Message forwarding method and equipment

Abstract

The invention provides a message forwarding method and equipment. The method comprises the following steps of: searching a matched three-layer forwarding table item based on a destination IP (Internet Protocol) address of a forwarded Ethernet message, and acquiring a next hop with overlay from the three-layer forwarding table item; modifying a source MAC (Media Access Control) address and a destination MAC address of the forwarded Ethernet message based on two-layer packaging information associated with the next hop of the overlay; the method comprises the following steps of: acquiring an associated underlying ECMP group (underlying ECMP group) from a middle table item in which a next hop of an overlay is recorded, and selecting a member path from an underlying ECMP table item of the underlying ECMP group; according to tunnel packaging information of a virtual port of the VXLAN tunnel, the Ethernet message is packaged into a VXLAN message; and sending the VXLAN message through the selected member path.

Description

Technical Field

[0001] This application relates to communication technology, specifically a message forwarding method and device. Background Technology

[0002] With the continuous expansion of data center network scale, multi-tenant (Tenan) network virtualization technology based on Virtual Extensible Local Area Network (VXLAN) is widely adopted. In a typical leaf-spine architecture, leaf switches act as virtual tunnel endpoints (VTEPs), carrying Layer 2 / Layer 3 services of each tenant on the overlay network through VXLAN tunnels. At the same time, load balancing and link redundancy are achieved in the underlay network through equal-cost multi-path forwarding (ECMP).

[0003] When the outgoing port of a VXLAN tunnel consists of multiple next-hops forming an Underlay ECMP Group, due to the hardware limitations of the switching chip, Address Resolution Protocol (ARP) entries, Neighbor Discovery (ND) entries, and routing table entries pointing to the VXLAN tunnel cannot be directly associated with the underlying ECMP group. Instead, they must first point to an Overlay ECMP Group, which then points back to the underlying ECMP group. This results in each Layer 3 entry pointing to the VXLAN tunnel requiring a dedicated Overlay ECMP Group and its corresponding ECMP Entry, making Overlay ECMP Group resources a critical bottleneck limiting the scale of VXLAN networks.

[0004] For example, when the overlay ECMP group capacity of a certain switching chip is only 4K, the total number of Layer 3 entries pointing to VXLAN tunnels is also limited to 4K. Assuming there are 8 tenants on a single device, each tenant corresponds to a Virtual Switch Instance (VSI), and each VXLAN tunnel under each VSI corresponds to at least one Layer 3 entry, then under the premise of "number of VSIs × number of VXLAN tunnels ≤ 4K", a maximum of 8 × 512 Layer 3 entries can be supported, that is, the upper limit of the number of VXLAN tunnels is 512. This limits the number of leaf nodes in the VXLAN network, making the actual buildable VXLAN network size much lower than the theoretical capacity of the overlay next-hop table or the routing table itself. As the number of tenants, VSIs, and service prefixes continue to grow, the entries pointing to VXLAN tunnels will quickly exhaust the resources of the overlay ECMP group and overlay ECMP entries. This will not only make it difficult to expand new VXLAN tunnels, but will also crowd out the hardware space of routing entries that truly need the overlay equal-cost routing function, thereby affecting the normal use of the overlay network's equal-cost routing function and severely restricting the scalability and reliability of large-scale VXLAN networks. Summary of the Invention

[0005] The purpose of this application is to provide a message forwarding method and device to improve system scalability.

[0006] To achieve the above objectives, this application also provides a packet forwarding method, which includes: finding a matching Layer 3 forwarding table entry based on the destination IP address of the forwarded Ethernet packet; obtaining the overlay next hop from the Layer 3 forwarding table entry; modifying the source MAC address and destination MAC address of the forwarded Ethernet packet based on the Layer 2 encapsulation information associated with the overlay next hop; obtaining the associated underlay ECMP group from the intermediate table entry recording the overlay next hop; selecting a member path from the underlay ECMP entries of the underlay ECMP group; encapsulating the Ethernet packet into a VXLAN packet according to the tunnel encapsulation information of the virtual port of the VXLAN tunnel; and sending the VXLAN packet through the selected member path.

[0007] To achieve the above objectives, this application also provides a packet forwarding device, which includes a processor, a machine-readable storage medium, a switching chip, and a network interface. The processor executes machine-executable instructions recorded on the machine-readable storage medium to perform the following operations: finding a matching Layer 3 forwarding table entry based on the destination IP address of the forwarded Ethernet packet; obtaining the overlay next hop from the Layer 3 forwarding table entry; modifying the source MAC address and destination MAC address of the forwarded Ethernet packet based on the Layer 2 encapsulation information associated with the overlay next hop; obtaining the associated underlay ECMP group from the intermediate table entry recording the overlay next hop; selecting a member path from the underlay ECMP table entry of the underlay ECMP group; encapsulating the Ethernet packet into a VXLAN packet according to the tunnel encapsulation information of the virtual port of the VXLAN tunnel; and sending the VXLAN packet through the selected member path.

[0008] The beneficial effect of this application is that it decouples overlay forwarding from underlay equivalent multipath forwarding, thereby improving the correctness of packet forwarding and the scalability of the system. Attached Figure Description

[0009] Figure 1 A flowchart of an embodiment of a message forwarding method provided in this application; Figure 2 This is a schematic diagram of the network architecture provided in the embodiments of this application; Figure 3 A flowchart illustrating an embodiment of the virtual port for a VXLAN tunnel provided in this application; Figure 4 for Figure 2 The flowchart shown in the embodiment illustrates the learning of three-layer forwarding table entries; Figure 5 for Figure 2 The flowchart for performing Layer 3 forwarding is shown in the example. Figure 6 for Figure 2 Flowchart for deleting a three-level forwarding table entry in the example; Figure 7 This is a schematic diagram illustrating an embodiment of a message forwarding device provided in this application. Detailed Implementation

[0010] The following detailed description will be provided with reference to several examples illustrated in the accompanying figures. In this detailed description, numerous specific details are used to provide a comprehensive understanding of the present application. Known methods, steps, components, and circuits are not described in detail in the examples to avoid obscuring their meaning.

[0011] In the terminology used, the term "including" means including but not limited to; the term "containing" means including but not limited to; the terms "above," "within," and "below" include the number itself; the terms "greater than" and "less than" mean not including the number itself. The term "based on" means based on at least a portion of them.

[0012] Figure 1 This application provides a flowchart of an embodiment of a message forwarding method; the embodiment includes, Step 101: Obtain the associated underlying ECMP group from the intermediate table entry that records the next hop of the overlay. Step 102: Select a member path from the underlay ECMP entry of the underlay ECMP group; Step 103: Encapsulate the Ethernet packet into a VXLAN packet according to the tunnel encapsulation information of the virtual port of the VXLAN tunnel. Step 104: Send a VXLAN message through the selected member path.

[0013] Figure 1 The beneficial effect of this embodiment is that it decouples overlay forwarding from underlay equivalent multipath forwarding, thereby improving the correctness of packet forwarding and the scalability of the system.

[0014] Figure 2 A schematic diagram of the network architecture provided for this application. Figure 2 In the Spine-Leaf architecture shown, the backbone nodes Spine1 and Spine2 and the leaf nodes Leaf11-Leaf14 are all Layer 3 switching devices. Through routing protocols such as OSPF or EBGP, they achieve equal cost multicast path (ECMP) load balancing and link backup.

[0015] On Leaf11-Leaf14, ECMP resources are divided into underlay (lower-level) available resources and overlay (overlay) available resources. The size of the ECMP group and its entries on Leaf11-Leaf14 is fixed. In existing technology, the underlay ECMP group and overlay ECMP group of the leaf node device each occupy half of the ECMP group resources, which remains unchanged. Figure 2In the illustrated embodiment, a portion of the overlay ECMP entry resources on Leaf11-Leaf14 are allocated as intermediate entry resources for intermediate tables. Each intermediate entry records the correspondence between an overlay next hop and an underlay ECMP group, thereby saving overlay ECMP group resources and expanding the scale of VXLAN networking.

[0016] In this application, the key for each intermediate table entry is the overlay next hop, and the corresponding data is the underly ECMP group.

[0017] In this application, every two intermediate table entries occupy one ECMP table entry resource. In one example, the overlay next-hop table has a size of 16K, and 8K of the overlay next-hop table entries are allocated to the intermediate table entries to store the 16K mappings. Therefore, the size of the intermediate table entries is equal to the size of the overlay next-hop table.

[0018] Figure 3 A flowchart illustrating an embodiment of the virtual port for a VXLAN tunnel provided in this application; Figure 3 for Figure 2 A flowchart illustrating the process of requesting a virtual port for a VXLAN tunnel in this example; this embodiment includes the following steps: Step 301: Create a VXLAN tunnel connecting the remote VTEP; The VTEP IP address of Leaf11 is 1.1.1.1; the VTEP IP address of Leaf14 is 10.1.1.1. Create a VXLAN tunnel named Tunnel 11 on Leaf11, connecting it to Leaf14.

[0019] Step 302: Request a VPN for the created VXLAN tunnel.

[0020] Leaf11 requests a virtual port VP 100 for the VXLAN tunnel Tunnel11. In this way, the interface information of the virtual port VP100 records the tunnel encapsulation information of the VXLAN tunnel Tunnel 11, including the local device VTEP IP address of 1.1.1.1 and the peer VETP IP address of 10.1.1.1.

[0021] Step 303: Apply for an underlay ECMP group for the created VXLAN tunnel.

[0022] Leaf11 requests an underlay ECMP group1000 for the VXLAN tunnel Tunnle and requests an underlay ECMP table entry to record the member interfaces of the underlay ECMP group1000.

[0023] In existing technology, when the outgoing port of a VLXAN tunnel created by a leaf node changes from a single next hop to multiple next hops, it is necessary to apply for an underlay ECMP group and the corresponding underlay ECMP entry, and record the multiple next hops of the outgoing port of the VXLAN tunnel in the underlay ECMP entry; when the outgoing port of a VLXAN tunnel changes from multiple next hops to a single next hop, it is necessary to release the underlay ECMP group and the corresponding underlay ECMP entry.

[0024] In this application, Leaf11 requests an underlay ECMP group1000 and an underlay ECMP entry when creating a VXLAN tunnel. This is to avoid repeated requests and releases caused by the next-hop route oscillation of VXLAN tunnel Tunnel 11.

[0025] Step 304: Bind the virtual port of the VXLAN tunnel to the requested underlay ECMP group.

[0026] Leaf11 records underlay ECMP group 1000 in the interface information of virtual port VP100.

[0027] Step 305: Associate the created VXLAN tunnel with the requested underlay ECMP group.

[0028] Leaf 11 records the virtual port VP100 and the underlay ECMP group 1000 in the tunnel information of VXLAN tunnel Tunnel11, and indirectly associates the VXLAN tunnel Tunnel11 with the underlay ECMP group 1000 through the virtual port VP100.

[0029] Figure 4 for Figure 2 The flowchart shown in the embodiment illustrates the learning of three-layer forwarding table entries; this embodiment includes the following steps: Step 401: Determine that the next-hop address of the learned routing table entry is the IP address of the peer VTEP of the created VXLAN tunnel; Leaf11 learns a network segment routing table entry with the prefix 192.168.10.0 / 24. Leaf11 obtains the next-hop address IP address 192.168.10.1 for this network segment routing table entry. Leaf11 recognizes this next-hop address IP address 10.1.1.1 as the peer VTEP IP address, and a VXLAN tunnel tunnel11 has already been established to reach this VTEP IP address 10.1.1.1. The processing mechanism when Leaf11 learns a host routing table entry is the same, and will not be described again here.

[0030] Step 402: Request an overlay next hop, record the virtual port associated with the VXLAN tunnel for the overlay next hop; record the source MAC address and destination MAC address of the inner MAC encapsulation information.

[0031] Leaf11 requests an overlay next-hop NH20 for routes whose next-hop address is the VETP IP address of the peer of VXLAN tunnel Tunnel11. In the next-hop information of the overlay next-hop NH20, the virtual port VP100 associated with VXLAN tunnel Tunnel11 is recorded, as well as the inner source MAC address and inner destination MAC address of the VXLAN packets sent through VXLAN tunnel Tunnel11.

[0032] Step 403: Record the mapping relationship between the requested overlay next hop and the underlay ECMP group associated with the VXLAN tunnel in the intermediate table entry.

[0033] Step 404: Synchronize the learned routing table entries into Layer 3 forwarding table entries, which record the IP prefix of the routing table entry corresponding to the requested overlay next hop.

[0034] Leaf11 synchronizes the routing table entry for IP address prefix 192.168.10.0 / 24 to the FIB table (Forwarding Information Based). This FIB table entry records the IP address prefix 192.168.10.0 / 24 of the network segment routing table entry and the overlay next hop NH20, so as to quickly find the VXLAN tunnel Tunnle11 that reaches the network segment.

[0035] Figure 5 for Figure 2 The flowchart for performing Layer 3 forwarding in this embodiment includes the following steps: Step 501: Find a matching Layer 3 forwarding table entry based on the destination IP address of the forwarded Ethernet packet, and obtain the overlay next hop.

[0036] The IP packet sent from terminal T1 to terminal T2 is carried in Ethernet packet 500. Leaf11 receives Ethernet packet 500, determines the destination MAC address and the local gateway MAC address, and performs Layer 3 forwarding.

[0037] Leaf11, based on the destination IP address 192.168.10.10 of Ethernet packet 500, looks up the Layer 3 forwarding table entry with the IP address prefix 192.168.10.0 / 24; and reads the overlay next hop NH20 from that entry.

[0038] Step 502: Based on the Layer 2 encapsulation information associated with the next hop of the overlay, modify the source MAC address and destination MAC address of the forwarded Ethernet packet.

[0039] Leaf11 obtains the inner source MAC address and inner destination MAC address from the next-hop information of the overlay next-hop NH20, and replaces the source MAC address and destination MAC address of the Ethernet packet 500.

[0040] Step 503: Locate the underlay ECMP group associated with the next hop of the overlay, and select a member path from the underlay ECMP entries of the associated underlay ECMP group.

[0041] Leaf11 searches the intermediate table based on the overlay next-hop NH20, retrieves the underlay ECMP group1000 from the found intermediate table entry, calculates the hash value based on the 5-tuple of the inner IP packet of Ethernet packet 500, and selects a member path corresponding to the hash value in the underlay ECMP table entry of underlay ECMP group1000.

[0042] Step 504: Encapsulate the modified Ethernet packet into a VXLAN packet based on the tunnel encapsulation information of the virtual port of the VXLAN tunnel.

[0043] Leaf11 encapsulates the Ethernet packet 500, which has modified the source and destination MAC addresses, into a VXLAN packet based on the VXLAN encapsulation information of the virtual port VP100 of the Ethernet packet 500.

[0044] Step 505: Send a VXLAN message through the selected member path.

[0045] Leaf11 will send the encapsulated VXLAN packets through the ports on the selected member path.

[0046] In this application, the upper limits on topology size and number of tenants are relaxed. Compared to the prior art, where the product of the number of VSIs and the number of VXLAN tunnels cannot exceed the overlay ECMP group specification (e.g., 4K) when multiple tenants exist on a single device, otherwise Layer 3 forwarding table entries pointing to VXLAN tunnels cannot be further distributed, this application significantly improves the scalability of leaf devices in VXLAN networking by introducing an intermediate table and expanding the number of overlay next hops and the number of supported VXLAN tunnels.

[0047] Figure 6 for Figure 2 Flowchart for deleting a three-level forwarding table entry in the example; Step 601: Based on the next hop of the overlay of the three-layer forwarding table entry of the deleted object, search in the intermediate table and delete the matching intermediate table entry found.

[0048] Leaf11 determines to delete the FIB entry with the IP prefix 192.168.10.0 / 24; Leaf11 performs a lookup in the intermediate table based on the overlay next-hop NH20 recorded in this FIB entry; an intermediate entry is found that records both the overlay next-hop NH20 and the underlay ECMP group 1000. Leaf11 deletes this found intermediate entry.

[0049] Step 602: Proceed to the next hop of the overlay of the three-level forwarding table entry of the deleted object.

[0050] Leaf11 will release the overlay next hop NH20 resource for the VXLAN tunnel application to the network segment 192.168.10.0 / 24.

[0051] Step 603: Delete the third-level forwarding table entry for the object to be deleted from the third-level forwarding table.

[0052] Leaf11 removes the FIB entry with the IP prefix 192.168.10.0 / 24 from the FIB table.

[0053] Figure 6 The beneficial effects of this embodiment are that it achieves unified cleanup of three-layer entries, superimposed next hops, and intermediate entries, avoids dangling next hops and intermediate entries from occupying system resources, ensures consistency among multiple forwarding tables, and improves system resource utilization and controllability of forwarding behavior.

[0054] Figure 7This is a schematic diagram of an embodiment of a message forwarding device provided in this application. The device includes a processor 71, a machine-readable storage medium 72, a switching chip 73, and a network interface 74.

[0055] Processor 71 executes the machine-executable instructions recorded in machine-readable storage medium 72 to perform the following operations: finding a matching Layer 3 forwarding table entry based on the destination IP address of the forwarded Ethernet packet; obtaining the overlay next hop from the Layer 3 forwarding table entry; modifying the source MAC address and destination MAC address of the forwarded Ethernet packet based on the Layer 2 encapsulation information associated with the overlay next hop; obtaining the associated underlay ECMP group from the intermediate table entry recording the overlay next hop; selecting a member path from the underlay ECMP entries of the underlay ECMP group; encapsulating the Ethernet packet into a VXLAN packet according to the tunnel encapsulation information of the VXLAN tunnel's virtual port; and sending the VXLAN packet through the selected member path.

[0056] The processor 71 performs the following operations by executing machine-executable instructions recorded in the machine-readable storage medium 72: creating a VXLAN tunnel connecting to the remote VTEP; requesting a virtual port for the VXLAN tunnel; requesting an underlay ECMP group for the VXLAN tunnel; binding the virtual port to the underlay ECMP group index; and associating the VXLAN tunnel with the underlay ECMP group.

[0057] Processor 71 executes the machine-executable instructions recorded in machine-readable storage medium 72 to perform the following operations: determine that the next-hop address of the learned routing table entry is the IP address of the peer VTEP of the VXLAN tunnel; request an overlay next hop and record the overlay next hop associated with the virtual port; record the source MAC address and destination MAC address associated with the inner MAC encapsulation information of the overlay; record the mapping relationship between the overlay next hop and the underlay ECMP group associated with the VXLAN tunnel in an intermediate table entry; and synchronize the routing table entry into a Layer 3 forwarding table entry; wherein the Layer 3 forwarding table entry records that the IP prefix of the routing table entry corresponds to the overlay next hop.

[0058] Processor 71 performs the following operations by executing machine-executable instructions recorded in machine-readable storage medium 72: determining that a Layer 3 forwarding table entry is to be deleted; deleting the intermediate table entry that records the overlay next hop from the intermediate table; deleting the overlay next hop of the Layer 3 forwarding table entry; and deleting the Layer 3 forwarding table entry from the Layer 3 forwarding table. Each pair of intermediate table entries shares one ECMP table entry resource.

[0059] In this application, a machine-readable storage medium can be any electronic, magnetic, optical, or other physical storage device used to store or contain information (such as executable instructions, data, etc.). For example, any machine-readable storage medium herein can be any type of random access memory (RAM), volatile memory, non-volatile memory, flash memory, storage drive (such as a hard disk drive), solid-state drive, any type of optical disc (such as an optical disc, DVD, etc.), and similar devices, or combinations thereof. Furthermore, any machine-readable storage medium described herein can be a non-transitory machine-readable storage medium.

[0060] The above description is merely a preferred embodiment of this application and is not intended to limit this application. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of protection of this application.

Claims

1. A message forwarding method, characterized in that, The method includes, Based on the destination IP address of the forwarded Ethernet packet, a matching Layer 3 forwarding table entry is found, and the next hop of the overlay is obtained from the Layer 3 forwarding table entry; Based on the Layer 2 encapsulation information associated with the next hop of the overlay, the source MAC address and destination MAC address of the forwarded Ethernet packet are modified. From the intermediate table entry that records the next hop of the overlay, obtain the associated underlying equivalent multipath group (underlayECMP group), and select a member path from the underlay ECMP table entries of the underlay ECMP group; The Ethernet packets are encapsulated into VXLAN packets based on the tunnel encapsulation information of the virtual port of the VXLAN tunnel. The VXLAN message is sent via the selected member path.

2. The method according to claim 1, characterized in that, Before finding a matching Layer 3 forwarding table entry based on the destination IP address of the forwarded Ethernet packet, the method further includes... Create the VXLAN tunnel connecting the remote VTEP; Request the virtual port for the VXLAN tunnel; Apply for the underlay ECMP group for the VXLAN tunnel; Bind the virtual port to the underlay ECMP group, and associate the VXLAN tunnel with the underlay ECMP group.

3. The method according to claim 2, characterized in that, Before finding a matching Layer 3 forwarding table entry based on the destination IP address of the forwarded Ethernet packet, the method further includes, The next-hop address of the learned routing table entry is determined to be the IP address of the peer VTEP of the VXLAN tunnel; Request the overlay next hop and record the overlay next hop associated with the virtual port; Record the source MAC address and destination MAC address associated with the inner MAC encapsulation information of the overlay; The mapping relationship between the overlay next hop and the underlay ECMP group associated with the VXLAN tunnel is recorded in the intermediate table entry; The routing table entry is synchronized to the Layer 3 forwarding table entry; wherein the Layer 3 forwarding table entry records the IP prefix of the routing table entry corresponding to the next hop of the overlay.

4. The method according to claim 3, characterized in that, The method also includes, The three-level forwarding entries have been determined to be deleted; From the intermediate table, delete the intermediate table entry that records the next hop of the overlay; Delete the next hop of the overlay in the three-layer forwarding table entry; Delete the aforementioned three-level forwarding table entry from the three-level forwarding table.

5. The method according to claim 4, characterized in that, Each pair of intermediate table entries shares one ECMP entry resource.

6. A message forwarding device, the device comprising a processor, a machine-readable storage medium, a switching chip, and a network interface; characterized in that, The processor performs the following operations by executing machine-executable instructions recorded on the machine-readable storage medium. Based on the destination IP address of the forwarded Ethernet packet, a matching Layer 3 forwarding table entry is found, and the next hop of the overlay is obtained from the Layer 3 forwarding table entry; Based on the Layer 2 encapsulation information associated with the next hop of the overlay, the source MAC address and destination MAC address of the forwarded Ethernet packet are modified. From the intermediate table entry that records the next hop of the overlay, obtain the associated underlying equivalent multipath group (underlayECMP group), and select a member path from the underlay ECMP table entries of the underlay ECMP group; The Ethernet packets are encapsulated into VXLAN packets based on the tunnel encapsulation information of the virtual port of the VXLAN tunnel. The VXLAN message is sent via the selected member path.

7. The device according to claim 6, characterized in that, The processor, by executing machine-executable instructions recorded on the machine-readable storage medium, also performs the following operations: Create the VXLAN tunnel connecting the remote VTEP; Request the virtual port for the VXLAN tunnel; Apply for the underlay ECMP group for the VXLAN tunnel; Bind the virtual port to the underlay ECMP group index, and associate the VXLAN tunnel with the underlay ECMP group.

8. The device according to claim 7, characterized in that, The processor, by executing machine-executable instructions recorded on the machine-readable storage medium, also performs the following operations: The next-hop address of the learned routing table entry is determined to be the IP address of the peer VTEP of the VXLAN tunnel; Request the overlay next hop and record the overlay next hop associated with the virtual port; Record the source MAC address and destination MAC address associated with the inner MAC encapsulation information of the overlay; The mapping relationship between the overlay next hop and the underlay ECMP group associated with the VXLAN tunnel is recorded in the intermediate table entry; The routing table entry is synchronized to the Layer 3 forwarding table entry; wherein the Layer 3 forwarding table entry records the IP prefix of the routing table entry corresponding to the next hop of the overlay.

9. The device according to claim 8, characterized in that, The processor, by executing machine-executable instructions recorded on the machine-readable storage medium, also performs the following operations: The three-level forwarding entries have been determined to be deleted; Delete the intermediate table entry that records the next hop of the overlay; Delete the next hop of the overlay in the three-layer forwarding table entry; Delete the aforementioned three-level forwarding table entry from the three-level forwarding table.

10. The device according to claim 9, characterized in that, Each pair of intermediate table entries shares one ECMP entry resource.

Images