WO2022001287A1

WO2022001287A1 - Message processing method and device

Info

Publication number: WO2022001287A1
Application number: PCT/CN2021/087107
Authority: WO
Inventors: 王辉登
Original assignee: 华为技术有限公司
Priority date: 2020-07-03
Filing date: 2021-04-14
Publication date: 2022-01-06

Abstract

Disclosed in embodiments of the present application are a message processing method and device. When a first network device determines that a preset condition is satisfied, according to a first message, a second message comprising a hash value and content related to the first message is generated, the hash value being obtained after hash calculation according to message feature information of the first message, and the first network device sends the second message to a second network device. In this way, because the hash value is obtained on the basis of an inner layer message feature of the message, the diversity of the message is presented, and to a certain extent, the uniform distribution of hash values is ensured. Therefore, carrying the hash value in the message to be sent and sending to another network device can ensure that the another network device implements a load balancing effect on the basis of the hash value.

Description

Method and device for processing message

This application claims the priority of the Chinese patent application with the application number 202010631135.3 and the invention titled "Method, Device and Network System for Processing Data" filed on July 3, 2020, the entire contents of which are incorporated into this application by reference; In addition, this application also claims the priority of the Chinese patent application submitted to the State Intellectual Property Office of China on September 28, 2020, the application number is 202011041940.7, and the application name is "a method and apparatus for message processing", the entire content of which is approved by Reference is incorporated in this application.

technical field

The present application relates to the field of communication technologies, and in particular, to a method and device for processing messages.

Background technique

In the network, through the forwarding of packets between network devices, users are provided with corresponding business services. When a network device forwards packets, load balancing is usually performed in a uniform hash method. That is, the hash value of each packet to be forwarded is calculated, and different packets to be forwarded are mapped to the outgoing interface of the network device based on the hash value. on the processing unit of a network device.

However, in some scenarios, the current hash makes it impossible for network devices to process the packets to be forwarded in a balanced manner, resulting in low resource utilization.

SUMMARY OF THE INVENTION

Based on this, the embodiments of the present application provide a packet processing method and device, so as to improve load balancing efficiency.

In a first aspect, an embodiment of the present application provides a method for processing a message. In the method, when a first network device determines that a preset condition is met, the first network device can obtain a message feature of the first message The hash value of the information is generated, and a second packet including the hash value and content related to the first packet is generated according to the first packet, so that the first network device can use the second packet carrying the hash value. The message is sent to the second network device. In this way, since the hash value is calculated based on the inner packet characteristics of the packet, it reflects the diversity of the packet and ensures the uniform distribution of the hash value to a certain extent. Therefore, the hash value is carried in the Sending the to-be-sent packet to other network devices can ensure that other network devices can achieve a better load balancing effect based on the hash value; moreover, when one network device performs a hash calculation, subsequent network devices will be loaded from the network device when there is a load balancing requirement. By directly reading the hash value in the message, high-quality load balancing can be performed, which saves the computing resources of each network device, improves the forwarding efficiency of the message, and thus improves the forwarding performance of the network device.

In some implementation manners, the preset condition may be that the length of the first packet is greater than a maximum transmission unit (English: maximum transmission unit, MTU for short) of the first network device. Then, the content related to the first packet includes fragmented packets of the first packet.

As an example, the header of the second packet may be used to carry the hash value.

As another example, the second packet may also include an IP extension header, and the Option field in the IP extension header may be used to carry the hash value. For example, the IP extension header of the second packet includes a Reserved field, where the Reserved field is used to carry the hash value.

In this way, in the scenario of packet fragmentation, each fragmented packet includes a hash value that preserves the diversity of packet characteristics, which can make the load in the network more balanced; moreover, it can also ensure that the same packet The fragmented packets of the text are distributed to the same link or the same processing unit when hashing is performed uniformly, so as to realize the orderly forwarding of the fragmented packets.

In other implementation manners, the preset condition may also be that the first network device is the head end point of the first tunnel. Then, the content related to the first packet includes the entire content of the first packet.

As an example, the second packet may further include the first tunnel header. For example, if the first tunnel is a GRE tunnel, then the second packet is a packet obtained by encapsulating a first generic routing encapsulation (English: generic routing encapsulation, GRE for short) header on the first packet. The GRE header is the first tunnel header in the second packet; if the first tunnel is an Internet protocol security (English: internet protocol security, IPSec for short) tunnel, then the second packet is the first packet encapsulated on the first packet. In the packet obtained after the IPSec header, the first IPSec header is the first tunnel header in the second packet.

As an example, the hash value may be carried in the first tunnel header. For example, the first tunnel header includes a Reserved field, which is used to carry the hash value.

Specifically, the first tunnel may be a virtual private network (English: virtual private network, VPN for short) tunnel. The types of VPN tunnels include but are not limited to GRE tunnels, IPSec tunnels, Virtual Extensible Local Area Network (English: Virtual Extensible Local Area Network, referred to as: VXLAN) tunnels, Layer 2 Tunneling Protocol (English: Layer 2 Tunneling Protocol, referred to as: L2TP) tunnels, Layer 3 Tunneling Protocol (English: Layer 3 Tunneling Protocol Version 3, referred to as: L2TPv3) tunnel, Multi-Protocol Label Switch (English: Multi-Protocol Label Switch, referred to as: MPLS) tunnel, IPv6 Over IPv4 tunnel, IPv4 Over IPv6 tunnel, etc.

In this way, each packet in the encapsulation tunnel scenario includes a hash value that preserves the diversity of packet characteristics, which provides convenience for subsequent load balancing when each network device forwards the second packet, and, Since the hash value of the first packet reflects the diversity of the inner layer features of the first packet, the load balancing effect in the network is guaranteed.

In a second aspect, an embodiment of the present application further provides a method for processing a packet. In the method, when the second network device receives a second packet sent by the first network device, the second packet includes a hash value and content related to the first packet (that is, all or part of the content of the first packet), the hash value is calculated by the first network device or other network devices based on the packet feature information of the first packet; In this way, the second network device can process the second packet based on the hash value.

As an example, if the second network device includes multiple processing units, and the second network device processes the second packet based on the hash value, the process may include: the second network device processes the second packet based on the hash value. Assigned to a first processing unit of the second network device, the first processing unit corresponding to the hash value. Wherein, the processing unit may be, for example, a central processing unit (English: Central Processing Unit, referred to as: CPU), a network processor (English: Network Processor, referred to as: NP) or a forwarding chip and other units with processing and forwarding functions.

As another example, if the second network device determines that there are at least two outgoing interfaces for forwarding the second packet, at this time, the second network device processes the second packet based on the hash value, and may also include: the first The second network device forwards the second packet from the first interface of the second network device based on the hash value, where the first interface corresponds to the hash value.

In some possible implementations, if the second network device is the head end of the second tunnel, the method may further include that the second network device encapsulates the second tunnel header on the second packet, and carries the hash value Process at the second tunnel head. Specifically, the method in this embodiment of the present application may further include: the second network device determines that the second network device is the head end point of the second tunnel, then, obtaining the hash value from the second packet; then, second The network device encapsulates the second tunnel header corresponding to the second tunnel for the second message, and obtains a third message, where the second tunnel header of the third message includes a hash value; thus, the second network device can The three packets are sent to the third network device, so that the third network device processes the third packet based on the hash value. In this way, the hash value is always included in each tunnel header of the packet. No matter which tunnel end point is reached first, the current tunnel header of the packet can be guaranteed to carry the hash value, so as to achieve better load balancing. The effect is guaranteed.

In some other possible implementation manners, if the second network device determines that the length of the fragmented second packet is greater than the MTU of the second network device, the method may further include that the second network device responds to the second packet The process of fragmenting and carrying the hash value in the fragmented packet of the second packet. Specifically, the embodiment of the present application may further include: the second network device determines that the length of the second packet is greater than the MTU of the second network device, and then processes the second packet into at least two fragmented packets; then, The second network device adds a hash value to each fragmented packet to obtain at least two third packets; thus, the second network device sends at least two of the third packets to the third network device, and the second network device sends at least two of the third packets to the third network device. The three network devices may process at least two of the third packets based on the hash value. In this way, each fragmented packet carries the hash value corresponding to the packet feature information of the packet before the fragmentation, so that each fragmented packet includes a hash value that preserves the diversity of packet characteristics, which can It makes the load in the network more balanced; moreover, it can also ensure that the fragmented packets of the same packet are shared on the same link or the same processing unit when hashing is performed evenly, so as to realize the order of the fragmented packets. Forward.

In a third aspect, an embodiment of the present application provides a first network device, including a transceiver unit and a processing unit. The transceiving unit is configured to perform the transceiving operation in the method provided in the first aspect or any possible implementation manner of the first aspect; the processing unit is configured to perform the foregoing first aspect or any possible implementation manner of the first aspect Other operations in the provided method other than the transceiving operation. For example: when the first network device executes the method of the first aspect, the transceiver unit is configured to send a second packet to the second network device; the processing unit is configured to generate a second packet according to the first packet Second message.

In a fourth aspect, an embodiment of the present application further provides a second network device, where the second network device includes a transceiver unit and a processing unit. The transceiving unit is configured to perform the transceiving operation in the method provided in the second aspect or any possible implementation manner of the second aspect; the processing unit is configured to perform the foregoing second aspect or any possible implementation manner of the second aspect Other operations in the provided method other than the transceiving operation. For example: when the second network device executes the method of the second aspect, the transceiver unit is configured to receive the second packet sent by the first network device; the processing unit is configured to, according to the hash value, Process the second packet.

In a fifth aspect, an embodiment of the present application further provides a first network device, where the first network device includes a memory and a processor. Wherein, the memory includes a computer program or an instruction; a processor in communication with the memory is used to execute the computer program or instruction, so that the first network device is used to execute the above first aspect or any possible implementation of the first aspect method provided.

In some embodiments, the first network device may also not include a memory, and the processor obtains computer programs or instructions from an external memory or cloud storage, and executes the computer programs or instructions to enable the first network device to use The method provided for performing the above first aspect or any possible implementation manner of the first aspect.

In a sixth aspect, an embodiment of the present application further provides a second network device, where the second network device includes a memory and a processor. Wherein, the memory includes a computer program or an instruction; a processor in communication with the memory is used to execute the computer program or instruction, so that the second network device is used to execute the above second aspect or any possible implementation of the second aspect method provided.

In some embodiments, the second network device may also not include a memory, and the processor obtains computer programs or instructions from an external memory or cloud storage, and executes the computer programs or instructions to enable the second network device to use The method provided for performing the above second aspect or any possible implementation manner of the second aspect.

In a seventh aspect, an embodiment of the present application further provides a computer-readable storage medium, where a computer program or instruction is stored in the computer-readable storage medium, and when it runs on a computer, the computer enables the computer to perform the above first aspect , any possible implementation manner of the first aspect, the second aspect, or a method provided by any possible implementation manner of the second aspect.

In an eighth aspect, the embodiments of the present application also provide a computer program product, including a computer program or computer-readable instructions, when the computer program or the computer-readable instructions are run on a computer, the computer is made to execute the aforementioned first aspect, A method provided by any possible implementation manner of the first aspect, the second aspect, or any possible implementation manner of the second aspect.

In a ninth aspect, an embodiment of the present application further provides a communication system, where the communication system includes the first network device provided in the third aspect or the fourth aspect, and/or the fifth aspect or the sixth aspect. the second network device.

Description of drawings

1 is a schematic diagram of a network system framework involved in an application scenario in an embodiment of the present application;

FIG. 2 is a signaling flowchart of a method 100 for processing a packet in an embodiment of the present application;

3a is a schematic diagram of a format of a fragmented packet in a packet fragmentation scenario according to an embodiment of the present application;

FIG. 3b is a schematic diagram of the format of another fragmented packet in a packet fragmentation scenario according to an embodiment of the present application;

4a is a schematic diagram of a format of a second packet in an encapsulation tunnel scenario in an embodiment of the present application;

4b is a schematic diagram of another format of a second packet in an encapsulation tunnel scenario according to an embodiment of the present application;

4c is a schematic diagram of another format of a second packet in an encapsulation tunnel scenario according to an embodiment of the present application;

FIG. 4d is a schematic diagram of another format of a second packet in an encapsulation tunnel scenario according to an embodiment of the present application;

FIG. 5 is a schematic structural diagram of a first network device 500 in an embodiment of the present application;

FIG. 6 is a schematic structural diagram of a second network device 600 in an embodiment of the present application;

FIG. 7 is a schematic structural diagram of a first network device 700 in an embodiment of the present application;

FIG. 8 is a schematic structural diagram of a second network device 800 in an embodiment of the present application;

FIG. 9 is a schematic structural diagram of a first network device 900 in an embodiment of the present application;

FIG. 10 is a schematic structural diagram of a second network device 1000 in an embodiment of the present application;

FIG. 11 is a schematic structural diagram of a communication system 1100 according to an embodiment of the present application.

detailed description

Load balancing of forwarding packets in the network is an important means to realize the rational use of resources and ensure that there is no congestion. At present, load balancing is usually performed in a hash uniform manner, that is, based on the hash values of the packets to be forwarded, the packets to be forwarded are evenly distributed to multiple outgoing interfaces of the network device or multiple processing units of the network device. . In specific implementation, after each network device receives a message, it first obtains message feature information from the message (for example, the quintuple of the message: source Internet Protocol (English: Internet Protocol, referred to as: IP) address, destination IP address, source port number, destination port number and protocol number), and then perform hash calculation on the packet feature information to obtain the hash value corresponding to the packet; thus, if the network device includes multiple processing units, then the The packet is sent to the target processing unit corresponding to the hash value among the multiple processing units; if it is determined based on the forwarding table that the next hop information of the packet indicates that there are multiple outgoing interfaces, then the packet is sent from the multiple The target outgoing interface corresponding to the hash value among the outgoing interfaces is forwarded to another network device.

In the uniform hash method, whether the load balance in the network can be effectively guaranteed depends on the diversity of the packet feature information on which the hash calculation is based. However, in the current uniform hashing method, the obtained packet feature information is often not comprehensive enough, and it is impossible or difficult to preserve the diversity of the packet, resulting in a relatively single hash value obtained after hashing the packet feature information. It is not dispersed enough to effectively achieve load balancing in the network. For example, some processing units in a network device are congested and some processing units are idle. For example, some interfaces of a network device are congested and some interfaces are idle, which greatly reduces network performance. resource utilization in .

For example, for the scenario of packet fragmentation, a packet is divided into multiple fragmented packets, since only one fragmented packet includes transmission control protocol (English: transmission control protocol, referred to as: TCP) or user data message protocol (English: user datagram protocol, referred to as: UDP) information, and other fragmented packets do not include TCP or UDP information. In order to ensure that the same load balancing operation is performed on multiple fragmented packets of a packet, the source IP address, destination IP address and protocol number included in each fragmented packet are used as the packet feature information, and the hash calculation is performed to obtain the same load balancing operation. This ensures that all fragmented packets of the packet are allocated to the same processing unit or outgoing interface of the network device based on the same hash value. However, hash calculation based on triples loses the diversity of packets to a certain extent. For example, the five-tuple (ie, source IP address, destination IP address, source port number, destination port number and protocol number) are different but For multiple packets with the same triple (ie source IP address, destination IP address and protocol number), by fragmenting and hashing the triples of the fragmented packets, the same hash value will be obtained. , all the fragmented packets corresponding to the multiple packets will be allocated to the same processing unit and the same outbound interface in each network device, which affects the effect of load balancing.

In the virtual private network (English: virtual private network, referred to as: VPN) scenario, the packet to be forwarded has been encapsulated with the packet header corresponding to the VPN tunnel. When the network device receives the packet including the packet header corresponding to the VPN tunnel, it cannot The quintuple (also known as the inner quintuple) before the VPN tunnel is not encapsulated in the packet can only be obtained based on the quintuple (also known as the outer quintuple) reflected in the packet header corresponding to the VPN tunnel encapsulated in the packet. quintuple) to hash uniformly. For example, when the packet to be forwarded is encapsulated with a general routing encapsulation (English: generic routing encapsulation, referred to as: GRE) tunnel, in one case, the network device can obtain the quintuple of the packet based on the GRE header of the packet, and Based on the quintuple, the hash is uniform, but the diversity of the inner quintuple is lost, and the load balancing effect is poor; in another case, the network device can also read the inner quintuple of the packet, based on the The inner layer quintuple is hashed evenly. However, to read the inner layer quintuple of the packet, it is necessary to calculate the length of the GRE header, and accurately strip the GRE header based on the length of the GRE header to obtain the inner layer packet, which is a complicated and time-consuming process. . Another example: when the packet to be forwarded is encapsulated with an Internet protocol security (English: internet protocol security, referred to as: IPSec) tunnel, the inner packet is encrypted, and the inner quintuple is not carried in plaintext in the packet. In the outer IPSec header, the network device cannot obtain it and hash it evenly based on the inner quintuple, which makes the load balancing effect poor. Moreover, for many scenarios where multiple layers of VPN tunnels are encapsulated, the inner quintuple of the packet to be forwarded is more difficult to obtain or even impossible to obtain. Therefore, for multiple packets encapsulated in the same VPN tunnel, the same hash value will be obtained, so that the multiple packets will be allocated to the same processing unit in each network device that the VPN tunnel passes through. and the same outgoing interface, the feature diversity of the inner layer of the packet (that is, the part before the VPN tunnel is not encapsulated) is lost, which affects the effect of load balancing.

Moreover, in the above-mentioned uniform hashing method, each network device that forwards the packet needs to perform a hash calculation after receiving the packet, which affects the forwarding rate of the packet, thereby reducing the forwarding performance of the network device.

Based on this, an embodiment of the present application provides a packet processing method. When a packet is about to enter a tunnel or is about to be fragmented, a network device obtains packet feature information of the packet (for example, the content of the packet). Layer 5-tuple), perform hash calculation based on the feature information of the message, obtain the hash value corresponding to the message, and compare the hash value with the content related to the message (such as the entire content of the message or the In this way, the network device that receives the packet can perform load balancing operations on the packet based on the hash value in the received content, because the hash value is based on the It is obtained by calculating the inner layer packet characteristics of the message, which reflects the diversity of the message, ensures the uniform distribution of the hash value to a certain extent, and achieves a better load balancing effect; moreover, when a network device performs a hash calculation, Subsequent network devices can read the hash value from the packet for load balancing when there is a load balancing requirement, which saves the computing resources of each network device, improves packet forwarding efficiency, and improves the forwarding performance of the network device.

Taking the scenario shown in FIG. 1 as an example, the scenario includes network devices 101 to 104, wherein the network device 101 includes a processing unit 11, a processing unit 12, an interface A1, an interface A2 and an interface A3, and the network device 102 includes processing Unit 21, interface B1 and interface B2, network device 103 includes processing unit 31, interface C1 and interface C2, network device 104 includes processing unit 41, processing unit 42, processing unit 43, processing unit 44, interface D1, interface D2 and interface D3. Assuming that the network device 101 is the head end of the GRE tunnel 1, when the processing unit 11 of the network device 101 receives the packet 1 and the packet 2 through the interface A1, it determines that the network device 101 is the head end of the tunnel 1, and the network device 101's The processing unit (such as the processor 11 and/or the processor 12) calculates the hash value 1 corresponding to the packet 1, and encapsulates the GRE header 1 for the packet 1 to obtain the packet 1', and the GRE header 1 carries the hash value 1; Calculate the hash value 2 corresponding to the packet 2, and encapsulate the GRE header 2 for the packet 2 to obtain the packet 2', and the GRE header 2 carries the hash value 2. The network device 101 determines that the outgoing interface of the packet 1 is the interface A2 according to the hash value 1, and then forwards the packet 1' from the interface A2 to the network device 102, which is received by the interface B1 of the network device 102 and sent to the network through the interface B2. device 104; similarly, network device 101 determines that the outgoing interface of packet 2 is interface A3 according to hash value 2, and then forwards packet 2' from interface A3 to network device 103, where it is received by interface C1 of network device 103 and sent to the network device 103. Sent to network device 104 through interface C2. Next, when the interface D1 of the network device 104 receives the packet 1, it can determine to allocate the packet 1 to the processing unit 41 according to the hash value 1 of the packet 1, and the processing unit 41 determines that the outgoing interface of the packet 1' is an interface D2, thereby forwarding the packet 1 from the interface D2 to the subsequent network device; when the interface D1 of the network device 104 receives the packet 2', it can determine to assign the packet 2' to the processing unit 43 according to the hash value 2, and process the packet 2'. The unit 43 determines that the outgoing interfaces of the packet 2' are the interface D2 and the interface D3. At this time, the processing unit 43 continues to determine the outgoing interface that actually forwards the packet 2 as the interface D3 according to the hash value 2, so as to send the packet 2' from the interface to the interface D3. D3 is forwarded to subsequent network devices. In this way, effective load balancing in the network can be achieved, forwarding efficiency can be improved, and network forwarding performance can be improved.

In some embodiments, one or more of the

processing units

11, 12, 21, 31, 41, 42, 43, and 44 may be a central processing unit (English: Central Processing Unit, referred to as: CPU), a network processor ( English: Network Processor, abbreviation: NP) or forwarding chip and other units with processing and forwarding functions, the specific form is not limited.

In some embodiments, the hash value 1 may be carried in other parts of the packet 1', such as the IP extension header of the packet.

In the embodiments of this application, the network device and the node refer to the same meaning in this application, and can be used interchangeably. In this embodiment of the present application, a network device refers to a device with a forwarding function, which may specifically include, but is not limited to, a switch, a router, or a firewall.

A specific implementation manner of a packet processing method in an embodiment of the present application will be described in detail below with reference to the accompanying drawings.

FIG. 2 is a signaling flowchart of a packet processing method 100 in an embodiment of the present application. In one case, the method 100 can be applied to the scenario of packet fragmentation, to perform the packet fragmentation operation of the first network device and any network device (referred to as the second network device) after the first network device. The interaction between them introduces the embodiments of the present application; in another case, the method 100 can also be applied to a tunnel encapsulation scenario, where the head point (also referred to as the first network device) of the first tunnel and the first tunnel The interaction between any one network device (referred to as a second network device) other than the first network device introduces the embodiments of the present application. During specific implementation, referring to FIG. 2 , the method 100 may include, for example, the following S101 to S104:

S101, when a preset condition is met, the first network device generates a second packet according to the first packet, the second packet includes a hash value and content related to the first packet, and the hash value is based on The packet feature information of the first packet is calculated.

The message feature information of the first message refers to the information in the first message that can reflect the diversity of the inner layer information of the first message, and ensures the distribution of hash values obtained by performing hash calculation based on the feature information of the message. more uniform, so as to achieve better load balancing effect. For example, the packet feature information of the first packet may be an inner layer quintuple of the first packet, that is, the source IP address, destination IP address, source port number, destination port number, and protocol number of the first packet. The first network device may acquire the packet feature information of the first packet by reading the IP header of the first packet.

The preset condition is an indication condition that the network device needs to calculate the hash value corresponding to the packet. Each network device in the network is configured with the preset condition. When receiving the message, it is judged whether the preset condition is met. If the preset condition is met, the method 100 is executed to perform load balancing; if the preset condition is not met, the load balancing is performed. condition, the packet is forwarded normally. The preset conditions are different according to different scenarios.

In some possible implementation manners, for the scenario of packet fragmentation, the preset condition may be: the length of the first packet is greater than the maximum transmission unit (English: maximum transmission unit, MTU for short) of the first network device. Each network device in the network is configured with the MTU of the interface, which is used to limit the maximum data length sent by the interface. After receiving a packet, it is judged whether the total length of the packet is greater than the MTU. If the total length of the packet is greater than the MTU, Then, fragment the packet and ensure that the length of each fragmented packet after fragmentation is less than the MTU, and forward each fragmented packet; if the total length of the packet is not greater than the MTU, there is no need for the packet. Fragmentation can directly forward the packet. For example, assuming that the MTU on the first network device is 1500 bytes and the first network device receives the first packet, if the length of the first packet is 1457 bytes, since 1457 bytes is less than 1500 bytes, the first A network device does not process the first packet, but simply forwards it; if the length of the first packet is 4678 bytes, since 4678 bytes is greater than 1500 bytes, the first network device forwards the first packet. Fragmentation, for example, at least 4 fragmented packets can be obtained: fragmented packet 1, fragmented packet 2, fragmented packet 3, and fragmented packet 4. The length of each fragmented packet is less than 1500 words. Festival.

When specifically implemented, S101 may specifically include: S11, when the first network device determines that the first packet needs to be fragmented, obtain packet feature information of the first packet; S12, the first network device calculates based on the packet feature information the hash value corresponding to the first packet; S13, the first network device processes the first packet into at least two fragmented packets; S14, the first network device adds a hash to each fragmented packet value, obtain at least two second packets. In this case, in addition to the hash value, the second packet also includes content corresponding to a fragmented packet of the first packet (that is, the content related to the first packet referred to in S101).

For S12, the first network device may perform a hash calculation on the packet feature information of the first packet according to a preset hash algorithm to obtain a hash value. The preset hash algorithm may be, for example, an XOR algorithm or a cyclic redundancy check (English: cyclic redundancy check, CRC for short) 16 or CRC32. Since the hash value carried in the second packet and used to guide subsequent load balancing, the number of bits of the hash value needs to match the number of interfaces of the network device. For example, if the network device has 8 interfaces, then the hash value The value can be a 3-bit long binary number. Therefore, the network device may preset the preset length of the hash value, and then, S12 may include, for example: the first network device uses a preset hash algorithm to perform hash calculation on the packet feature information to obtain an initial hash value; The initial hash value obtains a target hash value with a preset length, and the target hash value is the hash value carried in the second message in the embodiment of the present application. In one case, the network device may intercept the preset length in the initial hash value, and use the intercepted bit value as the target hash value; in another case, the first network device may also interpret the initial hash value. Perform modulo (such as XOR) processing to obtain a target hash value with a preset length.

After the first packet is fragmented to obtain multiple fragmented packets, the hash value may be carried in the IP extension header of each fragmented packet.

As an example, for an IP packet, the IP extension header may include a copy flag (C), a type (Class) field, and an option (English: Option) field. The hash value can be carried in the Option field in the IP extension header of the fragmented packet. For example, the value of the unoccupied Option field can be newly defined and carried in the value (English: Value) field corresponding to the Option field. the hash value. The unoccupied values of the Option field include 25, 26, and 31.

For example: assuming that the first packet satisfies the preset conditions, the hash value calculated by the first network device based on the packet feature information of the first packet is a, and the value of 26, which is an unoccupied Option field, is used to carry the hash value a. . The first network device may divide the first packet into N (N is an integer greater than 1) fragmented packets, as shown in Figure 3a, each fragmented packet may include: an IP header, an IP extension header, and a net header. Payload (English: payload), in the IP extension header, C=1, indicating that the IP extension header is a fragmented packet copy extension header; Class field=0, indicating control; Option field=26, indicating that the corresponding Value field carries Value; the value of the Length (English: Length) field is used to indicate the length of the content corresponding to the Option field; Value field = hash value a.

As another example, for an IPv6 packet, the IP extension header in the fragmentation scenario may include: a next extension header (English: Next Header) field, a reserved (English: Reserved) field, and a fragmentation offset (English: Fragment Offset) field, reserved (Res) field and M flag bit and identification (English: Identification) field. Among them, the length of the first Reserved field is an 8-bit field, which is currently unused and can be used to carry a hash value.

For example, assuming that the first packet is an IPv6 packet and satisfies a preset condition, the hash value calculated by the first network device based on the packet feature information of the first packet is a, and the Reserved field is used to carry the hash value a. The first network device may divide the first packet into N (N is an integer greater than 1) fragmented packets, as shown in Figure 3b, each fragmented packet may include: an IP header, an IP extension header, and a net Payload (English: payload), in the IP extension header, the first Reserved field is equal to the hash value a.

In some embodiments, the hash value corresponding to the first packet may also be carried in each fragmented packet in other ways. In this way, in the scenario of packet fragmentation, the diversity of packet characteristics is preserved, which can make Load balancing in the network; moreover, it can also ensure that the fragmented packets of the same packet are shared on the same link or the same processing unit when hashing is performed evenly, so as to realize the orderly forwarding of the fragmented packets.

In some other possible implementation manners, for the scenario of tunnel encapsulation, the preset condition may be: the first network device is the head end point of the first tunnel. When the first packet arrives at the first network device, the first tunnel is not encapsulated. If the first network device checks the forwarding table and determines that the next hop information of the forwarding table entry corresponding to the first packet indicates that the outbound interface type is the first tunnel, then , the first network device can determine that it is the head end of the first tunnel, and the first packet is transmitted from the first network device through the first tunnel. At this time, the first packet is encapsulated with the first tunnel corresponding to the first tunnel. After the header, the encapsulated packet is forwarded; if the first network device checks the forwarding table and determines that the next hop information of the forwarding table entry corresponding to the first packet indicates that the outbound interface type is not the first tunnel, then the first network The device may determine that it is not the head-end point of the first tunnel, and in this case, it can directly forward the first packet without encapsulating the first tunnel for the first packet. For example, assuming that the next hop information of the forwarding entry corresponding to the first packet on the first network device indicates that the outbound interface type is a GRE tunnel, after the first packet reaches the first network device, the first network device is the first network device. The packet encapsulates a GRE header; for another example, assuming that the next hop information of the forwarding entry corresponding to the first packet on the first network device indicates that the outbound interface type is an IPSec tunnel, then, after the first packet reaches the first network device, The first network device encapsulates an IPSec header for the first packet, and the IPSec header may specifically be an authentication header (English: Authentication Header, abbreviated: AH) or an encapsulating security payload (English: Encapsulating Security Payload, abbreviated: ESP).

In some embodiments, S101 may specifically include: S21, when the first network device determines that it is the head end point of the first tunnel, the first network device determines whether the first packet includes a hash value, and if so, executes S24 , otherwise perform S22; S22, the first network device obtains the message feature information of the first message; S23, the first network device calculates the hash value corresponding to the first message based on the message feature information; S24, the first The network device generates a second packet based on the first packet and the hash value, where the second packet includes the hash value. In this case, in addition to the hash value, the second packet also includes the entire content of the first packet (that is, the content related to the first packet referred to in S101).

If the first packet has been fragmented or encapsulated with other tunnels before reaching the first network device, the first packet may already include a hash that reflects the feature diversity of the first packet. At this time, in order to save computing resources, when the first network device determines that it is the head end point of the first tunnel, it can directly obtain the hash value from the first packet, and generate the second packet based on S24. If the first packet does not carry the hash value, then the hash value can be obtained by calculation based on S22 to S23, and for the specific process, please refer to the above-mentioned description of S12.

In some embodiments, the process of generating the second packet by the first network device in S24 includes: the first network device encapsulates the first tunnel header corresponding to the first tunnel for the first packet, and obtains a second packet, the first The packet header in the second packet includes the hash value.

As an example, the packet header of the second packet carrying the hash value may be the first tunnel header. For example, the hash value may be carried in the Reserved field of the first tunnel header, or the hash value may be carried in the flag bit (English: Flags) of the first tunnel header, or the hash value may also be carried in the first tunnel header. An extension field in an optional (English: Optional) field of a tunnel header carries the hash value.

The first tunnel may be the first VPN tunnel, for example, the first GRE tunnel or the first IPSec tunnel.

For example, assuming that the first tunnel is a GRE tunnel, then the first tunnel header is a GRE header. The format of the second packet is shown in Figure 4a or Figure 4b, and includes: an outer IP header, a GRE header, an inner IP header, and a payload. Referring to Fig. 4a, the hash value a of the first packet may be carried in the Flags of the second packet, that is, Flags=hash value a. Referring to Fig. 4b, the hash value a of the first packet can also be carried in the Optional extension field of the second packet, and the value of Flags indicates the position where the hash value a is carried, that is, the position of Flags=Optional , Optional field = hash value a.

For another example, assuming that the first tunnel is an IPSec tunnel, then the first tunnel header is an IPSec header. The format of the second packet is shown in FIG. 4c or FIG. 4d , and the hash value a of the first packet may be carried in the Reserved field of the IPSec header in the second packet. As shown in Figure 4c, the second packet corresponds to the tunnel mode of IPSec, and the second packet includes: an outer IP header, an IPSec header, an inner IP header, and a payload. As shown in FIG. 4d , the second packet corresponds to the transmission mode of IPSec, and the second packet includes: an inner IP header, an IPSec header, and a payload. It should be noted that, the IPSec header in FIG. 4c and FIG. 4d may be AH or ESP, which is not specifically limited in this embodiment of the present application.

As another example, if the first tunnel is in the tunnel mode, when the first network device encapsulates the first VPN header for the first packet, it also needs to encapsulate the outer IP header in the outer layer of the first VPN header. In order to allow subsequent network devices to easily read the hash value during the packet forwarding process, the hash value can also be carried in the outer IP header, that is, the second packet carrying the hash value. The header can also be an outer IP header. The specific carrying manner is not limited in this embodiment.

In some embodiments, the types of VPN tunnels include but are not limited to GRE tunnels, IPSec tunnels, virtual extensible local area network (English: virtual extensible local area network, VXLAN for short) tunnels, layer 2 tunneling protocol (English: layer 2 tunneling protocol, Abbreviation: L2TP) tunnel, Layer 3 tunneling protocol version 3 (English: layer 3 tunneling protocol version 3, Abbreviation: L2TPv3) tunnel, Multi-protocol label switching (English: multi-protocol label switch, Abbreviation: MPLS) tunnel, IPv6 Over IPv4 tunnel, IPv4 Over IPv6 tunnel, etc. For a specific implementation manner of generating the second packet, reference may be made to the above-mentioned related descriptions taking the GRE tunnel and the IPSec tunnel as examples, and details are not repeated here.

The specific operations of the above S101 may be completed in a processing unit in the first network device.

In some embodiments, the preset conditions in S101 include, but are not limited to, the descriptions in the above two possible implementation manners. For example, if the first packet is an IP packet that does not need to be fragmented or encapsulated a tunnel, then it can also be considered that the first packet meets the preset conditions, and the hash value corresponding to the first packet is carried in the IP The second packet is obtained from the IP extension header of the packet. Therefore, in the process of forwarding the second packet, each network device only needs to read the hash value in the IP extension header of the second packet. The load balancing of the second packet does not require each network device to perform hash calculation to obtain the hash value corresponding to the first packet, which saves the time of packet forwarding, and can effectively improve the load balance on the basis of ensuring load balance. Efficiency of text forwarding.

It can be seen that the first network device carries the hash value corresponding to the first packet in a position in the newly generated second packet that is easy to read, which provides convenience for load balancing when each subsequent network device forwards the second packet. Moreover, since the hash value of the first packet reflects the diversity of the inner layer features of the first packet, the load balancing effect in the network is guaranteed.

S102, the first network device sends the second packet to the second network device.

S103, the second network device receives the second packet sent by the first network device.

In some embodiments, the first network device may also perform load balancing based on the hash value in the second packet. For example: when the first network device determines that there are multiple outgoing interfaces based on the forwarding entry corresponding to the second packet, at this time, the first network device determines one outgoing interface from the multiple outgoing interfaces based on the hash value, so that the The second packet is forwarded from the determined outbound interface. For the specific process, refer to the relevant description of the following S104.

In the case of encapsulating a multi-layer tunnel at the first network device, if the first tunnel header is encapsulated for the first packet according to the above S101, and the second packet is obtained, then the method 100 may further include: S31, the first The network device determines that it is the head end point of the second tunnel, then obtains the hash value from the second packet; S34, the first network device encapsulates the second tunnel header corresponding to the second tunnel for the second packet, and obtains the updated The updated second packet includes the hash value, for example, the second tunnel header includes the hash value. If the second tunnel header is first encapsulated for the first packet, that is, before S101, the method 100 may further include: S41, the first network device determines that it is the head end of the second tunnel, then the first network device based on the packet The characteristic information calculates the hash value corresponding to the first packet; S42, the first network device encapsulates the second tunnel header for the first packet to obtain a third packet, and the third packet includes the hash value; at this time , S101 may include, for example: the first network device determines that it is the head end point of the first tunnel, then the first network device obtains the hash value from the third packet, so that the first network device encapsulates the third packet for the third packet. A tunnel header is obtained, and a second packet is obtained, where the second packet includes a hash value, for example, the second tunnel header and the first tunnel header include hash values. It should be noted that, multiple layers of tunnels are encapsulated on the same network device at the same time, the encapsulation sequence of the multiple layers of tunnels is not specifically limited, and the encapsulation sequence does not affect the effect of subsequent load balancing.

In the scenario of encapsulating a multi-layer tunnel, the second packet sent by the first network device to the second network device is the updated second packet, that is, the packet transmitted from the outgoing interface after the first network device has completed the final processing. Arts.

It should be noted that, regardless of whether the tunnel is superimposed or not, as long as the header of the second packet sent to the second network device includes a hash value that can reflect the inner layer characteristics of the first packet, it will not affect subsequent packets. Load balancing effect.

S104, the second network device processes the second packet based on the hash value in the second packet.

The second packet is forwarded from the first network device to the second network device, and the second network device may be directly connected to the first network device, or may be indirectly connected through other network devices.

In some possible implementations, in order to enhance the performance of forwarding and other aspects, the second network device may include multiple processing units. Then, after receiving the second packet, the second network device may Based on the hash value carried in the second packet, a processing unit corresponding to the second packet is determined, and the second packet is allocated to the processing unit for processing.

As an example, the second network device may directly determine the processing unit for processing the second packet based on the hash value. For example: assign the second packet to the processing unit with the same number and hash value. Assuming that the second network device includes 8 processing units, numbered 0 to 7, then, when the hash value in the second packet is 0, the second network device allocates the second packet to processing unit 0; When the hash value in the second packet is 1, the second network device allocates the second packet to the processing unit 1 .

As another example, the corresponding relationship between the hash value and the processing unit may be preset in the second network device to determine the processing unit that processes the second packet. For example, the second network device may preset the corresponding relationship according to the capability of each processing unit. Assuming that the second network device includes two processing units, and the capability of processing unit 0 is stronger than that of processing unit 1, then the preset correspondence relationship of the second network device includes: correspondence relationship 1 "Hash values 0 to 5 correspond to processing Unit 0", corresponding relationship 2 "Hash values 6 and 7 correspond to processing unit 1", in this way, when the hash value in the second packet is any one of 0 to 5, the second network device according to the corresponding relationship 1 and Hash value, it is determined that the second packet needs to be allocated to processing unit 0; when the hash value in the second packet is 6 or 7, the second network device determines according to the corresponding relationship 2 and the hash value. The second message is assigned to the processing unit 1 .

In some scenarios, for a second network device including multiple processing units, the multiple processing units often process in parallel, and after the processing is completed, multiple packets are sent from the second network device uniformly, thereby improving parallel forwarding efficiency. However, if the effect of uniform hashing is not good, the distribution of hash values used to guide the selection of processing units is not uniform. For example, for a flow that loses some packet characteristics, the five-tuple is different but the hash calculated based on the three-tuple is used. When the values are equal, the problem of uneven distribution of hash values will occur, which is likely to cause one or some of the multiple processing units to be heavily loaded, and other processing units to be lightly loaded, resulting in different processing speeds of each processing unit. , the parallel forwarding efficiency becomes slower, thus affecting or losing the advantages of multi-processing units. However, based on the method provided by the embodiment of the present application, since the hash value that can reflect the inner layer characteristics of the message is carried in the message and forwarded to the network device of the multi-processing unit, the message to be forwarded can be based on the evenly distributed hash value. The values are relatively evenly distributed to each processing unit in the network device, so that the efficient parallel forwarding function of multiple processing units can be fully and effectively utilized to achieve efficient forwarding.

In some other possible implementation manners, if the processing unit in the second network device that receives the second packet determines that there are multiple outgoing interfaces according to the forwarding entry corresponding to the second packet, at this time, the second network device Based on the hash value, one outgoing interface is determined from the multiple outgoing interfaces, so that the second packet is forwarded from the determined outgoing interface.

As an example, the second network device may directly determine the outbound interface for forwarding the second packet based on the hash value. For example, assign the second packet to the outbound interface with the same number and hash value. Assuming that the second network device includes 8 outgoing interfaces, numbered 0 to 7, then when the hash value in the second packet is 0, the second network device forwards the second packet from the given interface 0 outgoing the second network device; when the hash value in the second packet is 1, the second network device forwards the second packet from the outgoing interface 1 to the second network device.

As another example, the correspondence between the hash value and the outgoing interface may be preset in the second network device to determine the outgoing interface for forwarding the second packet. For example, the second network device may preset a corresponding relationship according to parameters such as the bandwidth of each outgoing interface. Assuming that the second network device includes two interfaces, and the bandwidth of interface 0 is three times that of interface 1, then the preset correspondence relationship of the second network device includes: correspondence relationship 1 "Hash values 0-5 correspond to interface 0" , corresponding relationship 2 "hash values 6 and 7 correspond to interface 1", in this way, when the hash value in the second packet is any one of 0 to 5, the second network device according to the corresponding relationship 1 and the hash value, It is determined that the second packet needs to be forwarded from interface 0 to the second network device; when the hash value in the second packet is 6 or 7, the second network device determines, according to the corresponding relationship 2 and the hash value, that it needs to The second packet is forwarded from the interface 1 to the second network device.

If the second network device determines that it is the head end point of the third tunnel, the second network device can obtain the hash value from the second packet, and encapsulate the third tunnel header for the second packet to obtain the first Four packets, wherein the hash value is included in the third tunnel header of the fourth packet. At this time, if the second network device determines that the fourth packet has multiple outgoing interfaces, the second network device may determine to forward the fourth packet to the outgoing interface of the second network device based on the hash value in the third tunnel header. interface; and, after the second network device sends the fourth packet to the third network device, the hash value in the fourth packet can also be used by the third network device to select the processing unit and the outgoing interface to achieve corresponding load balancing. For example, the first tunnel can be a GRE tunnel, and the third tunnel can be an IPSec tunnel. Then, the fourth packet can be considered to encapsulate an IPSec Over GRE tunnel; another example: the first tunnel can be an IPSec tunnel, and the third tunnel can be If it is a GRE tunnel, then the fourth packet can be considered to be encapsulated with a GRE Over IPSec tunnel; for another example, if the first tunnel and the second tunnel are both GRE tunnels, then the fourth packet can be considered to be encapsulated with a GRE Over IPSec tunnel. GRE tunnel.

If the second network device is not the terminal point of the first tunnel, then S104 may forward the second packet to the subsequent network device of the first tunnel based on the hash value in the second packet. If the second network device is the terminal point of the first tunnel, after determining the outbound interface, the second network device can strip the first tunnel header to obtain a fifth packet, and send the fifth packet from the determined outbound interface. The interface is transferred out.

In the case where the second network device is the termination point of the first tunnel, in one case, the second network device may directly continue forwarding the fifth packet in the network without processing the fifth packet. In this case, if the fifth packet does not include other tunnel headers, the subsequent nodes can obtain the inner packet feature information of the fifth packet. Therefore, the direct forwarding of the fifth packet does not affect subsequent nodes. The effect of node load balancing on packets based on the hash uniformity. In another case, the second network device may carry the hash value in the IP extension header of the fifth packet, generate the sixth packet, and continue to forward the sixth packet in the network. In this case, not only the load balancing effect in the subsequent forwarding process can be ensured, but also the subsequent network device does not need to obtain the packet feature information of the fifth packet, nor does it need to calculate the hash calculation, which saves the computing resources of each network device. It also greatly improves the packet forwarding efficiency.

In some embodiments, if the network device that receives the message to be forwarded determines that the preset conditions are not met and does not carry the hash value, then the message feature information of the message can be obtained, and the hash value can be calculated based on the message feature information. Hash value, and load balancing based on the calculated hash value.

It can be seen that, with the method 100 provided in this embodiment of the present application, when it is determined that the preset conditions are met, the network device obtains the packet feature information of the packet, performs hash calculation based on the packet feature information, and obtains the hash value corresponding to the packet. The hash value is carried in the packet for subsequent forwarding, so that the network device that receives the packet can perform load balancing operations on the packet based on the hash value carried in the packet. Since the hash value is calculated based on the inner quintuple of the message, it reflects the diversity of the message, ensures the even distribution of the hash value to a certain extent, and achieves a better load balancing effect; moreover, only A network device is required to perform a hash calculation, and subsequent network devices only need to read the hash value from the message when there is a load balancing requirement, which saves the computing resources of each network device and improves the forwarding efficiency of the message. Therefore, the forwarding performance of the network device can be improved.

In addition, an embodiment of the present application further provides a first network device 500, as shown in FIG. 5 . The first network device 500 includes a processing unit 501 and a sending unit 502 . The processing unit 501 is configured to perform the processing operation performed by the first network device in the embodiment shown in FIG. 2 ; the sending unit 502 is configured to perform the sending operation performed by the first network device in the embodiment shown in FIG. 2 . For example, the processing unit 501 may perform the operation in the embodiment in FIG. 2 : when the preset condition is satisfied, the second packet is generated according to the first packet. For example, the sending unit 502 may perform the operation in the embodiment in FIG. 2: send the second packet to the second network device.

In addition, an embodiment of the present application further provides a second network device 600, as shown in FIG. 6 . The second network device 600 includes a receiving unit 601 and a processing unit 602 . The receiving unit 601 is configured to perform the receiving operation performed by the second network device in the above-mentioned embodiment shown in FIG. 2 ; the processing unit 602 is configured to perform the processing operation performed by the second network device in the above-mentioned embodiment shown in FIG. 2 . For example, the receiving unit 601 may perform the operation in the embodiment in FIG. 2: receive the second packet sent by the first network device; the processing unit 602 may perform the operation in the embodiment in FIG. 2: based on the hash in the second packet The value is the value, and the second packet is processed.

In addition, an embodiment of the present application further provides a first network device 700, as shown in FIG. 7 . The first network device 700 includes a second communication interface 702 and a processor 703 . The second communication interface 702 is configured to perform the sending operation performed by the first network device in the aforementioned embodiment shown in FIG. 2 ; the processor 703 is configured to perform the operations performed by the first network device in the aforementioned embodiment shown in FIG. 2 except for the receiving operation and the sending operation. operations other than operations. For example, the processor 703 may perform the operations in the embodiment in FIG. 2 : when the preset condition is satisfied, the second packet is generated according to the first packet. The first network device 700 may further include a first communication interface 701, and the first communication interface 701 is configured to perform the aforementioned receiving operation performed by the first network device in the embodiment shown in FIG. 2, such as receiving a first message Arts.

In addition, an embodiment of the present application further provides a second network device 800, as shown in FIG. 8 . The second network device 800 includes a first communication interface 801 and a processor 803 . The first communication interface 801 is configured to perform the receiving operation performed by the second network device in the foregoing embodiment shown in FIG. 2 ; the processor 803 is configured to perform the receiving operation performed by the second network device in the foregoing embodiment shown in FIG. 2 . and operations other than send operations. For example, the processor 803 may perform the operation in the embodiment in FIG. 2 : process the second packet according to the hash value in the second packet. The second network device 800 may further include a second communication interface 802, where the second communication interface 802 is configured to perform the sending operation performed by the second network device in the foregoing embodiment shown in FIG. 2 .

In addition, an embodiment of the present application further provides a first network device 900, as shown in FIG. 9 . The first network device 900 includes a memory 901 and a processor 902 in communication with the memory 901 . The memory 901 includes computer-readable instructions; the processor 902 is configured to execute the computer-readable instructions, so that the first network device 900 executes the method performed by the first network device in the above embodiment shown in FIG. 2 .

In addition, an embodiment of the present application further provides a second network device 1000, as shown in FIG. 10 . The second network device 1000 includes a memory 1001 and a processor 1002 in communication with the memory 1001 . The memory 1001 includes computer-readable instructions; the processor 1002 is configured to execute the computer-readable instructions, so that the second network device 1000 executes the method performed by the second network device in the above embodiment shown in FIG. 2 .

In the above embodiment, the processor may be a central processing unit (English: central processing unit, abbreviation: CPU), a network processor (English: network processor, abbreviation: NP), or a combination of CPU and NP. The processor may also be an application-specific integrated circuit (English: application-specific integrated circuit, abbreviation: ASIC), a programmable logic device (English: programmable logic device, abbreviation: PLD) or a combination thereof. The above-mentioned PLD can be a complex programmable logic device (English: complex programmable logic device, abbreviation: CPLD), field programmable logic gate array (English: field-programmable gate array, abbreviation: FPGA), general array logic (English: generic array logic, abbreviation: GAL) or any combination thereof. The processor may refer to one processor, or may include multiple processors. The memory may include volatile memory (English: volatile memory), such as random-access memory (English: random-access memory, abbreviation: RAM); the memory may also include non-volatile memory (English: non-volatile memory), For example, read-only memory (English: read-only memory, abbreviation: ROM), flash memory (English: flash memory), hard disk (English: hard disk drive, abbreviation: HDD) or solid-state hard disk (English: solid-state drive, Abbreviation: SSD); the memory may also comprise a combination of the above-mentioned kinds of memory. The memory may refer to one memory, or may include multiple memories. In a specific embodiment, a computer program or instruction is stored in the memory, and the computer program or instruction includes a plurality of software modules, such as a sending module, a processing module and a receiving module. After executing each software module, the processor can perform corresponding operations according to the instructions of each software module. In this embodiment, the operation performed by a software module actually refers to the operation performed by the processor according to the instruction of the software module. After the processor executes the computer program or instruction in the memory, it can execute all operations that can be performed by each network node in the packet processing method according to the instructions of the computer program or instruction.

In the above embodiment, the second communication interface 702 of the first network device 700 can be specifically used as the sending unit 502 in the first network device 500 to implement data communication between the first network device and the second network device; The first communication interface 701 of a network device 700 may specifically be used as a receiving unit in the first network device 500, for example, may be used to receive a first packet sent by an upstream network device. Similarly, the first communication interface 801 of the second network device 800 can be specifically used as the receiving unit 601 in the second network device 600 to implement data communication between the first network device and the second network device; the second network device 800 The second communication interface 802 can be specifically used as a sending unit in the second network device 600 to implement data communication between the second network device and the downstream network device.

In addition, an embodiment of the present application further provides a communication system 1100, as shown in FIG. 11 . The communication system 1100 includes a first network device 1101 and a second network device 1102, wherein the first network device 1101 may specifically be the first network device 500, the first network device 700 or the first network device 900, the second network device 1102 may specifically be the above-mentioned second network device 600 , second network device 800 or second network device 1000 .

In addition, an embodiment of the present application also provides a computer-readable storage medium, where a computer program or instruction is stored in the computer-readable storage medium, and when it runs on a computer, the computer is made to execute the implementation shown in FIG. 2 above. the method described in the example.

In addition, the embodiments of the present application also provide a computer program product, including a computer program or a computer-readable instruction, when the computer program or the computer-readable instruction is run on a computer, the computer is made to execute the foregoing embodiment shown in FIG. 2 . the method described in .

The "first" in the names such as "first packet" and "first tunnel" mentioned in the application embodiments is only used for name identification, and does not represent the first in order. The same rule applies to "second" etc.

From the description of the above embodiments, those skilled in the art can clearly understand that all or part of the steps in the methods of the above embodiments can be implemented by means of software plus a general hardware platform. Based on this understanding, the technical solution of the present application can be embodied in the form of a software product, and the computer software product can be stored in a storage medium, such as read-only memory (English: read-only memory, ROM)/RAM, magnetic disk, An optical disc, etc., includes several instructions for causing a computer device (which may be a personal computer, a server, or a network communication device such as a router) to execute the methods described in various embodiments or some parts of the embodiments of the present application.

Each embodiment in this specification is described in a progressive manner, and the same and similar parts between the various embodiments may be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the system embodiments and device embodiments, since they are basically similar to the method embodiments, the description is relatively simple, and reference may be made to some descriptions of the method embodiments for related parts. The device and system embodiments described above are only illustrative, wherein the modules described as separate components may or may not be physically separated, and the components displayed as modules may or may not be physical modules, that is, they may be located in One place, or it can be distributed over multiple network elements. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution in this embodiment. Those of ordinary skill in the art can understand and implement it without creative effort.

The above descriptions are only preferred embodiments of the present application, and are not intended to limit the protection scope of the present application. It should be pointed out that for those of ordinary skill in the art, without departing from the present application, several improvements and modifications can also be made, and these improvements and modifications should also be regarded as the protection scope of the present application.

Claims

A method for processing a message, comprising:

When a preset condition is met, the first network device generates a second packet according to the first packet, where the second packet includes a hash value and content related to the first packet, and the hash value is Obtained by performing hash calculation based on the message feature information of the first message;

The first network device sends the second packet to the second network device.
The method according to claim 1, wherein the preset condition is that the length of the first packet is greater than the maximum transmission unit MTU of the first network device.
The method according to claim 2, wherein the content related to the first packet includes a fragmented packet of the first packet.
The method according to claim 3, wherein the header of the second packet includes the hash value.
The method according to claim 3, wherein the second packet includes an IP extension header, and an Option field in the IP extension header includes the hash value.
The method according to claim 5, wherein the IP extension header of the second packet includes a reserved Reserved field, and the Reserved field includes the hash value.
The method according to claim 1, wherein the preset condition is that the first network device is the head point of the first tunnel.
The method according to claim 7, wherein the second packet further includes a first tunnel header.
9. The method of claim 8, wherein the first tunnel header includes the hash value.
The method according to claim 9, wherein the first tunnel header includes a Reserved Reserved field, and the Reserved field includes the hash value.
The method according to any one of claims 7-10, wherein the first tunnel is a virtual private network VPN tunnel.
A method for processing a message, comprising:

The second network device receives a second packet sent by the first network device, where the second packet includes a hash value and content related to the first packet, and the hash value is based on the first packet Packet feature information is calculated;

The second network device processes the second packet based on the hash value.
The method according to claim 12, wherein, the second network device processing the second packet based on the hash value, comprising:

The second network device allocates the second packet to a first processing unit of the second network device based on the hash value, where the first processing unit corresponds to the hash value.
The method according to claim 12 or 13, wherein, the second network device processing the second packet based on the hash value, comprising:

The second network device forwards the second packet from a first interface of the second network device based on the hash value, where the first interface corresponds to the hash value.
The method according to any one of claims 12-14, further comprising:

The second network device determines that the second network device is the head point of the second tunnel, then, obtains the hash value from the second message;

The second network device encapsulates the second tunnel header corresponding to the second tunnel for the second packet, and obtains a third packet, where the second tunnel header of the third packet includes the hash. Greek value;

The second network device sends the third packet to a third network device, so that the third network device processes the third packet based on the hash value.
The method according to any one of claims 12-14, further comprising:

The second network device determines that the length of the second packet is greater than the maximum transmission unit MTU of the second network device, and then processes the second packet into at least two fragmented packets;

The second network device adds the hash value to each fragmented packet to obtain at least two third packets;

The second network device sends at least two of the third packets to a third network device, so that the third network device processes at least two of the third packets based on the hash value.
A network device, characterized in that it includes:

memory comprising computer readable programs or instructions;

A processor in communication with the memory for executing the computer-readable program or instructions to cause the network device to perform the method of any one of claims 1-11.
A network device, characterized in that it includes:

memory comprising computer readable programs or instructions;

A processor in communication with the memory for executing the computer-readable program or instructions to cause the network device to perform the method of any one of claims 12-16.
A communication system, characterized in that, the communication system comprises the network device of claim 17 and/or the network device of claim 18 .
A computer-readable storage medium, characterized in that, program codes or instructions are stored in the computer-readable storage medium, which, when run on a computer, cause the computer to execute the above-mentioned claims 1-11. the method described or the method described in any one of claims 12-16.