement the functions or steps of the first forwarding NVE device, or the use of a general physical server and the use of NFV technology to implement the functions or steps of the first forwarding NVE device can be specifically referred to. Fig.12 Embodiment of the invention.
[0208] It should be understood that the controller 1210A in the first NVE device 1300 and the first NVE forwarding device 1200A in this embodiment can implement various functions and steps implemented by the first NVE device in the method embodiment, and for the sake of brevity, they will not be repeated here. In addition, this embodiment is only described by taking the first NVE device as an example, and it should be understood that the second NVE device, the third NVE device, or more NVE devices have the same functions as the first NVE device, and they will not be repeated here one by one.
[0209] In addition, an embodiment of the present application provides a computer storage medium for storing computer software instructions used for the above-mentioned first NVE device, which includes a program designed for executing the above-mentioned method embodiment.
[0210] like Figure 1 As shown, the embodiment of the present application also includes a network system for processing routing, the network system includes at least two network virtual edge NVE devices, each of the at least two NVE devices is the aforementioned Fig.11 or Fig.12 or Fig.13 or Fig.14 The first NVE device in the market.
[0211] The steps of the method or algorithm described in conjunction with the disclosure of the present application can be implemented in a hardware manner, or can be implemented by a processor executing software instructions. The software instructions can be composed of corresponding software modules, and the software modules can be stored in a RAM memory, a flash memory, a ROM memory, an EPROM memory, an EEPROM memory, a register, a hard disk, a mobile hard disk, a CD-ROM or any other form of storage medium well known in the art. An exemplary storage medium is coupled to the processor so that the processor can read information from the storage medium and can write information to the storage medium. Of course, the storage medium can also be a component of the processor. The processor and the storage medium can be located in an ASIC. In addition, the ASIC can be located in a user device. Of course, the processor and the storage medium can also be present in the user device as discrete components.
[0212] Those skilled in the art should be aware that in one or more of the above examples, the functions described in this application can be implemented using hardware, software, firmware, or any combination thereof. When implemented using software, these functions can be stored in a computer-readable medium or transmitted as one or more instructions or codes on a computer-readable medium. Computer-readable media include computer storage media and communication media, wherein communication media include any media that facilitates the transmission of a computer program from one place to another. The storage medium can be any available medium that a general or special-purpose computer can access.
[0213] The specific implementations described above further describe the purpose, technical solutions and beneficial effects of the present application in detail. It should be understood that the above description is only a specific implementation of the present application.
Claims
1. A method for processing routing, the method being applied to a virtual extended local area network VXLAN, the VXLAN comprising a first network device and a second network device, characterized in that: The method comprises: The first network device receives a second update message from the second network device, the second update message includes second multi-protocol reachable network layer reachability information MP_REACH_NLRI, the second MP_REACH_NLRI includes a public virtual extended local area network tunnel endpoint VTEP address stored by the second network device, wherein the second update message also includes a second VTEP address; In response to the public VTEP address in the second MP_REACH_NLRI being the same as the public VTEP address stored by the first network device, the first network device establishes a VXLAN tunnel from the first network device to the second network device according to the second VTEP address in the second update message.
2. The method according to claim 1, characterized in that The method further comprises: The first network device sends a first update message to the second network device, the first update message includes a first MP_REACH_NLRI, the first MP_REACH_NLRI includes a public VTEP address saved by the first network device, wherein the first update message also includes a first VTEP address, and the first VTEP address is used by the second network device to establish a VXLAN tunnel from the second network device to the first network device.
3. According to the method of claim 2, the VXLAN further includes a provider edge PE device, characterized in that: The method further comprises: The first network device sends the first update message to the PE device; The first network device receives an update message from the PE device, where the update message from the PE device includes a VTEP address of the PE device; The first network device establishes a VXLAN tunnel from the first network device to the PE device according to the VTEP address of the PE device in the update message from the PE device.
4. The method according to claim 2 or 3, characterized in that: The VXLAN attribute in the first update message carries the first VTEP address; The VXLAN attribute in the second update message carries the second VTEP address.
5. The method according to any one of claims 1 to 3, characterized in that: The VXLAN further includes a third network device, and the method further includes: The first network device receives a third update message from the third network device, the third update message includes a third MP_REACH_NLRI, the third MP_REACH_NLRI includes a public VTEP address, wherein the third update message also includes a third VTEP address; In response to the public VTEP address included in the third MP_REACH_NLRI being the same as the public VTEP address stored by the first network device, the first network device establishes a VXLAN tunnel from the first network device to the third network device according to the third VTEP address in the third update message.
6. The method according to claim 5, characterized in that A user edge CE device is multi-homed to the first network device, the second network device and the third network device via multiple Ethernet links, and the multiple Ethernet links constitute an Ethernet segment ES; The method further comprises: The first network device determines the first network device as a designated forwarder DF in the ES according to the ES routes from the second network device and the third network device.
7. The method according to claim 5, characterized in that The method further comprises: The first network device receives broadcast, unknown unicast, and multicast BUM traffic sent by the second network device via a VXLAN tunnel from the second network device to the first network device; The first network device avoids forwarding the BUM traffic to the third network device via the VXLAN tunnel from the first network device to the third network device based on the fact that the BUM traffic is received by the first network device via the VXLAN tunnel from the second network device to the first network device.
8. The method according to claim 2 or 3, characterized in that: The CE device is dual-homed to the first network device and the second network device via multiple Ethernet links, and the multiple Ethernet links constitute an ES; The method further comprises: The first network device determines the first network device as a DF in the ES according to the ES route from the second network device; The first network device sends an Ethernet AD per ES route to the second network device, where the Ethernet AD per ES route carries an Ethernet segment identifier ESI label allocated by the first network device to the Ethernet link between the CE device and the first network device; The first network device receives a VXLAN message sent by the second network device via a VXLAN tunnel from the second network device to the first network device, wherein the original Ethernet payload in the VXLAN message includes BUM traffic, and the VXLAN message carries the ESI label, wherein the BUM traffic comes from the CE device; The first network device carries the ESI label based on the VXLAN message to avoid forwarding the BUM traffic to the CE device via the ES between the CE device and the first network device.
9. The method according to claim 8, characterized in that The ESI label is encapsulated between the VXLAN header and the original Ethernet payload in the VXLAN message, and the VXLAN header carries indication information, where the indication information is used to indicate that the VXLAN message carries the ESI label.
10. The method according to any one of claims 1 to 3, characterized in that: The CE device is dual-homed to the first network device and the second network device via multiple Ethernet links, and the multiple Ethernet links constitute an ES; The method further comprises: The first network device receives a media access control / Internet protocol announcement MAC / IP Advertisement route from the second network device, the MAC / IP Advertisement route includes a media access control-virtual local area network identifier MAC-VLAN ID, wherein the MAC-VLAN ID is used to indicate the VLAN to which the MAC address carried by the MAC / IP Advertisement route belongs, the MAC address carried by the MAC / IP Advertisement route is the MAC address of the CE device, or the MAC address carried by the MAC / IP Advertisement route is the MAC address of a host in an Ethernet virtual private network site EVPN site governed by the CE device; The first network device determines, according to the ESI carried by the MAC / IP Advertisement route and the MAC-VLANID, a local interface of the first network connecting the Ethernet link between the CE device and the first network device, wherein the ESI carried by the MAC / IP Advertisement route is used to indicate the ES between the CE device and the second network device; The control plane of the first network device sends a first MAC table entry to the forwarding plane of the first network device, wherein the outgoing interface in the first MAC table entry is the local interface of the first network device, wherein the first MAC table entry includes the MAC address carried by the MAC / IP Advertisement route and the MAC-VLANID.
11. According to the method according to any one of claims 1-3, both the first network device and the second network device are network virtual edges NVE.
12. A first network device, characterized in that: The first network device comprises: A receiving unit, configured to receive a second update message from a second network device, wherein the second update message includes second multi-protocol reachable network layer reachability information MP_REACH_NLRI, wherein the second MP_REACH_NLRI includes a public virtual extended local area network tunnel endpoint VTEP address stored by the second network device, wherein the second update message also includes a second VTEP address; In response to the public VTEP address in the second MP_REACH_NLRI being the same as the public VTEP address stored in the first network device, the processing unit is used to establish a virtual extended local area network VXLAN tunnel from the first network device to the second network device according to the second VTEP address in the update message.
13. The first network device according to claim 12, characterized in that: The first network device also includes: A sending unit, used to send a first update message to the second network device, the first update message includes a first MP_REACH_NLRI, the first MP_REACH_NLRI includes a public VTEP address stored by the first network device, wherein the first update message also includes a first VTEP address, and the first VTEP address is used by the second network device to establish a VXLAN tunnel from the second network device to the first network device.
14. The first network device according to claim 13, characterized in that: The sending unit is further configured to send the first update message to the operator edge PE device; The receiving unit is further configured to receive an update message from the PE device, wherein the update message from the PE device includes a VTEP address of the PE device; The processing unit is further configured to establish a VXLAN tunnel from the first network device to the PE device according to the VTEP address of the PE device in the update message from the PE device.
15. The first network device according to claim 13 or 14, characterized in that: The VXLAN attribute in the first update message carries the first VTEP address; The VXLAN attribute in the second update message carries the second VTEP address.
16. The first network device according to any one of claims 12 to 14, characterized in that: The receiving unit is further configured to receive a third update message from a third network device, the third update message including a third MP_REACH_NLRI, the third MP_REACH_NLRI including a public VTEP address, wherein the third update message also includes a third VTEP address; In response to the public VTEP address included in the third MP_REACH_NLRI being the same as the public VTEP address stored in the first network device, the processing unit is further configured to establish a VXLAN tunnel from the first network device to the third network device according to the third VTEP address in the third update message.
17. The first network device according to claim 16, characterized in that: A user edge CE device is multi-homed to the first network device, the second network device and the third network device via multiple Ethernet links, and the multiple Ethernet links constitute an Ethernet segment ES; The processing unit is further configured to determine the first network device as a designated forwarder DF in the ES according to the ES routes from the second network device and the third network device.
18. The first network device according to claim 16, characterized in that: The receiving unit is further configured to receive broadcast, unknown unicast, and multicast BUM traffic sent by the second network device via a VXLAN tunnel from the second network device to the first network device; The processing unit is also used to avoid forwarding the BUM traffic to the third network device via the VXLAN tunnel from the first network device to the third network device based on the fact that the BUM traffic is received by the first network device via the VXLAN tunnel from the second network device to the first network device.
19. The first network device according to claim 13 or 14, characterized in that: The CE device is dual-homed to the first network device and the second network device via multiple Ethernet links, and the multiple Ethernet links constitute an ES; The processing unit is further configured to determine the first network device as a DF in the ES according to the ES route from the second network device; The sending unit is further used to send an Ethernet AD per ES route to the second network device, where the Ethernet AD per ES route carries an Ethernet segment identifier ESI label allocated by the first network device to the Ethernet link between the CE device and the first network device; The receiving unit is further configured to receive a VXLAN message sent by the second network device via a VXLAN tunnel from the second network device to the first network device, wherein the original Ethernet payload in the VXLAN message includes BUM traffic, and the VXLAN message carries the ESI label, wherein the BUM traffic comes from the CE device; The processing unit is also used to avoid forwarding the BUM traffic to the CE device via the ES between the CE device and the first network device based on the VXLAN message carrying the ESI label.
20. The first network device according to claim 19, characterized in that: The ESI label is encapsulated between the VXLAN header and the original Ethernet payload in the VXLAN message, and the VXLAN header carries indication information, where the indication information is used to indicate that the VXLAN message carries the ESI label.
21. The first network device according to any one of claims 12 to 14, characterized in that: The CE device is dual-homed to the first network device and the second network device via multiple Ethernet links, and the multiple Ethernet links constitute an ES; The receiving unit is further used to receive a media access control / Internet protocol announcement MAC / IP Advertisement route from the second network device, the MAC / IP Advertisement route includes a media access control-virtual local area network identifier MAC-VLAN ID, wherein the MAC-VLAN ID is used to indicate the VLAN to which the MAC address carried by the MAC / IP Advertisement route belongs, the MAC address carried by the MAC / IP Advertisement route is the MAC address of the CE device, or the MAC address carried by the MAC / IP Advertisement route is the MAC address of a host in an Ethernet virtual private network site EVPN site governed by the CE device; The processing unit is further configured to determine, according to the ESI carried by the MAC / IP Advertisement route and the MAC-VLANID, a local interface of the first network connecting the Ethernet link between the CE device and the first network device, wherein the ESI carried by the MAC / IP Advertisement route is used to indicate the ES between the CE device and the second network device; The processing unit is also used to trigger sending a first MAC table entry to the forwarding plane of the first network device, wherein the outgoing interface in the first MAC table entry is the local interface of the first network device, wherein the first MAC table entry includes the MAC address carried by the MAC / IP Advertisement route and the MAC-VLAN ID.
22. The first network device according to any one of claims 12 to 14, wherein both the first network device and the second network device are network virtual edges (NVEs).
23. A network system for processing routing, the network system comprising at least two network devices, each of the at least two network devices being the first network device as described in any one of claims 12 to 22.
