CN107995110B

CN107995110B - Traffic forwarding method and device

Info

Publication number: CN107995110B
Application number: CN201711133675.3A
Authority: CN
Inventors: 黄李伟; 王伟
Original assignee: Hangzhou H3C Technologies Co Ltd
Current assignee: Hangzhou H3C Technologies Co Ltd
Priority date: 2017-11-16
Filing date: 2017-11-16
Publication date: 2020-12-01
Anticipated expiration: 2037-11-16
Also published as: CN107995110A

Abstract

The disclosure relates to a traffic forwarding method and device. The method is applied to a first VTEP in a distributed aggregation system, wherein the distributed aggregation system comprises at least two VTEPs, and different VTEPs are configured with different MAC addresses, and the method comprises the following steps: acquiring the forwarding performance of each VTEP in the distributed aggregation system under the condition that a virtual machine connected with the distributed aggregation system is detected to be on line; and returning the MAC address of the VTEP for load sharing to the virtual machine according to the current traffic and forwarding performance of each VTEP. In the distributed aggregation system, different MAC addresses can be set for different VTEPs as actual MAC addresses, and load sharing of the virtual machine service traffic is dynamically determined according to the current traffic and forwarding performance of two VTEP devices in the distributed aggregation system, so that all VTEPs in the distributed aggregation system can be efficiently used.

Description

Traffic forwarding method and device

Technical Field

The present disclosure relates to the field of communications technologies, and in particular, to a traffic forwarding method and apparatus.

Background

Fig. 1 is a schematic diagram of a multi-chassis link aggregation group (MLAG) proposed based on distributed aggregation and a peer link of an Ethernet Virtual Private Network (EVPN). As shown in fig. 1, the meaning of each component is as follows:

the EVPN is a two-layer VPN (Virtual Private Network) technology, the control plane uses MP-BGP (multi Protocol Border Gateway Protocol) to notify EVPN routing information, and the data plane uses VXLAN (Virtual eXtensible Local Area Network) encapsulation to forward messages.

VM (Virtual Machine): multiple virtual machines can be created on one server, and different virtual machines can belong to different VXLANs. Virtual machines belonging to the same VXLAN are in the same logic two-layer network and are communicated with each other in two layers; two levels of isolation between virtual machines belonging to different VXLANs. VXLAN is identified by VXLAN ID, also known as VNI (VXLAN Network Identifier), which is 24 bits long.

VTEP (VXLAN Tunnel End Point ): edge device of VXLAN. The VXLAN processing is performed on the VTEP, for example, to identify the VXLAN to which the ethernet data frame belongs, to perform two-layer forwarding on the data frame based on the VXLAN, and to encapsulate/decapsulate the packet. The VTEP may be an independent physical device or a server where the virtual machine is located.

VXLAN tunnel: a point-to-point logical tunnel between two VTEPs. After encapsulating a VXLAN header, a UDP header and an IP header for a data frame, the VTEP forwards the encapsulated message to a far-end VTEP through a VXLAN tunnel, and the far-end VTEP decapsulates the encapsulated message.

Core equipment: devices in an IP core network. The core device does not participate in VXLAN processing, and only needs to forward the message in three layers according to the destination IP address of the encapsulated message.

VSI (Virtual Switch Instance): a virtual switching instance on the VTEP provides a two-layer switching service for VXLAN. The VSI can be viewed as a VXLAN-based virtual switch on a VTEP that performs layer two forwarding, with all the functions of a traditional ethernet switch, including source MAC address learning, MAC address aging, flooding, etc. VSIs correspond one-to-one to VXLANs.

AC (Attachment Circuit, access Circuit): the VTEP connects physical or virtual circuits of the local site. On a VTEP, the three-tier interface or Ethernet service instance (service instance) associated with a VSI is referred to as the AC. Wherein the ethernet service instance is created on a layer two ethernet interface, the ethernet service instance defining a series of matching rules for matching data frames received from the layer two ethernet interface. Service instances are configured below the 1 two-layer physical port.

The significance of this AC is: if the data message entering from the physical port ten1/0/1 carries tag10, the data message enters vsi vpnb for forwarding, namely, the mapping from the vlan tag10 to the encapsulated vxlan id 100 message is completed.

DRNI (Distributed Resilient Network Interconnect) is a cross-device link aggregation technology, where two physical devices are virtualized into one device on an aggregation layer to implement cross-device link aggregation, thereby providing device-level redundancy protection and traffic load sharing.

IPP (Intra-Portal Port, internal control Link Port): and a two-layer aggregation interface connected with the DR neighbor equipment at the opposite end and used for internal control. Each DR equipment has only one IPP port. DRNI protocol messages are transmitted between the DR devices through IPLs (Intra-Portal links). A DR system has only one IPL. In FIG. 2, the link between VTEP B and VTEP C is the link of the IPP port.

DR interface (Distributed Relay interface): and the two-layer aggregation interface is connected with an external device. DR interfaces connected to the same aggregation group on the external device belong to the same DR group (Distributed-Relay group). As shown in fig. 2, the two-layer aggregation interface connected to VM a on VTEP B and the two-layer aggregation interface connected to VM a on VTEP C belong to the same DR group.

As shown in fig. 1, traffic forwarding from VM1 to VM2 may pass through any one of the devices VTEP1 or VTEP3 devices of the distributed aggregation. When one device is hung, the other device can not be influenced to take over work.

Disclosure of Invention

In view of this, the present disclosure provides a traffic forwarding method and apparatus.

According to an aspect of the present disclosure, a traffic forwarding method is provided, where the traffic forwarding method is applied to a first VTEP in a distributed aggregation system, where the distributed aggregation system includes at least two VTEPs, and different VTEPs are configured with different MAC addresses, and the method includes:

acquiring the forwarding performance of each VTEP in the distributed aggregation system under the condition that a virtual machine connected with the distributed aggregation system is detected to be on line;

and returning the MAC address of the VTEP for load sharing to the virtual machine according to the current traffic and forwarding performance of each VTEP.

According to another aspect of the present disclosure, there is provided a traffic forwarding apparatus applied in a first VTEP in a distributed aggregation system, where the distributed aggregation system includes at least two VTEPs, and different VTEPs are configured with different MAC addresses, the apparatus including:

the acquisition module is used for acquiring the forwarding performance of each VTEP in the distributed aggregation system under the condition that a virtual machine connected with the acquisition module is detected to be online;

and the returning module is used for returning the MAC address of the VTEP for load sharing to the virtual machine according to the current traffic and forwarding performance of each VTEP.

In the distributed aggregation system, different MAC addresses can be set for different VTEPs as actual MAC addresses, and load sharing of the virtual machine service traffic is dynamically determined according to the current traffic and forwarding performance of two VTEP devices in the distributed aggregation system, so that all VTEPs in the distributed aggregation system can be efficiently used.

Other features and aspects of the present disclosure will become apparent from the following detailed description of exemplary embodiments, which proceeds with reference to the accompanying drawings.

Drawings

The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate exemplary embodiments, features, and aspects of the disclosure and, together with the description, serve to explain the principles of the disclosure.

Fig. 1 is a schematic diagram of MLAG and EVPN based distributed aggregation networking.

Fig. 2 is a schematic diagram of integrated networking equivalent load sharing for distributed aggregation and EVPN.

Fig. 3 shows a flow chart of a traffic forwarding method according to an embodiment of the present disclosure.

Fig. 4 shows another flow chart of a traffic forwarding method according to an embodiment of the present disclosure.

Fig. 5 is a schematic diagram illustrating an application scenario of a traffic forwarding method according to an embodiment of the present disclosure.

Fig. 6 shows a block diagram of a traffic forwarding apparatus according to an embodiment of the present disclosure.

Fig. 7 shows another block diagram of a traffic forwarding apparatus according to an embodiment of the present disclosure.

Fig. 8 is a schematic diagram illustrating a traffic forwarding device, according to an example embodiment.

Detailed Description

Various exemplary embodiments, features and aspects of the present disclosure will be described in detail below with reference to the accompanying drawings. In the drawings, like reference numbers can indicate functionally identical or similar elements. While the various aspects of the embodiments are presented in drawings, the drawings are not necessarily drawn to scale unless specifically indicated.

The word "exemplary" is used exclusively herein to mean "serving as an example, embodiment, or illustration. Any embodiment described herein as "exemplary" is not necessarily to be construed as preferred or advantageous over other embodiments.

Furthermore, in the following detailed description, numerous specific details are set forth in order to provide a better understanding of the present disclosure. It will be understood by those skilled in the art that the present disclosure may be practiced without some of these specific details. In some instances, methods, means, elements and circuits that are well known to those skilled in the art have not been described in detail so as not to obscure the present disclosure.

At present, in distributed aggregation and EVPN integrated networking, traffic forwarded from a virtual machine side through a DR port of distributed aggregation is forwarded through equivalent load sharing. As shown in fig. 2, when a service data packet forwarded from VM a to VM E goes out of an interface, it is forwarded to VTEP B and VTEP C according to load sharing of service balancing.

However, the current solution is based on the fact that VTEP B and VTEP C belong to switching devices with comparable forwarding performance. If VTEP B and VTEP C belong to devices with large difference in forwarding performance, it is not appropriate to use balanced load sharing, which may result in a situation where a device with poor forwarding performance is busy and a device with high forwarding performance is idle.

In view of the above problems, the present disclosure proposes a scheme for dynamically distinguishing traffic by constructing a virtual MAC of a gateway under an integrated networking based on distributed aggregation and EVPN.

In a distributed aggregation system, different MAC addresses are set as actual MAC addresses for all EVPN gateways (e.g., VTEPs). During load sharing, after hash is performed on one VTEP in the distributed aggregation system, the VTEP may return an actual MAC address table entry for packet forwarding to the VM according to a forwarding performance ratio of two VTEPs in the distributed aggregation system. By the scheme, the load sharing of the service flow of the virtual machine can be dynamically determined according to the forwarding performance of the VTEP, so that two VTEPs of the distributed aggregation system can be efficiently used.

Fig. 3 shows a flow chart of a traffic forwarding method according to an embodiment of the present disclosure. As shown in fig. 3, the method is applied to a first VTEP in a distributed aggregation system, where the distributed aggregation system includes at least two VTEPs, and different VTEPs are configured with different MAC addresses, and the method includes:

step 101, acquiring the forwarding performance of each VTEP in the distributed aggregation system under the condition that a virtual machine connected with the distributed aggregation system is detected to be online;

and step 102, returning the MAC address of the VTEP for load sharing to the virtual machine according to the current traffic and forwarding performance of each VTEP.

In a possible implementation manner, the distributed aggregation system includes a first VTEP and a second VTEP, and the first VTEP may obtain forwarding performance and current traffic of the first VTEP and the second VTEP through protocol interaction of the EVPN. The second VTEP can acquire the forwarding performance and the current traffic of the second VTEP and the first VTEP through the protocol interaction of the EVPN. As shown in fig. 4, step 102 includes:

step 201, determining a VTEP for load sharing according to the current traffic and forwarding performance of a first VTEP and a second VTEP in a distributed aggregation system;

step 202, if the VTEP for load sharing is the first VTEP, returning the MAC address of the first VTEP to the virtual machine.

In one possible implementation, as shown in fig. 4, step 102 further includes:

and 203, if the VTEP for load sharing is a second VTEP, returning the MAC address of the second VTEP to the virtual machine.

In a distributed aggregation system, the forwarding performance of each VTEP is related to the respective forwarding performance, e.g., the maximum amount of traffic that can be forwarded. The forwarding performance of the VSI (virtual switching instance) is set to a certain value at the first VTEP and the second VTEP. For example, if the forwarding performance of the second VTEP is 2 times that of the first VTEP, the forwarding performance of the VSI may be set to 2 on the second VTEP and 1 on the first VTEP. The first VTEP acquires the forwarding performance of the first VTEP and the second VTEP through the protocol interaction of the EVPN. And, the first VTEP may also obtain current traffic of itself and the second VTEP, for example, the number of currently received traffic, through EVPN protocol interaction. On the first VTEP, if the VTEP used for load sharing is determined to be the first VTEP according to the current traffic volume and forwarding performance of the first VTEP and the second VTEP, the first VTEP returns the MAC address of the first VTEP to the virtual machine. On the first VTEP, if the VTEP used for load sharing is determined to be the second VTEP, the first VTEP returns the MAC address of the second VTEP to the virtual machine.

In one possible implementation, step 201 includes:

if the ratio of the current traffic of the first VTEP and the second VTEP in the distributed aggregation system is smaller than a first threshold value, the VTEP used for load sharing is the first VTEP, and the first threshold value is determined according to the forwarding performance of the first VTEP and the second VTEP. For example, the first threshold is a ratio of the forwarding performance of the first VTEP to the forwarding performance of the second VTEP.

If the ratio of the current traffic of the first VTEP to the second VTEP is smaller than the first threshold, it shows that the first VTEP is lighter in load relative to the forwarding performance which can be carried by the first VTEP and the second VTEP is heavier in load relative to the forwarding performance which can be carried by the second VTEP. Therefore, the first VTEP may bear more traffic and is selected to share the current load.

In a possible implementation manner, step 201 further includes:

and if the ratio of the current traffic of the first VTEP to the second VTEP in the distributed aggregation system is greater than or equal to a first threshold, the VTEP for load sharing is the MAC address of the second VTEP. At this point, the first VTEP may return the MAC address of the second VTEP to the virtual machine, such that the virtual machine modifies the next hop in the forwarding entry from the first VTEP to the second VTEP.

If the ratio of the current traffic of the first VTEP to the current traffic of the second VTEP is larger than or equal to the first threshold, the first VTEP is heavier in load relative to the self-bearable forwarding performance, and the second VTEP is lighter in load relative to the self-bearable forwarding performance. Therefore, the second VTEP may bear more traffic and is selected to share the current load.

It should be noted that the above burden-sharing implementation is only an example, and not a limitation. For example, when the ratio of the current traffic volumes of the first VTEP and the second VTEP in the distributed aggregation system is equal to the first threshold, it may be stated which VTEP shares the current load, and the load of the first VTEP and the second VTEP is balanced. Therefore, at this time, the first VTEP may return the MAC address of the first VTEP or the second VTEP to the virtual machine at random, instead of directly returning the MAC address of the second VTEP to the virtual machine.

In this embodiment, each VTEP may acquire forwarding performance of other VTEPs in the distributed aggregation system through EVPN protocol interaction, so as to determine a ratio of its own forwarding performance in the distributed aggregation system.

As an example, in the first VTEP, the ratio 1/2 of the forwarding performance of the first VTEP and the second VTEP may be used as the first threshold. The second threshold value may be 2/1, which is the ratio of the forwarding performance of the second VTEP to the first VTEP. The first threshold and the second threshold may be used for subsequent decisions on which VTEP to use for load sharing.

When determining which VTEP is used for load sharing, the first VTEP in the distributed aggregation system may determine its current traffic volume according to the number of received services, and obtain the current traffic volume of the second VTEP in the distributed aggregation system.

In the first VTEP, if the ratio of the current traffic of the first VTEP to the second VTEP is calculated to be smaller than the first threshold, the first VTEP may be considered to be lightly loaded with respect to forwarding performance that the first VTEP can carry, and therefore, the MAC address of the first VTEP may be replied to the VM. If the ratio of the current traffic of the first VTEP to the second VTEP is calculated to be larger than or equal to the first threshold value, the second VTEP can be considered to be lighter in load relative to the forwarding performance which can be carried by the second VTEP, and therefore, the MAC address of the second VTEP can be answered to the VM.

In the second VTEP, if the ratio of the current traffic of the second VTEP to the first VTEP is calculated to be smaller than the second threshold, the second VTEP may be considered to be less loaded with respect to the forwarding performance that the second VTEP can carry, and therefore, the MAC address of the second VTEP may be replied to the VM. If the ratio of the current traffic of the second VTEP to the first VTEP is calculated to be greater than or equal to the second threshold value, the first VTEP can be considered to be lightly loaded relative to the forwarding performance which can be carried by the first VTEP, and therefore, the MAC address of the first VTEP can be answered to the VM.

In the distributed aggregation system, different MAC addresses can be set for different VTEPs as actual MAC addresses, and the load sharing of the virtual machine service flow is dynamically determined according to the current service volume and the forwarding performance of two VTEP devices in the distributed aggregation system, so that all the VTEPs in the distributed aggregation system can be efficiently used.

Fig. 5 is a schematic diagram illustrating an application scenario of a traffic forwarding method according to an embodiment of the present disclosure. As shown in fig. 5, the traffic forwarding method includes the following steps:

step 1, accessing a VM A to a distributed aggregation system, wherein the distributed aggregation system comprises two VTEP devices, namely VTEP B and VTEP C. The actual IP address configured on the VTEP B is 1.1.1.1, the EVPN IP address of the virtual machine is 3.3.3.3, and the MAC address of the gateway is 1-1-1. The actual IP address configured on the VTEP C is 2.2.2.2, the virtual machine EVPN IP address is 3.3.3.3, and the MAC address of the gateway is 2-2-2.

And 2, carrying out proportional configuration setting of forwarding performance on the VTEP B and the VTEP C. For example, if the forwarding performance of VTEP C is double that of VTEP B, the forwarding performance of VSI is set to 2 above VTEP C; on top of VTEP B, the forwarding performance of VSI (Virtual Switch Interface) is set to 1. The VTEP B and the VTEP C expand the protocol interaction of the EVPN, so that the VTEP B and the VTEP C know the forwarding performance proportion of each other mutually, and the occupation proportion of the forwarding performance occupied by the VTEP B and the VTEP C in distributed aggregation is determined.

Step 3, the VTEP B and the VTEP C interact the MAC address of the opposite gateway and record the MAC address locally.

And 4, after the virtual machine service is subjected to distributed aggregation from a DR port (distributed aggregation interface) of the distributed aggregation system, determining whether to use a gateway MAC address of the VTEP B or the VTEP C for response according to the number of the services received by the two current devices by the VTEP B and the VTEP C.

After step 5, e.g. VM a goes online, it is hashed (hash) onto VTEP B. If the ratio of the number of the VXLAN service online data, VTEP B and the number of the service received by VTEP C on the VTEP B does not reach 1:2 at the moment, then the confirmation VM A can be hash-forwarded on the VTEP B. The VTEP B replies a gateway MAC address 1-1-1 of the VTEP B to the VM A equipment;

and 6, if the VM A is online, the hash is carried out on the VTEP B. If the ratio of the number of the services received by the VTEP B to the number of the services received by the VTEP C exceeds 1:2 according to the online data of the VXLAN service on the VTEP B, the VM A is confirmed to be hash to the VTEP C for forwarding. VTEP B replies to VM A with the gateway MAC address 2-2-2 of the peer VTEP C. At the same time, the forwarding next hop for VM A100.1.1.2 is switched from VTEP B to VTEP C on top of the forwarding entry for VM a.

In comparison, if VM A is online, it is hashed (hash) to VTEP C. And if the ratio of the number of the VXLAN service on the VTEP C to the number of the service received by the VTEP B does not reach 2:1 at the moment, the VM A can be hash-forwarded to the VTEP C. The VTEP C replies with its own gateway MAC address 2-2-2 to the VM a device. If the ratio of VTEP C to the number of traffic received by VTEP B exceeds 2:1, then the acknowledgment VM A may be hash-forwarded onto VTEP B. VTEP C replies the gateway MAC address 1-1-1 of the peer VTEP B to VM A device. At the same time, the forwarding next hop for VM A100.1.1.2 is switched from VTEP C to VTEPB on top of it in the forwarding entry of VM a.

The direct effect that this disclosure brings: in the distributed aggregation system, different MAC addresses are set for all EVPN gateways to serve as actual MAC addresses, hash is carried out according to the forwarding performance ratio of two VTEP devices in the distributed aggregation system, MAC address table items forwarded by device messages can be used for dynamically determining load sharing of service flow of a virtual machine according to the forwarding performance of the VTEP devices, and therefore the two VTEPs of the distributed aggregation system can be efficiently used.

According to the method, in the distributed aggregation system, the service flow is forwarded in proportion according to the forwarding performance of each VTEP, different MAC addresses are used for indicating that different VTEPs respond to different virtual machine services, and the virtual machine services can select correct EVPN gateway communication.

Fig. 6 shows a block diagram of a traffic forwarding apparatus according to an embodiment of the present disclosure. As shown in fig. 6, the traffic forwarding apparatus is applied to a first VTEP in a distributed aggregation system, where the distributed aggregation system includes at least two VTEPs, and different VTEPs are configured with different MAC addresses, and the apparatus includes:

an obtaining module 41, configured to obtain forwarding performance of each VTEP in the distributed aggregation system when a virtual machine connected to the obtaining module is detected to be online;

and a returning module 43, configured to return the MAC address of the VTEP for load sharing to the virtual machine according to the current traffic volume and forwarding performance of each VTEP.

In one possible implementation, as shown in fig. 5, the return module 43 includes:

a comparison submodule 51, configured to determine, according to current traffic and forwarding performance of a first VTEP and a second VTEP in a distributed aggregation system, a VTEP used for load sharing;

the first returning submodule 53 is configured to return the MAC address of the first VTEP to the virtual machine if the VTEP used for load sharing is the first VTEP.

In a possible implementation manner, the return module 43 further includes:

and a second returning submodule 55, configured to return the MAC address of the second VTEP to the virtual machine if the VTEP used for load sharing is the second VTEP.

In a possible implementation manner, the comparing sub-module 51 is further configured to determine that the VTEP for load sharing is the first VTEP if a ratio of current traffic volumes of the first VTEP and the second VTEP in the distributed aggregation system is smaller than a first threshold, where the first threshold is determined according to forwarding performances of the first VTEP and the second VTEP. For example, the first threshold is a ratio of the forwarding performance of the first VTEP to the forwarding performance of the second VTEP.

In a possible implementation manner, the comparing sub-module 51 is further configured to determine that the VTEP used for load sharing is the MAC address of the second VTEP if the ratio of the current traffic volumes of the first VTEP and the second VTEP in the distributed aggregation system is greater than or equal to the first threshold.

The second returning sub-module 55 is further configured to, when the comparing sub-module 51 determines that the VTEP for load sharing is the second VTEP, return the MAC address of the second VTEP to the virtual machine, so that the virtual machine modifies the next hop in the forwarding table entry from the first VTEP to the second VTEP.

With regard to the apparatus in the above-described embodiment, the specific manner in which each module performs the operation has been described in detail in the embodiment related to the method, and will not be elaborated here.

Fig. 8 is a schematic diagram illustrating a traffic forwarding device, according to an example embodiment. Referring to fig. 8, the apparatus 900 may include a processor 901, a machine-readable storage medium 902 having stored thereon machine-executable instructions. The processor 901 and the machine-readable storage medium 902 may communicate via a system bus 903. Also, the processor 901 performs the traffic forwarding method described above by reading machine executable instructions in the machine readable storage medium 902 corresponding to the traffic forwarding logic.

The machine-readable storage medium 902 referred to herein may be any electronic, magnetic, optical, or other physical storage device that can contain or store information such as executable instructions, data, and the like. For example, the machine-readable storage medium may be: a RAM (random Access Memory), a volatile Memory, a non-volatile Memory, a flash Memory, a storage drive (e.g., a hard drive), a solid state drive, any type of storage disk (e.g., an optical disk, a dvd, etc.), or similar storage medium, or a combination thereof.

Having described embodiments of the present disclosure, the foregoing description is intended to be exemplary, not exhaustive, and not limited to the disclosed embodiments. Many modifications and variations will be apparent to those of ordinary skill in the art without departing from the scope and spirit of the described embodiments. The terms used herein were chosen in order to best explain the principles of the embodiments, the practical application, or technical improvements to the techniques in the marketplace, or to enable others of ordinary skill in the art to understand the embodiments disclosed herein.

Claims

1. A traffic forwarding method is applied to a first VTEP in a distributed aggregation system, the distributed aggregation system comprises at least two VTEPs, different VTEPs are configured with different MAC addresses, and the method comprises the following steps:

2. The method according to claim 1, wherein returning the MAC address of the VTEP for load sharing to the virtual machine according to the current traffic volume and forwarding performance of each VTEP comprises:

determining a VTEP for load sharing according to the current traffic and forwarding performance of a first VTEP and a second VTEP in a distributed aggregation system;

and if the VTEP for load sharing is the first VTEP, returning the MAC address of the first VTEP to the virtual machine.

3. The method according to claim 2, wherein the MAC address of the VTEP used for load sharing is returned to the virtual machine according to the current traffic volume and forwarding performance of each VTEP, further comprising:

and if the VTEP for load sharing is a second VTEP, returning the MAC address of the second VTEP to the virtual machine.

4. The method according to claim 2 or 3, wherein determining the VTEP for load sharing according to the current traffic volume and forwarding performance of the first VTEP and the second VTEP in the distributed aggregation system comprises:

if the ratio of the current traffic of the first VTEP and the second VTEP in the distributed aggregation system is smaller than a first threshold value, the VTEP used for load sharing is the first VTEP, and the first threshold value is determined according to the forwarding performance of the first VTEP and the second VTEP.

5. The method according to claim 2 or 3, wherein determining the VTEP for load sharing according to the current traffic volume and forwarding performance of the first VTEP and the second VTEP in the distributed aggregation system comprises:

and if the ratio of the current traffic of the first VTEP to the second VTEP in the distributed aggregation system is greater than or equal to a first threshold, the VTEP for load sharing is the MAC address of the second VTEP.

6. A traffic forwarding apparatus applied to a first VTEP in a distributed aggregation system, the distributed aggregation system including at least two VTEPs, different VTEPs being configured with different MAC addresses, the apparatus comprising:

7. The apparatus of claim 6, wherein the return module comprises:

the comparison submodule is used for determining the VTEP for load sharing according to the current traffic and forwarding performance of the first VTEP and the second VTEP in the distributed aggregation system;

and the first returning submodule is used for returning the MAC address of the first VTEP to the virtual machine if the VTEP for load sharing is the first VTEP.

8. The apparatus of claim 7, wherein the return module further comprises:

and the second returning submodule is used for returning the MAC address of the second VTEP to the virtual machine if the VTEP for load sharing is the second VTEP.

9. The apparatus according to claim 7 or 8, wherein the comparing sub-module is further configured to determine the VTEP for load sharing to be the first VTEP if a ratio of current traffic volumes of the first VTEP and the second VTEP in the distributed aggregation system is smaller than a first threshold, where the first threshold is determined according to forwarding performance of the first VTEP and the second VTEP.

10. The apparatus of claim 7 or 8, wherein the comparing sub-module is further configured to determine that the VTEP for load sharing is the MAC address of the second VTEP if the ratio of the current traffic volumes of the first VTEP and the second VTEP in the distributed aggregation system is greater than or equal to the first threshold.