WO2024001553A1

WO2024001553A1 - Routing publishing method, electronic device and computer-readable storage medium

Info

Publication number: WO2024001553A1
Application number: PCT/CN2023/093828
Authority: WO
Inventors: 荣欣; 舒晔
Original assignee: 中兴通讯股份有限公司
Priority date: 2022-06-30
Filing date: 2023-05-12
Publication date: 2024-01-04
Also published as: CN117377020A

Abstract

The present disclosure provides a routing publishing method, an electronic device and a computer-readable storage medium. The routing publishing method is executed by a first provider edge (PE) device, and comprises: receiving a first routing message from a second PE device, wherein the first routing message is used for advertising a MAC/IP advertisement route, the first routing message comprises a first label field and a second label field, the first label field comprises a first label, and the second label field is vacant; filling the second label field with a second label so as to generate a second routing message, wherein the first label is used for indicating layer 2 service traffic forwarding, and the second label is used for indicating layer 3 service traffic forwarding; and sending the second routing message to a third PE device.

Description

Route publishing method, electronic device and computer-readable storage medium

Cross-references to related applications

This application claims priority from Chinese patent application No. 202210760692.4 submitted on June 30, 2022. The content of this Chinese patent application is incorporated herein by reference in its entirety.

Technical field

The present disclosure relates to the field of communications, and in particular, to a route publishing method, electronic device and computer-readable storage medium.

Background technique

In IP-based wireless access networks (IP Radio Access Network, IPRAN), in order to reduce the routing pressure on access-side devices, the access ring is generally deployed using Layer 2 Virtual Private Network (L2VPN), aggregation The core ring is deployed using Layer 3 Virtual Private Network (L3VPN). In order to achieve end-to-end service interoperability, the Layer 2 and Layer 3 bridging technology is used to forward L2VPN messages and L3VPN messages on the intermediate aggregation provider edge (PE) device.

The principle of Layer 2 and Layer 3 bridging is to configure a Layer 2 and Layer 3 bridge interface on the bridge device. This type of interface can be used on both Layer 2 and Layer 3. With the development of 5G networks, traditional Virtual Private Network (VPN) bridging is gradually replaced by Ethernet Virtual Private Network (EVPN) bridging. The types of bridging technologies include traditional L2VPN and L3VPN bridging. There is also the bridging of the new network's Layer 2 Ethernet Virtual Private Network (L2EVPN) and Layer 3 Ethernet Virtual Private Network (L3EVPN).

In a VPN Layer 2 and Layer 3 bridged network, how to implement Layer 2 when the Layer 2 network fails? Layer 3 linkage, which switches services to backup paths in a timely manner to reduce service damage, is an important link that needs to be considered in the 2nd and 3rd bridging scenarios.

For example, in related technologies, as shown in Figure 3, PE1, PE2, and PE3 deploy L2EVPN, and PE2, PE3, and PE4 deploy L3EVPN to achieve traffic interworking between CE1 and CE2 as customer edge devices (Customer Edge, CE). , configure the Integrated Routing and Bridging (IRB) gateway on the Layer 2 and Layer 3 bridging devices PE2 and PE3 to implement Layer 2 and Layer 3 bridge forwarding.

The IRB interfaces configured on bridge points PE2 and PE3 will each generate a MAC/IP advertised route (RT-2), which will be published to PE1 through EVPN. After PE1 receives it, it will write it to the local Virtual Private LAN Service. , VPLS), the next hop is PE2, and a private network label L2 is assigned. For packets sent from CE1 to CE2, after the packet is forwarded to PE1, the MAC table is checked for routing and forwarding based on the destination MAC of the packet (the MAC address of the IRB interface). The next hop is the PE2 device. When the packet reaches the bridge point, it finds the IRB interface according to the destination MAC, searches for the destination address route under the virtual routing forwarding (VRF) where the IRB interface is located, and performs Layer 3 forwarding.

The traffic sent from CE1 to CE2 is called uplink traffic, and the traffic sent from CE2 to CE1 is called downlink traffic. In live network deployment, redundant backup protection is generally used. The IRB interfaces on PE2 and PE3 serve as dual gateways and are configured with the same IP address and the same MAC. The IRB interface and the interface int1 on the CE1 device are in the same network segment. When a link or node fails, it can quickly switch to the backup path to ensure uninterrupted services. As shown in Figure 3, PE3 is the backup device of PE2. When the PE2 device fails or the main link (such as PE1-PE2-PE4) fails, the service can be switched to the backup path (such as PE1-PE3-PE4) to ensure the business No interruption.

Contents of the invention

In a first aspect, the present disclosure provides a route publishing method, executed by a first provider edge PE device, including: receiving a first routing message from a second PE device; the first routing message is used to advertise MAC/IP advertisements Routing, the first routing message includes a first label field and a second label field, the first label field includes a first label, and the second label field is blank; filling the second label field with a second label, To generate a second routing message, the first label is used to represent Layer 2 service traffic forwarding, and the second label is used to means forwarding Layer 3 service traffic; sending the second routing message to the third PE device.

In a second aspect, the present disclosure provides an electronic device. The electronic device includes: one or more processors; a memory on which one or more computer programs are stored. When the one or more computer programs are The one or more processors execute, so that the one or more processors implement the route publishing method according to the first aspect; one or more I/O interfaces are connected between the processor and the memory, Configured to implement information interaction between the processor and the memory.

In a third aspect, the present disclosure provides a computer-readable storage medium. A computer program is stored on the computer-readable storage medium. When the computer program is executed by a processor, the route publishing method according to the first aspect is implemented. .

Description of drawings

Figure 1 is a flow chart of a route publishing method provided by an embodiment of the present disclosure.

Figure 2 is a flow chart of some steps in a route publishing method provided by an embodiment of the present disclosure.

Figure 3 is a schematic diagram of a typical EVPN layer 2 and layer 3 bridge topology.

Figure 4 is a topology diagram of fault scenarios in a traditional VPN Layer 2 and Layer 3 bridge network.

Figure 5 is a topology diagram of fault scenarios in the traditional EVPN Layer 2 and Layer 3 bridge network.

Figure 6 is a schematic diagram of the message format of EVPN MAC/IP advertised route (RT-2).

Figure 7 is a schematic diagram of a fault scenario in Embodiment 1 of the present disclosure.

Figure 8 is a schematic diagram of a fault scenario in Embodiment 2 of the present disclosure.

Figure 9 is a schematic diagram of a fault scenario in Embodiment 3 of the present disclosure.

FIG. 10 is a schematic diagram of an electronic device provided by an embodiment of the present disclosure.

FIG. 11 is a schematic diagram of a computer-readable storage medium provided by an embodiment of the present disclosure.

Detailed ways

It should be understood that the specific embodiments described here are only used to explain the present disclosure and are not used to limit the present disclosure.

In the following description, terms such as "module" and "component" used to represent elements are used. The suffix "unit" or "unit" is only used to facilitate the description of the present disclosure and has no special meaning in itself. Therefore, "module", "component" or "unit" may be used interchangeably.

In a traditional VPN Layer 2 and Layer 3 bridge network, the Layer 3 bridge interface is associated with the status of L2VPN Bidirectional Forwarding Detection (BFD) or public network tunnel BFD to perform linkage from Layer 2 to Layer 3. As shown in Figure 4, the link between the bridge point PE2 and the access PE1 (the link corresponding to int2 and int7) fails. L2VPN deploys virtual link (Pseudo Wire, PW) BFD or public network BFD, L2VPN VPLS PW There is a problem with BFD or the BFD status of the public network tunnel (down), which triggers the shutdown of the Layer 3 bridge interface (down). The network segment route of the bridge interface is published and revoked through L3VPN. L3VPN on PE4 performs route switching, and the next hop is switched to PE3. This directs downlink traffic to PE3 to achieve layer 2 and layer 3 linkage.

With the development of 5G networks, traditional VPN bridging is gradually replaced by EVPN bridging. Currently, EVPN does not have a separate BFD detection mechanism, and public network tunnel BFD uses more automatic creation technology, and no specific BFD session is tracked. . When a fault occurs between the main bridge point and the access device, layer 2 and layer 3 linkage cannot be realized well, and traffic will be interrupted or detoured.

In such a scenario where manual tracking of associations is not possible, host routing is needed to achieve intelligent linkage. In the EVPN distributed gateway scenario, EVPN IRB can automatically generate host routes to achieve linkage. However, when the network adopts centralized gateway deployment, the EVPN gateway only has the host routes of the gateway itself and cannot generate host routes for connecting to the CE. Therefore, it is impossible to achieve integration with the CE. Intelligent linkage of host routing. If manual BFD binding is used, it will not be easy to change dynamically with business changes, and it will increase the configuration complexity of the device and consume a large amount of BFD resources on the device. At the same time, a large number of BFD detection packets will also increase the pressure on the network.

In view of the above problems in the centralized gateway deployment scenario, the inventor found a solution to solve the problem of layer 2 and layer bridging linkage through host routing linkage. This disclosure carries both the first label field and the second label field for the RT-2 route on the bridge point PE, triggering the three-layer PE to generate a 32-bit host route, so that the three-layer PE forwards according to the 32-bit host route. When the two-layer network When a network element or link fails and the RT-2 route associated with the main path fails, the linked Layer 3 PE deletes the 32-bit host route corresponding to the main path and switches to the backup path, thereby realizing Layer 2 and Layer 3 bridge linkage. .

In view of this, as a first aspect of the present disclosure, the present disclosure provides a route publishing method, which is executed by the first provider edge PE device. As shown in Figure 1, the route publishing method includes the following steps S100 to S300.

In step S100, a first routing message from the second PE device is received; the first routing message is used to advertise a MAC/IP Advertisement Route (RT-2), and the first routing message includes a A label field and a second label field, the first label field includes a first label, and the second label field is blank.

In step S200, fill the second label field with a second label to generate a second routing message. The first label is used to represent Layer 2 business traffic forwarding, and the second label is used to represent Layer 3 business traffic. Forward.

In step S300, the second routing message is sent to the third PE device.

As shown in Figure 5, in the traditional implementation method, the IRB interface of CE1 and the bridge PE are in the same network segment. The bridge PE will generate an EVPN IP prefix route from the IRB interface network segment (such as a 24-bit network segment address) and send it to Layer 3 PE4, but the 32-bit host address of the CE1 interface will not be generated on the bridge PE, so the 32-bit host route of the CE1 interface cannot be advertised to PE4.

In the embodiment of the present disclosure, Address Resolution Protocol (ARP) proxy is configured on PE1 on the access side. When CE1 issues an ARP request, access PE1 will generate EVPN RT-2 MAC and EVPN RT- 2 single-label routes and published to EVPN. The bridging PE (PE2 and PE3) devices receive the EVPN RT-2 route. The IP address of the route and the IRB interface are in the same network segment, forming an ARP entry to CE1, and the outgoing interface is the IRB. The message format of the EVPN RT-2 route is shown in Figure 6. According to the L3VPN instance to which the IRB belongs, the EVPN RT-2 route is imported into the routing table of each L3VPN instance. The L3VPN instance allocates a Layer 3 label to the route and adds the The label is filled in the second label field label2 of Multi-Protocol Label Switching (MPLS) in the routing packet. On this basis, add the outbound routing target (export RT) of the local VRF, and publish the route through EVPN. After PE4 receives the RT-2 route, it compares the export RT with the local inbound routing target (import RT). ) cross-matching, import the route into the local L3VPN instance, and generate a 32-bit host route corresponding to CE1 based on the host IP address of CE1 carried in the RT-2 route.

When searching for routes when forwarding traffic, 32-bit host routing is better than 24-bit network segment routing. When the link between PE1 and PE2 and the link between PE2 and PE3 fail, the EVPN RT-2 routing entry sent by CE1 to PE4 through PE1 and PE2 is revoked, and the EVPN RT sent by CE1 to PE4 through PE1 and PE3 The -2 routing entry is still there, so the next hop of the 32-bit host route to the CE1 int1 address is switched to PE3, thereby realizing traffic switching.

In the technical solution provided by this disclosure, in the scenario of centralized gateway deployment, EVPN RT-2 routes are converted from single labels to dual labels, and are published to EVPN, triggering the generation of 32-bit host routes on the PEs that receive the routes. To realize intelligent linkage in EVPN layer 2 and layer bridging scenarios.

In some implementations, as shown in Figure 2, step S200 can be specifically decomposed into the following steps S210 and S220.

In step S210, determine the L3VPN private network label corresponding to the first routing message based on the first routing message.

In step S220, the L3VPN private network label is filled into the second label field of the first routing message to generate a second routing message.

The first routing message will carry export RT, and the corresponding L3VPN instance can be determined based on the local import RT of the first PE. The first routing message will carry the IP address of the CE1 interface, which can determine the network segment corresponding to the IRB interface on the first PE. Therefore, the private network label assigned by the L3VPN instance to this network segment can be determined. When the first PE forwards the RT-2 route to the third PE, it fills the determined private network label value into the second label field of the RT-2 route, and then sends it to the third PE.

In some embodiments, the L3VPN instance corresponding to RT-2 is set to per instance per label, that is, each L3VPN instance shares the same private network label, and all routes in an L3VPN instance use the same label. The difference from per-route per label is that you do not need to first determine the network segment and L3VPN instance, and then confirm the private network label of the network segment route. You only need to determine the L3VPN instance to know the private network label of the L3VPN instance. Of course, the private network label of the L3VPN instance can also be set statically.

After filling in the label value in the second label field, the second route will be sent to the three-layer PE. The three-layer PE will determine whether the received RT-2 route is a double label. If it is a double label, it will use the IP in the RT-2 route. Address generation for 32-bit host routes. At this time the second label field acts as the host The role of the route generation identifier. Therefore, the L3VPN private network label value corresponding to the IRB interface network segment needs to be filled in the second label field.

In some implementations, step S210 can be specifically decomposed into the following steps S211 to S213.

In step S211, the corresponding L3VPN interface is determined according to the first routing message.

In step S212, the L3VPN instance corresponding to the L3VPN interface is determined.

In step S213, the private network label assigned by the L3VPN instance to the network segment address of the L3VPN interface is used as the L3VPN private network label corresponding to the first routing message.

In other embodiments, the first routing message carries the identifier of the virtual LAN VLAN, and step S210 can be specifically decomposed into the following steps S214 to S216.

In step S214, the corresponding L3VPN sub-interface is determined according to the identifier of the VLAN.

In step S215, the L3VPN instance corresponding to the L3VPN sub-interface is determined.

In step S216, the private network label assigned by the L3VPN instance to the network segment address of the L3VPN sub-interface is used as the L3VPN private network label corresponding to the first routing message.

The VPLS EVPN scenario is divided into two modes: qualified and unqualified. The learning of ordinary Ethernet source MAC addresses is unqualified mode; if in addition to learning the source MAC, it is also necessary to learn VLAN tags (service delimiter P-tag), it is qualified mode.

For unqualified VPLS EVPN scenarios, this disclosure proposes a method for converting EVPN RT-2 routes from single label to dual label. Different settings are required for RT-2 routes generated by different access circuit (Attachment Circuit, AC) entrances. For the L2 export RT value, configure the IRB sub-interface on the bridge device, bind it to different private network VRFs, find the corresponding IRB sub-interface by matching L2 export RT and L3 import RT, and add the corresponding IRB The L3VPN private network label to which the sub-interface belongs is filled in the label2 field of the EVPN RT-2 routing packet. This achieves the purpose of converting single label to dual label.

For the qualified VPLS EVPN scenario, this disclosure proposes a method of converting EVPN RT-2 routing from single label to dual label. Multiple CE devices are connected to PE1 through different VLAN sub-interfaces, and access to PE1 generates EVPN RT-2 single-label routes carry different VLAN IDs and are published to the bridge device. The VLAN ID value is carried in the packet to find the route. Go to the IRB sub-interface corresponding to its VLAN ID, import the route into the L3VPN instance corresponding to the IRB sub-interface, and fill in the L3VPN private network label assigned by the VRF into the label2 field in the EVPN RT-2 routing packet, through EVPN is advertised to remote PE4 and generates a 32-bit host route.

Further, the method also includes step S400.

In step S400, when the first route parsed according to the first routing message satisfies the inactivity condition, the second routing message is withdrawn from the third PE.

In order to ensure the validity of the route, when the first route status as the source is inactive, the second route message that has been sent needs to be revoked. After receiving the revocation of the second route message, the third PE will also delete the 32-bit host route generated based on the second route synchronously. Therefore, when the main path fails, the RT-2 route of the main path is revoked, and the 32-bit host route of the main path generated based on this RT-2 route is deleted. Then, the backup path generated based on the RT-2 route of the backup path is deleted. The 32-bit host route is selected as the optimal route. Therefore, business traffic will be forwarded according to the newly selected optimal route.

The inactivity condition includes at least one of the following conditions: the next hop of the first route is unreachable, the first route cannot iterate to a tunnel, and a withdrawal message of the first route is received.

When the above common reasons for the inactivity of the first route occur, the first route will become inactive, which will in turn cause the second route message to be withdrawn from the third PE.

In some embodiments, the first PE is a bridging PE device in an EVPN Layer 2 and Layer 3 bridging scenario.

Further, the second PE is a Layer 2 PE in an EVPN Layer 2 and Layer 3 bridging scenario, and the third PE is a Layer 3 PE in an EVPN Layer 2 and Layer 3 bridging scenario.

The specific applications of the route publishing method described in the first aspect of this disclosure in different scenarios are introduced below with reference to three embodiments.

Example 1

This disclosed embodiment is a basic EVPN Layer 2 and Layer 3 bridging scenario. EVPN RT-2 routes are converted from single labels to dual labels, published to the EVPN network and host routes are generated to achieve intelligent linkage in EVPN Layer 2 and Layer 3 bridging scenarios.

As shown in Figure 7, configure ARP proxy on PE1 on the access side. When CE1 issues ARP Upon request, access to PE1 will generate EVPN RT-2 MAC and EVPN RT-2 single-label routes and publish them to the EVPN network. The bridging PE (PE2 and PE3) devices receive the EVPN RT-2 route. The IP address of the route and the IRB interface are in the same network segment, forming an ARP entry to CE1, and the outgoing interface is the IRB. At the same time, according to the VRF to which the IRB interface belongs, the EVPN RT-2 route is imported into the corresponding VRF table. The VRF assigns a three-layer label to the route and fills the label into the label2 field in the routing message (EVPN RT-2 report The file format is shown in Figure 6), then add the export RT of the local VRF, and publish the route through EVPN. After PE4 receives the RT-2 route, it matches the export RT of the route with the import RT of the local VRF. The route is imported into the local VRF and a 32-bit host route corresponding to CE1 is generated.

When searching for routes when forwarding traffic, 32-bit host routing is better than 24-bit network segment routing. When the link between PE1 and PE2 and the link between PE2 and PE3 both fail, the EVPN RT-2 routing entry sent by CE1 to PE4 through PE1 and PE2 is revoked, and the EVPN RT-2 routing entry sent by CE1 to PE4 through PE1 and PE3 2 The routing entry is still there, so the next hop of the 32-bit host route to the CE1 int1 address is switched to PE3, thereby realizing traffic switching.

The specific process includes the following steps 101 to 109.

In step 101, configure the VPLS EVPN instance on PE1, PE2, and PE3. On the PE1 device, the link int1 between CE1 and PE1 serves as the AC and connects to the VPLS EVPN instance VPLS 1. PE1, PE2, and PE3 establish neighbors through Border Gateway Protocol (BGP) EVPN to deliver RT-2 routes. There is an L3EVPN network between PE2, PE3 and PE4. Neighbors are established through BGP EVPN to deliver RT-2 routes.

In step 102, CE1 int1 is configured with an IP address (such as 10.0.0.1). Configure bridge interface irb1 on bridge devices PE2 and PE3. Bridge interface irb1 and CE1 int1 are in the same network segment (such as 10.1.1.2), and PE2 and PE3 are configured with dual gateways with the same address and MAC (10.20.30.40.50.60). The irb1 interface is bound to the VPLS EVPN instance VPLS1 as a Layer 2 interface and as a Layer 3 interface. The irb1 interface is referenced in the L3VPN instance vrf1 to realize the interconnection between L2EVPN and L3EVPN.

In step 103, configure ARP proxy on access PE1. When CE1 issues an ARP request, access PE1 will generate EVPN RT-2 MAC and EVPN RT-2 single label (labelA) routes destined for the CE1 int1 interface, and Publish to the EVPN network. The bridge PE (PE2 and PE3) equipment receives the EVPN RT-2 route, and the IP address of the route is in the same network as irb1. segment, forming an ARP entry to CE1 int1, the outgoing interface is irb1. According to the VRF (vrf1) to which irb1 belongs, the EVPN RT-2 route is imported into the routing table of vrf1. vrf1 assigns a three-layer label (labelB) to the route and fills the label into the label2 field (EVPN RT) in the routing message. -2 route message format (as shown in Figure 6), then add the export RT of local vrf1, and publish the route through EVPN. After PE4 receives the route, it imports the route into the local VRF according to the export RT and generates the corresponding 32-bit host route for CE1 int1 interface address.

In step 104, in order for IP1 (20.0.0.1), an address on CE1, to communicate with IP2 (30.0.0.1) on CE2, they need to connect routes to IP1 and IP2 addresses. The following uses static routing as an example.

In step 105, open the route from CE1 to CE2 IP2, configure a private network static route to CE2 IP2 on PE4, the next hop is the int6 address of CE2, and publish the private network static route to the bridge device PE2 through BGP EVPN. On PE3. On PE2 and PE3, the next hop of the route to IP2 is PE4. Configure a static route on CE1 to CE2 IP2, with the next hop being irb1. The route from CE1 to CE2 IP2 is opened.

In step 106, on the contrary, open the route from CE2 to CE1 IP1, configure a static route on CE2 to CE1 IP1, and the next hop is the int6 interface address on PE4. Configure a private network static route to CE1 IP1 on the bridge devices PE2 and PE3. The next hop is the int1 interface address of CE1. The private network static route is published to PE4 through L3EVPN. The next hops of the route from PE4 to IP1 are PE2 and PE3. Configure a static route on CE2 to CE1 IP1, and the next hop is the int6 interface address on PE4. The route from CE2 to CE1 IP1 is opened.

In step 107, as shown in Figure 7, the PE1 device is configured with the ARP proxy function. PE2 and PE3 obtain the EVPN RT-2 route to CE1 int1, which carries a layer of labels (label 100, and fill in 100 in label 1 of the message). This label is the private network label of L2EVPN. On PE2 and PE3, import the EVPN RT-2 route into the routing table of vrf1 according to vrf1 to which irb1 belongs. vrf1 assigns a layer 3 label (label 200) to the route, and fills label 200 into the label2 field in the routing packet ( The EVPN RT-2 route message format is shown in Figure 6), then add the export RT (11.0.0.1:1) of local vrf1, and publish the EVPN RT-2 route through the EVPN network. After PE4 receives the route, it discovers export RT (11.0.0.1:1) matches import RT(11.0.0.1:1) in local vrf1. Import the route into local vrf1 and generate a 32-bit host route corresponding to CE1 int1 (10.0.0.1).

In step 108, CE2 sends the service flow to CE1, and the PE4 device searches for routes when forwarding the traffic. The 32-bit host route to CE1 int1 (10.0.0.1) is better than the 24-bit network segment route. Check the host route for forwarding. The next hop of the main path is PE2.

In step 109, when the link between PE1 and PE2 and the link between PE2 and PE3 fail, the EVPN RT-2 routing entry sent by CE1 to PE4 through PE1 and PE2 is revoked, and the EVPN RT-2 routing entry sent by CE1 to PE4 through PE1 and PE3 is revoked. The EVPN RT-2 routing entry of PE4 is still there. Therefore, on the PE4 device, the next hop of the 32-bit host route to the CE1 int1 address is switched to PE3, and the traffic is switched to the PE3-PE1-CE1 path for forwarding.

Example 2

This disclosed embodiment is an unqualified bridging scenario for L2EVPN VPLS. By matching L2 export RT and L3 import RT, the L3 private network label is inserted into the label2 field of the EVPN RT-2 route, thereby converting the route from single label to dual label. Publish the EVPN RT-2 route to the EVPN network and generate host routes on the remote PE to realize intelligent linkage of routes in EVPN Layer 2 and Layer 3 bridge scenarios.

As shown in Figure 8, multiple CEs on the access side are connected to PE1. In the remote Layer 3 network, multiple CEs are connected through different private network L3VPN instances. To achieve intercommunication between CE1 and remote CE2 through Layer 3 vrf1, CE3 and remote CE4 communicate through the Layer 3 network vrf2, and ARP proxy is configured on access PE1. When CE1 issues an ARP request, access PE1 will generate EVPN RT-2 MAC and EVPN RT-2 single-label route MAC1. When CE3 When issuing an ARP request, access PE1 generates EVPN RT-2 MAC and EVPN RT-2 single-label routing MAC2, and publishes MAC1 and MAC2 to the EVPN network. PE1 sets different L2 export RT values according to different AC access MACs, MAC1 sets export RT1, and MAC2 sets export RT2. Use IRB sub-interfaces for bridging on the bridging PE (PE2 and PE3) devices. Different sub-interfaces are bound to different VRFs. Find the corresponding L3VPN private network label by matching the import RT value of the VRF with the L2 export RT value, and Fill in the label2 field in the routing packet (the EVPN RT-2 routing packet format is shown in Figure 6), and then add the local The export RT of the VRF publishes the route through the EVPN network. After PE4 receives the route, it imports the route into the local VRF according to the export RT and generates a 32-bit host route corresponding to the CE. A 32-bit host route to CE1 int1 is formed in vrf1 on PE4, and a 32-bit host route to CE2 int11 is formed in vrf2.

When searching for routes when forwarding traffic, 32-bit host routing is better than 24-bit network segment routing. When the link between PE1 and PE2 and the link between PE2 and PE3 both fail, the EVPN RT-2 routing entries sent by CE1 and CE3 to PE4 through PE1 and PE2 are revoked, and the EVPN RT-2 routing entries sent by CE1 and CE3 to PE4 through PE1 and PE3 The EVPN RT-2 routing entry is still there. Therefore, the next hop of the 32-bit host route to the CE1int1 address is switched to PE3, and the next hop of the 32-bit host route to the CE3 int11 address is switched to PE3, thereby realizing traffic switching.

The specific process includes the following steps 201 to 209.

In step 201, configure VPLS EVPN instances on PE1, PE2, and PE3. On the PE1 device, the link int1 between CE1 and PE1 serves as the AC to access the VPLS EVPN instance VPLS1. The link int11 between CE3 and PE1 serves as the AC. Enter the VPLS EVPN instance VPLS1. Under the VPLS EVPN instance, configure different MAC addresses of the AC port through policies, and set different export RTs. Int1 corresponds to export RT1, and int11 corresponds to export RT2. BGP EVPN neighbors are established between PE1, PE2, and PE3 to transmit RT-2 routes. The L3EVPN network is between PE2, PE3 and PE4, and BGP EVPN neighbors are established to transmit RT-2 routes.

In step 202, CE1 int1 is configured with an IP address (10.0.0.1), and CE3 int11 is configured with an IP address (30.0.0.1). On the bridge devices PE2 and PE3, create L3VPN instances vrf1 and vrf2, and configure the bridge interface IRB1 and its sub-interfaces irb1.1 and irb1.2. The sub-interfaces irb1.1 and irb1.2 are bound to vrf1 and vrf2 respectively. The IP address configurations of interfaces irb1.1 and irb1.2 are in the same network segment (10.1.1.2, 30.1.1.2) as CE1 int1 and CE3 int11 respectively, and PE2 and PE3 are configured with dual gateways, the same addresses and the same MAC. IRB is a Layer 2 interface bound to VPLS EVPN instance VPLS1. The sub-interface of the IRB interface is used as a Layer 3 interface. The irb1.1 sub-interface is referenced in L3VPN instance vrf1, and the irb1.2 sub-interface is referenced in L3VPN instance vrf2. Implementation Interconnection between L2EVPN and L3EVPN.

In step 203, configure ARP proxy on access PE1. When CE1 issues ARP When requested, accessing PE1 will generate EVPN RT-2 MAC and EVPN RT-2 single label (labelA) route MAC1 destined for the CE1 int1 interface, and publish them to the EVPN network. When CE3 issues an ARP request, the accessed PE1 generates the EVPN RT-2 MAC and EVPN RT-2 single label (labelC) route MAC2 destined for the CE3 int11 interface, and publishes them to the EVPN network. The bridge PE (PE2 and PE3) device receives the EVPN RT-2 routes MAC1 and MAC2. According to the export RT1 carried by MAC1, it finds the import RT and the corresponding vrf1 in the three-layer L3VPN instance, and adds the EVPN RT- 2 Routes are imported into the routing table of vrf1. vrf1 assigns a three-layer label (labelB) to the route, fills the label into the label2 field in the routing packet (the format of the EVPN RT-2 packet is shown in Figure 6), and then adds Go to the export RT3 of local vrf1 and publish the route through the EVPN network. After PE4 receives the route, it imports the route into the local vrf1 according to the export RT3 and generates a 32-bit host route corresponding to the CE1 int1 interface address. Similarly, based on the export RT2 carried by MAC2, PE2 and PE3 find the import RT and the corresponding vrf2 in the Layer 3 L3VPN instance, and import the EVPN RT-2 route into the routing table of vrf2, where vrf2 is the route. Assign a Layer 3 label (labelD), fill the label into the label2 field in the routing packet, add the export RT4 of the local vrf2, and publish the route through the EVPN network. After PE4 receives the route, according to the export RT4, Import the route into local vrf2 and generate a 32-bit host route corresponding to the CE3 int11 interface address.

In step 204, if an address IP1 (100.0.0.1) on CE1 wants to communicate with the address IP2 (200.0.0.1) on CE2, they need to open routes to IP1 and IP2 addresses. To communicate with the IP3 (110.0.0.1) address on CE3 and the IP4 (210.0.0.1) address on CE4, they need to open routes to IP3 and IP4 addresses. The following uses static routing as an example.

In step 205, open the route from CE1 to CE2 IP2, configure a static route on PE4 to the private network vrf1 of CE2 IP2, the next hop is the int6 address of CE2, and publish the private network static route to the bridge device through BGP EVPN On PE2 and PE3. On PE2 and PE3, the next hop of the route to IP2 is PE4. Configure a static route on CE1 to CE2 IP2, with the next hop being irb1.1. The route from CE1 to CE2 IP2 is opened.

Open the route from CE3 to CE4 IP4, and configure a route on PE4 to CE4 IP4. For the static route of private network vrf2, the next hop is the int16 address of CE4. The private network static route is published to the bridge devices PE2 and PE3 through BGP EVPN. On PE2 and PE3, the next hop of the route to IP4 is PE4. Configure a static route on CE3 to CE4 IP4, and the next hop is irb1.2. The route from CE3 to CE4 IP4 is opened.

In step 206, on the contrary, open the route from CE2 to CE1 IP1, configure a static route on CE2 to CE1 IP1, and the next hop is the int6 interface address on PE4. Configure a static route on the bridging device PE2 and PE2 to the private network vrf1 of CE1 IP1. The next hop is the int1 interface address of CE1. The private network static route is published to PE4 through L3EVPN. The next hops of the route from PE4 to IP1 are PE2 and PE3. Configure a static route on CE2 to CE1 IP1, and the next hop is the int6 interface address on PE4. The route from CE2 to CE1 IP1 is open.

Open the route from CE4 to CE3 IP3. Configure a static route on CE4 to CE3 IP3. The next hop is the int16 interface address on PE4. Configure a static route on the bridge device PE2 and PE2 to the private network vrf2 of CE3 IP3. The next hop is the int11 interface address of CE3. The private network static route is published to PE4 through L3EVPN. The next hops of the route from PE4 to IP3 are PE2 and PE3. Configure a static route on CE2 to CE3 IP3, and the next hop is the int16 interface address on PE4. The route from CE2 to CE1 IP1 is open.

In step 207, as shown in Figure 8, the PE1 device is configured with the ARP proxy function. PE2 and PE3 obtain the EVPN RT-2 route MAC1 to CE1 int1, which carries a layer of labels (label 100, fill in 100 at label 1 of the message). , this label is the private network label of L2EVPN; PE2 and PE3 obtain the EVPN RT-2 route MAC2 of CE3 int11, which carries a layer of label (label 110, fill in 110 in the packet label1), this label is the private network label of L2EVPN; Match the export RT and L3 import RT of L2 on PE2 and PE3, import the EVPN RT-2MAC1 route into the routing table of vrf1, vrf1 assigns a three-layer label (label 200) to the route, and fills label 200 into EVPN RT. -2 Route the label2 field in the MAC1 message, then add the export RT3 (11.0.0.1:1) of the local vrf1, and publish the route through the EVPN network; import the EVPN RT-2 MAC2 route into the routing table of vrf2, vrf2 is This route is assigned a layer 3 label (label 210), fill in label 210 into the label2 field in the EVPN RT-2 route MAC2 message, and then add local vrf2 export RT4(22.0.0.1:1), and publish the route through the EVPN network. After PE4 receives the route, it finds that export RT3 (11.0.0.1:1) matches the import RT3 (11.0.0.1:1) in local vrf1, imports the route into local vrf1, and generates the corresponding CE1 int1 (10.0.0.1) 32-bit host route; and the export RT4 (22.0.0.1:1) matches the import RT4 (22.0.0.1:1) in the local vrf2, import the route into the local vrf2, and generate the corresponding CE3 int11 (30.0.0.1) 32-bit host routing.

In step 208, CE2 sends the service flow to CE1, and the PE4 device searches for routes when forwarding the traffic. The 32-bit host route to CE1 int1 (10.0.0.1) is better than the 24-bit network segment route. Check the host route for forwarding. The next hop of the main path is PE2. CE4 sends service flows to CE3, and the PE4 device searches for routes when forwarding traffic. The 32-bit host route to the CE3 int11 (30.0.0.1) address is better than the 24-bit network segment route. Check the host route for forwarding. The next hop of the main path is PE2.

In step 209, when the link between PE1 and PE2 and the link between PE2 and PE3 fail, the EVPN RT-2 routing entry sent by CE1 to PE4 through PE1 and PE2 is revoked, and the EVPN RT-2 routing entry sent by CE1 to PE4 through PE1 and PE3 is revoked. The EVPN RT-2 routing entry of PE4 is still there. Therefore, on the PE4 device, the next hop of the 32-bit host route to the CE1 int1 address is switched to PE3, and the traffic is switched to the PE3-PE1-CE1 path for forwarding. In the same way, the EVPN RT-2 routing entries sent by CE3 to PE4 through PE1 and PE2 are revoked, but the EVPN RT-2 routing entries sent by CE3 to PE4 through PE1 and PE3 are still there. Therefore, on the PE4 device, to the int11 address of CE3 The next hop of the 32-bit host route is switched to PE3, and the traffic is switched to the PE3-PE1-CE3 path for forwarding. When a link fails, traffic switching can be achieved without interruption.

Example 3

This disclosed embodiment is a qualified bridging scenario for L2EVPN VPLS. Through the VLAN ID carried in the L2 packet, the IRB sub-interface matches the VLAN ID, and the Layer 3 private network label assigned by the VRF where the corresponding IRB sub-interface is located is found. In EVPN RT The corresponding label2 field is inserted into the -2 routing packet, thereby converting the route from single label to dual label, publishing it to the EVPN network, and generating host routes on the remote PE to achieve intelligent linkage in EVPN Layer 2 and Layer 3 bridging scenarios.

As shown in Figure 9, multiple CEs on the access side are connected to PE1 and sub-interfaces are connected, in a qualified scenario. In the remote Layer 3 network, multiple CEs are connected through different private network L3VPN instances. To achieve intercommunication between CE1 and remote CE2 through Layer 3 vrf1, and CE3 and remote CE4 through Layer 3 network vrf2, configure ARP proxy on access PE1. When CE1 issues an ARP request, access PE1 will generate EVPN RT -2 MAC and EVPN RT-2 single label routing MAC1, carrying the corresponding sub-interface VLAN ID (vlan id1); when CE3 issues an ARP request, access PE1 generates EVPN RT-2 MAC and EVPN RT-2 single label routing MAC2, Carry the corresponding sub-interface VLAN ID (vlan id2) and publish it to the EVPN network. On the bridge PE (PE2 and PE3) device, configure the sub-interface of the bridge interface IRB for bridging. Bind different sub-interfaces to different VRFs. Find the corresponding IRB sub-interface by searching for the IRB sub-interface with the same VLAN ID value in the packet. The VRF to which the interface belongs imports the route into the routing table of the corresponding vrf, generates an L3 private network label, fills the label into the label2 field in the packet, and then adds the export RT of the local VRF to pass the route. The EVPN network publishes the route. After PE4 receives the route, it imports the route into the local VRF according to the export RT and generates a 32-bit host route corresponding to the CE. A 32-bit host route to CE1 int1 is formed in vrf1 on PE4, and a 32-bit host route to CE2 int11 is formed in vrf2.

When searching for routes when forwarding traffic, 32-bit host routing is better than 24-bit network segment routing. When the link between PE1 and PE2 and the link between PE2 and PE3 both fail, the EVPN RT-2 routing entries sent by CE1 and CE3 to PE4 through PE1 and PE2 are revoked, and the EVPN RT-2 routing entries sent by CE1 and CE3 to PE4 through PE1 and PE3 The EVPN RT-2 routing entry is still there. Therefore, the next hop of the 32-bit host route to the CE1 int1 address is switched to PE3, and the next hop of the 32-bit host route to the CE3 int11 address is switched to PE3, thereby realizing traffic switching. .

The specific process includes the following steps 301 to 309.

In step 301, configure VPLS EVPN qualified instances on PE1, PE2, and PE3. On the PE1 device, the link int1 between CE1 and PE1 serves as the AC, and accesses the VPLS EVPN instance VPLS1. The link int11 between CE3 and PE1 serves as the AC. Accessing VPLS In EVPN instance VPLS1, PE1 accesses different CEs through sub-interfaces. BGP EVPN neighbors are established between PE1, PE2, and PE3 to transmit RT-2 routes. The L3EVPN network is between PE2, PE3 and PE4, and BGP EVPN neighbors are established to transmit RT-2 routes.

In step 302, CE1 int1 is configured with an IP address (10.0.0.1), and CE3 int11 is configured with an IP address (30.0.0.1). On the bridge devices PE2 and PE3, create L3VPN instances vrf1 and vrf2, and configure the bridge interface IRB1 and its sub-interfaces irb1.1 and irb1.2. Sub-interfaces irb1.1 and irb1.2 are bound to vrf1 and vrf2 respectively. The IP address configurations of sub-interfaces irb1.1 and irb1.2 are in the same network segment (10.0.0.2, 30.0.0.2) as CE1 int1 and CE3 int11 respectively. , and PE2 and PE3 are configured with dual gateways, with the same addresses and the same MAC. The IRB is bound to the VPLS EVPN instance VPLS1 as the Layer 2 interface. The sub-interface of the IRB is used as the Layer 3 interface. The irb1.1 interface is referenced in the L3VPN instance vrf1, and the irb1.2 interface is referenced in the L3VPN instance vrf2 to implement L2EVPN and L3EVPN. interconnections between.

In step 303, configure ARP proxy on access PE1. When CE1 issues an ARP request, access PE1 will generate EVPN RT-2 MAC and EVPN RT-2 single label (labelA) route MAC1 destined for the CE1 int1 interface. The packet carries the VLAN ID (vlan id1) corresponding to AC sub-interface int1 and publishes it to the EVPN network. CE3 issues an ARP request and accesses PE1 to generate EVPN RT-2 MAC and EVPN RT-2 single label (labelC) route MAC2 destined for CE3 int11 interface, and the message carries the VLAN ID (vlan id2) corresponding to AC sub-interface int11 ) and publish it to the EVPN network. The bridge PE (PE2 and PE3) device receives the EVPN RT-2 routes MAC1 and MAC2. According to the vlan id1 carried by MAC1, it finds the IRB sub-interface (irb1.1) consistent with vlan id1, and imports this EVPN RT-2 route into vrf1. In the routing table, vrf1 assigns a three-layer label (labelB) to the route MAC1, fills the label into the label2 field in the routing message, and then adds the export RT3 of the local vrf1, publishes the route through the EVPN network, and PE4 receives After arriving at the route, import the route into the local vrf1 according to export RT3, and generate a 32-bit host route corresponding to the CE1 int1 interface address.

In the same way, PE2 and PE3 find the IRB sub-interface (irb1.2) consistent with vlan id2 based on the vlan id2 carried by MAC2, and import the EVPN RT-2 route into the routing table of vrf2. vrf2 allocates a three-layer label to the route MAC2. (labelD), and fill in the label2 field in the routing message, then add the export RT4 of the local vrf2, and publish the route through the EVPN network. After PE4 receives the route, it imports the route according to the export RT4 In local vrf2, and generate a 32-bit host route corresponding to the CE3 int11 interface address.

Unless otherwise specified, steps 304 to 309 are the same as steps 204 to 209 and will not be described again.

In a second aspect, an embodiment of the present disclosure provides an electronic device, as shown in Figure 10, The electronic device includes: one or more processors 501; a memory 502, on which one or more computer programs are stored. When the one or more computer programs are executed by one or more processors, the one or more processors Implement the route publishing method according to any one of the above first aspects; and, one or more I/O interfaces 503 are connected between the processor and the memory, and are configured to implement information interaction between the processor and the memory.

The processor 501 is a device with data processing capabilities, including but not limited to a central processing unit (CPU), etc.; the memory 502 is a device with data storage capabilities, including but not limited to random access memory (RAM, more specifically such as SDRAM). , DDR, etc.), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory (FLASH); the I/O interface (read-write interface) 503 is connected between the processor 501 and the memory 502, and can Implement information interaction between the processor 501 and the memory 502, which includes but is not limited to a data bus (Bus), etc.

In some implementations, processor 501, memory 502, and I/O interface 503 are connected to each other and, in turn, to other components of the computing device via bus 504.

In a third aspect, embodiments of the present disclosure provide a computer-readable storage medium. As shown in Figure 11, a computer program is stored on the computer-readable storage medium. When the computer program is executed by a processor, any one of the above-mentioned aspects of the first aspect is implemented. Route publishing method.

The route publishing method proposed in this disclosure uses MAC/IP Advertisement Route (RT-2) simultaneously in a centralized EVPN Layer 2 and Layer 3 bridging scenario when no specific BFD session can be traced. Carrying the first label field and the second label field triggers the three-layer PE to generate a 32-bit host route, realizing the linkage between RT-2 routing and host routing. When a fault occurs in the Layer 2 network, by deleting the RT-2 route of the main path, the Layer 3 PE is linked to delete the 32-bit host route corresponding to the main path to achieve switching to the backup path. Since there is no need to bind BFD detection to service routes, the present disclosure not only reduces the configuration complexity on the device, but also saves the BFD resources of the device and relieves network pressure.

Those of ordinary skill in the art can understand that all or some steps, systems, and functional modules/units in the devices disclosed above can be implemented as software, firmware, hardware, and appropriate combinations thereof.

In a hardware implementation, between the functional modules/units mentioned in the above description A partitioning does not necessarily correspond to a partitioning of physical components; for example, one physical component may have multiple functions, or one function or step may be performed cooperatively by several physical components. Some or all of the physical components may be implemented as software executed by a processor, such as a central processing unit, a digital signal processor, or a microprocessor, or as hardware, or as an integrated circuit, such as an application specific integrated circuit. circuit. Such software may be distributed on computer-readable media, which may include computer storage media (or non-transitory media) and communication media (or transitory media). As is known to those of ordinary skill in the art, the term computer storage media includes volatile and nonvolatile media implemented in any method or technology for storage of information such as computer readable instructions, data structures, program modules or other data. removable, removable and non-removable media. Computer storage media includes, but is not limited to, RAM, ROM, EEPROM, flash memory or other memory technology, CD-ROM, Digital Versatile Disk (DVD) or other optical disk storage, magnetic cassettes, tapes, disk storage or other magnetic storage devices, or may Any other medium used to store the desired information and that can be accessed by a computer. Additionally, it is known to those of ordinary skill in the art that communication media typically embodies computer readable instructions, data structures, program modules or other data in a modulated data signal such as a carrier wave or other transport mechanism, and may include any information delivery media .

The preferred embodiments of the present disclosure have been described above with reference to the accompanying drawings, but the scope of rights of the present disclosure is not thereby limited. Any modifications, equivalent substitutions and improvements made by those skilled in the art without departing from the scope and essence of the present disclosure shall be within the scope of rights of the present disclosure.

Claims

A route publishing method, executed by the first provider edge PE device, includes:

Receive a first routing message from the second PE device; wherein the first routing message is used to advertise a MAC/IP advertised route, the first routing message includes a first label field and a second label field, the first label The field includes a first tag, and the second tag field is blank;

Filling the second label field with a second label to generate a second routing message, wherein the first label is used to represent Layer 2 business traffic forwarding, and the second label is used to represent Layer 3 business traffic forwarding; and

Send the second routing message to the third PE device.
The route publishing method according to claim 1, wherein filling the second label field with a second label to generate a second routing message includes:

Determine the Layer 3 virtual private network L3VPN private network label corresponding to the first routing message according to the first routing message; and

Fill the L3VPN private network label into the second label field of the first routing message to generate a second routing message.
The route publishing method according to claim 2, wherein the determining the L3VPN private network label corresponding to the first routing message according to the first routing message includes:

Determine the corresponding L3VPN interface according to the first routing message;

Determine the L3VPN instance corresponding to the L3VPN interface; and

The private network label assigned by the L3VPN instance to the network segment address of the L3VPN interface is used as the L3VPN private network label corresponding to the first routing message.
The route publishing method according to claim 2, wherein the first routing message carries an identifier of a virtual LAN VLAN, and the L3VPN private number corresponding to the first routing message is determined based on the first routing message. Net tags, including:

Determine the corresponding L3VPN sub-interface according to the identifier of the VLAN;

Determine the L3VPN instance corresponding to the L3VPN sub-interface; and

The private network label assigned by the L3VPN instance to the network segment address of the L3VPN sub-interface is used as the L3VPN private network label corresponding to the first routing message.
The route publishing method according to any one of claims 1 to 4, said method further comprising:

When the first route parsed according to the first routing message meets the inactivity condition, the second routing message is withdrawn from the third PE device.
The route publishing method according to claim 5, wherein the inactivity condition includes at least one of the following conditions:

The next hop of the first route is unreachable, the first route cannot iterate to the tunnel, and a withdrawal message of the first route is received.
The route publishing method according to any one of claims 1 to 4, wherein the first PE device is a bridge PE device in an Ethernet Virtual Private Network (EVPN) Layer 2 and Layer 3 bridging scenario.
The route publishing method according to claim 7, wherein the second PE device is a Layer 2 PE device in an EVPN Layer 2 and Layer 3 bridging scenario, and the third PE device is a Layer 3 PE device in an EVPN Layer 2 and Layer 3 bridging scenario. PE equipment.
An electronic device, the electronic device includes:

one or more processors;

A memory having one or more computer programs stored thereon, which when the one or more computer programs are executed by the one or more processors, causes the one or more processors to implement the method according to claims 1 to 8 The route publishing method described in any one of the above; and

One or more I/O interfaces are connected between the processor and the memory, and are configured to implement information exchange between the processor and the memory.
A computer-readable storage medium. A computer program is stored on the computer-readable storage medium. When the computer program is executed by a processor, the route publishing method according to any one of claims 1 to 8 is implemented.