Flow processing method and related equipment

Abstract

The embodiment of the invention provides a traffic processing method and related equipment, which are used for reducing hardware resource occupation as much as possible. The method provided by the embodiment of the invention comprises the following steps: dividing each physical device PF with a hardware function into a plurality of virtual instance devices VFs based on a virtualization technology, and determining any PF and a virtual network interface VSI corresponding to any VF; determining a target index table according to all the VSIs between all the PFs and all the VFs; defining a significant bit value of an index significant bit and a copy bit value of a traffic copy bit of a target index relationship between any PF and any VF in a target index table to obtain a multicast member table corresponding to all PFs and all VFs; wherein the index effective bit is used for identifying an effective rule of the target index relation, and the flow replication bit is used for representing a replication rule of the network flow of the VSI associated with the target index relation; and copying the network flow for any PF and VF according to the effective bit value and the copy bit value in the multicast member table so as to complete the copying process of the network flow.

Description

Technical Field

[0001] The embodiments of the present application relate to the field of communication networks, and in particular, to a traffic processing method and related equipment. Background Art

[0002] The current software defined network (SDN) in the industry, such as the virtual switch (OVS, Open vSwitch), can realize the traffic entering a single in_port and grouping out from a multicast group by configuring the flow table of the network communication protocol OpenFLow, where the multicast group can be regarded as multiple traffic outlets.

[0003] At present, traffic replication is mostly achieved by increasing network bandwidth. However, if there are too many replication members in each flow table, it cannot meet the needs of large-scale scenarios, because the larger the leaf node, the more hardware resources it occupies. For example, a leaf node requires at least 64 bits. In large-scale scenarios, if there are many replication members in each flow table, a large amount of hardware resources will be occupied.

[0004] Therefore, how to achieve traffic configuration while occupying less hardware resources has become a technical problem that needs to be solved urgently. Summary of the invention

[0005] The embodiments of the present application provide a traffic processing method and related devices for minimizing hardware resource usage.

[0006] A first aspect of an embodiment of the present application provides a traffic processing method, which is applied to a network device, wherein the network device includes at least two physical devices PF having hardware functions, and the method includes:

[0007] Based on virtualization technology, each of the physical devices PF with hardware functions is divided into multiple virtual instance devices VF, and a virtual network interface VSI corresponding to any PF and any VF is determined;

[0008] Determine a target index table according to all VSIs between all PFs and all VFs; wherein the target index table is used to characterize the index relationship between all PFs and all VFs;

[0009] The valid bit value of the index valid bit and the replication bit value of the traffic replication bit of the target index relationship between any PF and any VF are defined in the target index table to obtain a multicast member table corresponding to all PFs and all VFs; wherein the index valid bit is used to identify the valid rule of the target index relationship, and the traffic replication bit is used to characterize the replication rule of the network traffic of the VSI associated with the target index relationship;

[0010] In the multicast member table, the network traffic for any of the PFs and the VFs is replicated according to the valid bit value and the replication bit value to complete the network traffic replication process.

[0011] Optionally, the PF includes a register, and the replicating the network traffic for any of the PF and the VF according to the valid bit value and the replica bit value includes:

[0012] Determine the leaf node of the network traffic in the multicast member table, which is marked as a valid bit in the valid bit value and as a copy bit in the copy bit value, and mark the leaf node identifier of the leaf node; wherein the leaf node identifier is used to indicate that the network traffic is traffic to be copied;

[0013] If the hardware performance of the register reaches a preset performance condition, the network traffic identified as traffic to be copied is input into a low priority scheduling queue of the register to copy the network traffic;

[0014] If the hardware performance of the register does not meet the preset performance condition, the network traffic identified as traffic to be copied is input into the high priority scheduling queue of the register to copy the network traffic.

[0015] Optionally, before defining the valid bit value of the index valid bit and the replication bit value of the traffic replication bit of the target index relationship between any PF and any VF in the target index table, the method further includes:

[0016] If the PF hits the flow table rule of the network communication protocol flow table, an action field table corresponding to the PF is obtained according to the flow table rule; wherein the action field table includes an action capability field and an action record field, the action capability field is used to describe the field for performing an action in the network communication protocol flow table, and the action record field is used to describe the field for recording an action in the action capability field;

[0017] Traversing the action capability field to determine an execution action bit value corresponding to the execution action in the network communication protocol flow table;

[0018] The execution action data corresponding to the execution action is determined in the action record field according to the execution action bit value, so as to set the valid bit value of the index valid bit and the copy bit value of the traffic copy bit according to the execution action data.

[0019] Optionally, the network communication protocol flow table includes at least a clear action interface, and the method further includes:

[0020] Determining the starting position of the action capability field;

[0021] A clear instruction field corresponding to the clear action interface is configured at the starting position to clear the field corresponding to the execution action in the action capability field and the field corresponding to the record action in the action record field according to the clear instruction field; wherein the clear instruction field is used to clear the offset length of the field for executing the action and the action length of the field for recording the action.

[0022] Optionally, the network traffic replication process includes:

[0023] When the network traffic enters the PF, the network traffic is cached in a hardware buffer of the PF;

[0024] Parsing the network traffic in the hardware buffer to obtain traffic information corresponding to the network traffic; wherein the traffic information at least includes a traffic weight and a traffic reference index of the network traffic, and the traffic reference index is associated with an index relationship of the target index table;

[0025] Determining, according to the traffic reference index, whether the network traffic is network traffic for traffic replication;

[0026] If the network traffic is network replication traffic, when any network replication traffic is replicated, the traffic weight of the network replication traffic is reduced by a preset weight value to obtain a next traffic weight, and the step of reducing the traffic weight of the network replication traffic by a preset weight value when any network replication traffic is replicated is performed; wherein the network replication traffic is used to characterize the network traffic for which traffic replication is performed;

[0027] When the next traffic weight is zero, all the network replication traffic that has completed replication is released to complete the replication of the network traffic for any of the PF and the VF.

[0028] Optionally, after obtaining the next traffic weight, the method further includes:

[0029] When the network replication traffic is the last network traffic in the multicast member table, the traffic weight corresponding to the last network traffic is cleared, and all the network replication traffic that has been replicated is released to complete the replication of the network traffic for any of the PF and the VF.

[0030] Optionally, the method further comprises:

[0031] If the network traffic is the unicast traffic, the traffic weight of the unicast traffic is cleared to zero, and the unicast traffic is released to execute the traffic replication process of another network traffic; wherein the unicast traffic is used to characterize other network traffic except the network traffic replication.

[0032] A second aspect of an embodiment of the present application provides a traffic processing system, which is applied to a network device, wherein the network device includes at least two physical devices PF having hardware functions, and the system includes:

[0033] A division unit, used for dividing each of the physical devices PF having hardware functions into a plurality of virtual instance devices VF based on virtualization technology, and determining a virtual network interface VSI corresponding to any PF and any VF;

[0034] A determination unit, configured to determine a target index table according to all VSIs between all PFs and all VFs; wherein the target index table is used to characterize the index relationship between all PFs and all VFs;

[0035] A definition unit, used to define the valid bit value of the index valid bit and the replication bit value of the traffic replication bit of the target index relationship between any PF and any VF in the target index table, so as to obtain a multicast member table corresponding to all PFs and all VFs; wherein the index valid bit is used to identify the valid rule of the target index relationship, and the traffic replication bit is used to characterize the replication rule of the network traffic of the VSI associated with the target index relationship;

[0036] The replication unit is used to replicate the network traffic for any of the PFs and the VFs in the multicast member table according to the valid bit value and the replication bit value to complete the replication process of the network traffic.

[0037] The traffic processing system provided in the second aspect of the embodiment of the present application is used to execute the traffic processing method described in the first aspect.

[0038] A third aspect of an embodiment of the present application provides a flow processing device, including:

[0039] CPU, memory, input and output interfaces, wired or wireless network interfaces, and power supply;

[0040] The memory is a short-term storage memory or a persistent storage memory;

[0041] The central processing unit is configured to communicate with the memory and execute instruction operations in the memory to perform the traffic processing method described in the first aspect.

[0042] A fourth aspect of an embodiment of the present application provides a computer-readable storage medium, wherein the computer-readable storage medium includes instructions. When the instructions are executed on a computer, the computer executes the traffic processing method described in the first aspect.

[0043] A fifth aspect of an embodiment of the present application provides a computer program product, which includes instructions. When the instructions are executed on a computer, the computer executes the traffic processing method described in the first aspect.

[0044] It can be seen from the above technical scheme that the embodiment of the present application has the following advantages: through a traffic processing method disclosed in the embodiment of the present application, first based on virtualization technology, each physical device PF with hardware functions is divided into multiple virtual instance devices VF, and the virtual network interface VSI corresponding to any PF and any VF is determined; then according to all VSIs between all PFs and all VFs, the target index table is determined; wherein the target index table is used to characterize the index relationship between all PFs and all VFs; then, the valid bit value of the index valid bit of the target index relationship between any PF and any VF and the replication bit value of the traffic replication bit are defined in the target index table to obtain a multicast member table corresponding to all PFs and all VFs; wherein the index valid bit is used to identify the valid rules of the target index relationship, and the traffic replication bit is used to characterize the replication rules of the network traffic of the VSI associated with the target index relationship; finally, in the multicast member table, the network traffic for any PF and VF is replicated according to the valid bit value and the replication bit value to complete the replication process of the network traffic. Thus, by moving the traffic replication function processed by the software to the hardware processing, the software pressure is released to avoid a large amount of resources being occupied by a single business. At the same time, because PF focuses on traffic replication services, it can occupy fewer resources and hardware processing is simpler. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present application. For ordinary technicians in this field, other drawings can also be obtained based on these drawings.

[0046] Figure 1 A flow chart of a flow processing method disclosed in an embodiment of the present application;

[0047] Figure 2 A flow chart of another flow processing method disclosed in an embodiment of the present application;

[0048] Figure 3 A flow chart of another flow processing method disclosed in an embodiment of the present application;

[0049] Figure 4 A flow chart of another flow processing method disclosed in an embodiment of the present application;

[0050] Figure 5A schematic diagram of the structure of a flow processing system disclosed in an embodiment of the present application;

[0051] Figure 6 A schematic diagram of the structure of a flow processing device disclosed in an embodiment of the present application. DETAILED DESCRIPTION

[0052] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0053] The current software defined network (SDN) in the industry, such as the virtual switch (OVS, Open vSwitch), can achieve a single in_port incoming traffic and out from a multicast group group by configuring the flow table of the network communication protocol OpenFLow, where the multicast group can be regarded as multiple traffic exports. However, using the multicast traffic replication function will consume more system resources and bandwidth, and may increase network latency. Therefore, there are two main problems with multi-export replication of traffic: one is that it consumes more bandwidth, and the other is that it consumes more software resources. There is nothing wrong with using traffic replication to increase bandwidth. At the same time, although the loss of software performance caused by multi-traffic replication can be avoided by offloading to hardware, when there are many multicast group members, it is impossible to reduce the burden on the OVS software by combining software and hardware.

[0054] Specifically, the main process of multicast traffic replication and offloading is currently: 1) OVS receives the message and performs software traffic replication; 2) If flow table offload is enabled, the generated precise traffic is sent to the hardware through the network offload interface; 3) First, according to the key+action of the OVS flow table, the precise flow table is sent to the field programmable gate array (FPGA) flow table matching module. The key sent to the hardware is the key of the OVS precise traffic, the action is to execute the traffic replication action, and the data of the action is the first node index ID of the multicast replication module; then according to the OVS action: multiple export dports (multicast members), each dport is sent to the multicast replication module, and the members are managed in the multicast replication module through a linked list. After the multicast traffic replication flow table is unloaded, the same traffic is directly replicated by hardware. The replication process is as follows: 1) The hardware receives the message and matches the flow table according to the message. If it hits, the hardware traffic replication is performed; 2) The flow table action obtained by the query includes the action of multicast traffic replication, and the data of the action is extracted: the index ID of the first node of the multicast replication module linked list; 3) According to the index ID, the first node is found in the multicast replication module and the replication action is executed. (Each node will record the position of the next node. As long as the first node is found, the next member can be found through the next of the current node).

[0055] As mentioned above, although the existing technology can complete the hardware unloading of multicast replication, its implementation scheme is too cumbersome and cannot meet the needs of large-scale multicast replication well, because multicast replication occupies too many resources of a single leaf node. If there are too many leaf nodes, the hardware will not be able to bear it.

[0056] To solve the technical problems mentioned above, please participate Figure 1 , Figure 1 This is a flow chart of a traffic processing method disclosed in an embodiment of the present application, including steps 101 to 104.

[0057] 101. Based on virtualization technology, each physical device PF with hardware functions is divided into multiple virtual instance devices VF, and a virtual network interface VSI corresponding to any PF and any VF is determined.

[0058] It should be explained in advance that the difference between the technical solution in this embodiment and the above-mentioned technical solution is that the hardware implementation of this technical solution is no longer implemented through the linked list of the core module (MCC, MultiCast Copy) tbl. Specifically, a leaf node in the previous MCC TBL requires at least 64 bits. In large-scale scenarios, when the number of replication members in each flow table is relatively large, a large amount of hardware resources will be occupied. In addition, the linked list management method may cause deadlock risks and bring hidden dangers to hardware stability. This technical solution is implemented through bitmap. It should also be explained that the MCC module will obtain the multicast replication linked list from the multicast replication table based on the index ID of the first node of the linked list from the multicast replication action, and replicate the original message according to the replication linked list; the table entry members in each multicast replication table include but are not limited to:

[0059] Port: indicates the copied export port_id;

[0060] next: represents the next pointer of the current node, pointing to the next node to be copied;

[0061] tail: indicates whether the current node is the last node; if it is 1, it means the current node is the tail node of the linked list, otherwise it is not the tail node;

[0062] Through the above three members, as long as the multicast replication action outputs an index ID of the first node, the replication list can be found and the replication action can be executed. The MCC TBL module is mainly used for the MCC module to provide replication information, find the replication list through MCCaction, and execute the replication action; at the same time, the module can also be dynamically added and deleted by software (OVS flow table unloading and aging use).

[0063] In this embodiment, it is mainly applied to network devices. Among them, the network device includes at least two physical devices (PF, physical function) with specific hardware functions. It should be noted that PF refers to the actual hardware function on the physical device. It represents the entity implementation of the physical device, such as a network adapter, a graphics card, etc. Each physical device usually has one or more PFs. Therefore, in this embodiment, each physical device PF with hardware functions can be divided into multiple virtual instance devices (VF, virtual function) based on virtualization technology, and the virtual network interface (VSI, virtual station interface) corresponding to any PF and any VF is determined. It should also be noted that VF is usually created by virtualization technology (SR-IOV, Single Root I / O Virtualization) to divide PF into multiple virtual instances so as to provide independent network functions for different virtual machines or containers in a virtualized environment. SR-IOV is an IO virtualization technology that enables multiple system images (SI, System Image) to share PCI hardware resources. System image SI refers to software, such as an operating system or I / O device driver for executing an application. Therefore, with SR-IOV, a physical PCIe device can virtualize multiple virtual PCIe devices, and then directly connect the virtual PCIe devices to each virtual machine, so as to achieve the application scenario of a single physical PCIe device supporting multiple virtual machines. Each vsi represents a virtual network interface, and the PF and VF devices can be indexed through the VSI.

[0064] In one specific embodiment, the network device may be a smart network card, wherein the smart network card generally has multiple PFs, and each PF may enable a VF through SR-IOV. Specifically, based on the SR-IOV technology, it supports enabling multiple VFs for each PF. Each PF may correspond to multiple VFs.

[0065] 102. Determine a target index table according to all VSIs between all PFs and all VFs.

[0066] Based on step 101, the corresponding target index table can be determined according to all VSIs between all PFs and all VFs. It should be noted that the target index table includes index relationships representing the relationships between all PFs and all VFs.

[0067] In one specific embodiment, for example, the network card has two PF devices (PF0 and PF1), each device supports opening 32 VFs (VF0-VF31), then a total of 2+32*2=66 VSIs (0-65) are required, which are encoded as follows:

[0068] PF0:0

[0069] PF0VF0-PF0VF31:1-32

[0070] PF1:33

[0071] PF1VF0-PF1VF31:34-65

[0072] That is, all PF / VF indexes can be completed through 66 bits. Therefore, the index relationship between VF and PF can be realized through bitmap instead of MCC TBL linked list. In this way, the index relationship between PF and VF is completed.

[0073] 103. Define the valid bit value of the index valid bit and the copy bit value of the traffic copy bit of the target index relationship between any PF and any VF in the target index table to obtain a multicast member table corresponding to all PFs and all VFs.

[0074] Based on step 102, the valid bit value of the index valid bit and the replication bit value of the traffic replication bit of the target index relationship between any PF and any VF can be defined in the target index table, so as to obtain a multicast member table corresponding to all PFs and all VFs. It should be noted that the index valid bit is used to identify the valid rule of the target index relationship, and the traffic replication bit is used to characterize the replication rule of the network traffic of the VSI associated with the target index relationship.

[0075] In one specific embodiment, the MCC TBL mentioned above can be defined. Specifically, it can be defined whether the table entries of all index relations in the target index relation are valid, and whether the network traffic between the corresponding PF and VF in each index relation needs to be replicated. For easy understanding, please refer to Table 1:

[0076] Vld Indicates whether the entry is valid. Lead_dport_btm Multicast member list

[0077] It should be noted that Vld is the index valid bit described above, and its corresponding setting is the valid bit value. Lead_dport_btm is a bitmap table, that is, the traffic replication bit. For example, the bitmap has a total of 128 bits. If the 1st to 10th bit value is 1, it means that the traffic is replicated to VF1 to VF10 of PF0. Correspondingly, the multicast member table is actually a bitmap. Which position is 1 represents that the corresponding vsi needs to be replicated, which is the replication bit value mentioned above.

[0078] 104. In the multicast member table, the network traffic for any PF and VF is replicated according to the valid bit value and the replication bit value to complete the replication process of the network traffic.

[0079] Therefore, through the copy bit value and the valid bit value in the multicast member table described in step 103, the network traffic before any PF and VF can be copied (needs to be set to copy), thereby completing the network traffic copy process.

[0080] In one of the specific embodiments, since the replication bit value of the network traffic flowing through each PF and VF is set in the multicast member table (bitmap), it is possible to determine whether the corresponding network traffic needs to be replicated, and thereby complete the above-mentioned network traffic replication process. Furthermore, compared with the existing solutions, the performance and resource utilization of this technical solution are greatly improved. The linked list function can be realized through a bitmap.

[0081] A traffic processing method disclosed in this embodiment moves the traffic replication function processed by software to hardware processing, thereby relieving software pressure and avoiding a large amount of resources being occupied by a single service. At the same time, since PF focuses on traffic replication services, it can occupy fewer resources and hardware processing is simpler. Specifically, the PF / VF of the smart network card is indexed by VSI. Based on this design, multicast replication is easier. Each bit can represent a VSI. The corresponding position is 1, which means that replication is required, otherwise it is not required. This design can greatly save multicast replication resources.

[0082] Furthermore, replication traffic supports priority scheduling configuration, for details, see Figure 2 , Figure 2 This is a flow chart of another traffic processing method disclosed in an embodiment of the present application, including steps 201 to 203.

[0083] 201. Determine a leaf node of network traffic whose valid bit value is marked as a valid bit and whose copy bit value is marked as a copy bit in a multicast member table, and mark a leaf node identifier of the leaf node.

[0084] As can be seen from the above description, this embodiment mainly implements priority scheduling configuration of replication traffic. It should be noted that the replication traffic can be understood as network traffic (or messages) that need to be replicated.

[0085] In this embodiment, the network traffic marked as valid in the valid bit value and as replicated in the replicated bit value in the multicast member table is first determined, and the network traffic is defined as a leaf node, and the leaf node identifier of the leaf node is marked. It should be noted that the leaf node identifier is used to indicate that the network traffic is traffic to be replicated (network traffic that needs to be replicated).

[0086] In one of the specific embodiments, the network traffic in the multicast member table that needs to be replicated in the subsequent process is first determined. Thus, all replicated traffic is marked with a leaf identifier, that is, a leaf node identifier, indicating that the network traffic is a replicated traffic. Specifically, Leaf represents replicated traffic, which is prioritized for distinguishing it from other traffic. At the same time, the network traffic at this time is a kind of metadata (MD, meta data), so determining that the network traffic is a leaf node can be understood as defining the MD that needs to be replicated as a Leaf node.

[0087] In another feasible technical solution, if the MD corresponding to the network traffic at this time is not a Leaf node, the MD is directly sent to a high priority scheduling queue.

[0088] 202. If the hardware performance of the register reaches a preset performance condition, the network traffic identified as traffic to be copied is input into a low priority scheduling queue of the register to copy the network traffic.

[0089] Therefore, based on step 201, when the hardware performance of the register reaches the preset performance condition, the network traffic identified as the traffic to be copied can be input into the low priority scheduling queue of the register, so as to copy the network traffic.

[0090] In one of the specific embodiments, an mcc_pri register is provided in the PF, so it is necessary to determine whether the register at this time has reached the performance bottleneck, that is, the preset performance condition described above. Specifically, the preset performance condition (or performance bottleneck) is the maximum processing capacity of the hardware. For example, if the hardware cannot handle it, it will start to back-pressure the subsequent module, and the internal module will be able to sense that it cannot handle it. Therefore, the MD corresponding to the network traffic at this time is sent to the low-priority schedule queue (LSQ), so that in the subsequent process, the replication of network traffic or other work can be realized.

[0091] Furthermore, according to the above description, when the register reaches a performance bottleneck, sending the MD to the LSQ can also be understood as discarding the multicast message (network traffic).

[0092] 203. If the hardware performance of the register does not meet the preset performance condition, the network traffic identified as the traffic to be copied is input into the high priority scheduling queue of the register to copy the network traffic.

[0093] Therefore, based on step 201, when the hardware of the register fails to meet the preset performance condition, the network traffic identified as the traffic to be copied can be input into the high priority scheduling queue of the register, so as to copy the network traffic.

[0094] In one of the specific embodiments, when the register has not reached a performance bottleneck, that is, when the register has sufficient processing memory to handle traffic replication, the MD corresponding to the network traffic at this time is sent to a high-priority scheduling queue (HSQ, HIGH-priority schedule queue), so as to realize the replication of network traffic or other work in subsequent processes.

[0095] Furthermore, according to the above description, sending the MD to the HSQ can also be understood as giving priority to processing multicast messages (network traffic).

[0096] Through a traffic processing method disclosed in this embodiment,

[0097] Furthermore, replication traffic also supports clearing redundant hardware actions. Figure 3 , Figure 3 This is a flow chart of another traffic processing method disclosed in an embodiment of the present application, including steps 301 to 305.

[0098] 301. If the PF hits the flow table rule of the network communication protocol flow table, an action field table corresponding to the PF is obtained according to the flow table rule.

[0099] It should be noted that in this embodiment, the action of the original traffic will be copied after the replication, that is, during the traffic replication process. If the user does not need the action of the original traffic, the action can also be cleared. Specifically, this embodiment mainly supports clearing redundant actions in the replicated traffic.

[0100] In this embodiment, when PF hits the flow table rule of the network communication protocol flow table, the action field table corresponding to PF can be obtained according to the flow table rule. It should be noted that the action field table includes an action capability field and an action record field. The action capability field is used to describe the field for performing actions in the network communication protocol flow table, and the action record field is used to describe the field for recording actions in the coordinated action capability field.

[0101] In one of the specific embodiments, it should be noted that the hardware action is implemented through AB and AD. Among them, AB: actionbits, that is, the action capability field described above, which mainly describes the action capability set executed by the flow table, such as edit ip action, pushvlan action, traffic replication MCC action, etc. AD: action data, that is, the action record field described above, which mainly cooperates with AB to record the data of the action, such as ip edit action, in addition to having the editing capability, it is also necessary to know the specific edited ip data.

[0102] Furthermore, after hitting the flow table rule key of the OpenFLow flow table, the hardware PF will obtain the action, that is, AB+AD; at this time, the hardware traverses AB and will know the actions contained in the flow table. At the same time, the specific action corresponding data can also be obtained from AD. Specifically, AB is mainly used to describe the capability set of aciton, but the specific actions require the corresponding AB setting of the action. For example, the hardware supports uninstalling up to 6 actions: fwd (forwarding), ip_ped (ip editing), port_ped (port editing), mcc (traffic replication), push_vlan (adding VLAN) or pop_vlan (removing VLAN), etc. Then the capability set of AB is defined as follows: "mcc: 1; Ip_ped: 2; Port_ped: 2; fwd: 1; Push_vlan: 1; Pop_vlan: 1". That is, 8 bits can define all action capability sets.

[0103] 302. Traverse the action capability field to determine the execution action bit value corresponding to the execution action in the network communication protocol flow table.

[0104] Based on step 301, the action capability field can be traversed to determine the execution action bit value corresponding to the execution action in the network communication protocol flow table.

[0105] In one of the specific embodiments, it can be known from the above description that the execution action bit values ​​of various execution actions can be determined by the capability set of AB defined above. For example, if there are three actions when the user sends down the flow table, such as mcc+port_ped+pop_vlan. Then set MCC to 1, indicating that the traffic replication action is to be executed. The specific VFs to be copied are stored in AD, and AB does not care; port_ped is set according to the situation. If the source ip is edited, then port_ped is set to 1, Pop_vlan only needs to set AB to 1, and AD has no data, because this action only needs to tell the hardware whether it needs to be executed. If the destination ip is edited (the destination ip is the ip that the user needs to replace), then port_ped is set to 2, if the source ip + edit the destination ip, then port_ped is set to 3; pop_vlan is set to 1. It should be added that AB represents the action type and whether it needs to be executed. For example, if MCC is 1, it means that traffic replication needs to be executed. The specific action content copied by MCC can be found in AD. The same is true for Port_ped.

[0106] Thus, the execution action bit value of the execution action can be determined, thereby determining the action that needs to be performed subsequently on the traffic that currently needs to be copied.

[0107] 303. Determine execution action data corresponding to the execution action in the action record field according to the execution action bit value, so as to set the valid bit value of the index valid bit and the copy bit value of the traffic copy bit according to the execution action data.

[0108] Then, based on step 302, the execution action data corresponding to the execution action can be determined in the action record field according to the execution action bit value, so as to set the valid bit value of the index valid bit and the copy bit value of the traffic copy bit according to the execution action data.

[0109] In one of the specific embodiments, since AD ​​stores specific action data, the specific data is determined by AB, and the length of the action data is bound to the action type, the data of AD is compactly arranged.

[0110] For example, the example above has three actions: mcc+port_ped+pop_vlan. Since AB has set the three actions, AD will compactly arrange the action data of the queue, mcc_idx+port+null. The specific meanings are as follows:

[0111] mcc_idx: indicates the specific MCC table entry content, that is, the index position of the bitmap of mcc;

[0112] port: the specific port to be edited, for example, 10010;

[0113] null: empty data, because pop_vlan is just a pop action and does not require specific action data.

[0114] Thus, through the execution action data corresponding to the execution action, the valid bit value and the copy bit value (such as copying traffic and other actions) in the multicast member table can be set based on the execution action data. At the same time, further, other execution actions for processing network traffic can also be set.

[0115] 304. Determine the starting position of the action capability field.

[0116] In other feasible technical solutions, the MCC action may also be cleared. Specifically, the starting position of the action capability field may be determined first.

[0117] In one of the specific embodiments, since the MCC module provides a clear action interface (clear_action interface) for each MCC flow table, the user can control whether to clear redundant actions according to needs. Specifically, the mcc action is at the starting position of AB in hardware design. If the copied traffic does not want to perform the subsequent work, it can be controlled through the clear_act related configuration.

[0118] In one specific embodiment, the AB action is generally set behind the AB field. Therefore, by determining the starting position of the action capability field, the positions of all executed actions can be obtained accordingly.

[0119] It should be noted that the clear_act related configuration is as follows:

[0120] clear_act_offset1: clear action offset 1;

[0121] clear_act_len1: clear the action length of offset1;

[0122] clear_act_offset2: clear action offset 2;

[0123] clear_act_len2: Clear the action length of offset2.

[0124] 305. Configure a clear instruction field corresponding to the clear action interface at the starting position to clear the field corresponding to the execution action in the action capability field and the field corresponding to the record action in the action record field according to the clear instruction field.

[0125] Then, based on step 304, a clear instruction field corresponding to the clear action interface can be configured at the starting position, so as to clear the field corresponding to the execution action in the action capability field and the field corresponding to the record action in the action record field according to the clear instruction field.

[0126] In one specific embodiment, the related clear instruction field can be configured through clear_act at the starting position of AB, so as to realize the clearing of the field for executing the action and the field corresponding to the recorded action in the action record field.

[0127] Furthermore, if the user only wants to execute mcc without any additional actions, then configure clear_act_offset1 to 1, which means clearing starts from AB's ip_ped, and then configure clear_act_len to 7, which means clearing all actions after mcc.

[0128] There are two offsets here to prevent users from clearing multiple but non-consecutive actions, such as clearing ip_ped and then clearing a pop_vlan action. Then configure clear_act_offset1 to 1 and clear_act_len1 to 2; then configure clear_act_offset2 to 7 and clear_act_len2 to 1, and then you can clear multiple non-consecutive actions.

[0129] Through the traffic processing method disclosed in this embodiment, the starting position of the action to be cleared is realized by offset, and the number of consecutive actions to be cleared is realized by len. Multiple non-consecutive actions are cleared by multiple offsets, which improves the feasibility of the solution. At the same time, flexible action configuration supports clearing redundant actions of duplicated traffic, and the processing flow is simpler and more efficient.

[0130] Furthermore, during the traffic replication process, performance can be improved by using reference counting. Figure 4 , Figure 4 This is a flow chart of another traffic processing method disclosed in an embodiment of the present application, including steps 401 to 407.

[0131] 401. When network traffic enters the PF, the network traffic is cached in the hardware buffer of the PF.

[0132] It should be noted that in this embodiment, during the traffic replication process, the performance can be improved by using the reference technology. At the same time, the replication of traffic does not copy the entire message (traffic), but references the original traffic. That is, the original traffic has a reference count. If n copies are made, the original message references -1 for each copy. The original message is released only when the reference is 0. The replication performance is improved by using the reference count.

[0133] Specifically, after the network traffic enters the PF, the network traffic is first cached in the hardware buffer of the PF.

[0134] In one specific embodiment, each network traffic (or message) is cached into a hardware buffer (FIFO, first in first on) after entering the hardware.

[0135] 402. Analyze the network traffic in the hardware buffer to obtain traffic information corresponding to the network traffic.

[0136] Then the network traffic in the hardware buffer is parsed to obtain the traffic information of the network traffic. It should be noted that the traffic information at least includes the traffic weight and traffic reference index of the network traffic, and the traffic reference index is associated with the index relationship of the target index table.

[0137] In one specific embodiment, the hardware can parse the network traffic to obtain basic information of the network traffic. For example, the five-tuple is stored in an MD, and each MD has a weight (the default weight is 1024). At the same time, the basic information also includes the source and destination of the network traffic, reflecting the index relationship between PF and VF. Furthermore, the basic information also includes the setting of determining whether the network traffic is a duplicate traffic, as described above, and the specifics are not limited here.

[0138] 403. Determine whether the network traffic is network traffic for traffic replication according to the traffic reference index.

[0139] Therefore, by parsing the traffic reference index, it can be determined whether the network traffic is the network traffic that needs to be replicated.

[0140] In one specific embodiment, for network traffic, the hardware (such as a chip therein, etc.) can analyze the settings corresponding to the network traffic through the traffic reference index, thereby determining whether the network traffic needs to be copied. The properties of the network traffic can also be analyzed, thereby executing step 404 or step 406 respectively.

[0141] 404. If the network traffic is network replication traffic, when any network replication traffic is replicated, the traffic weight of the network replication traffic is reduced by a preset weight value to obtain a next traffic weight.

[0142] Based on step 403, when the network traffic is network replication traffic and when any network replication traffic is completed, the traffic weight of the network replication traffic can be reduced by the preset weight value to obtain the next traffic weight. Thus, the cycle is executed, that is, in the next network traffic, the preset weight value is reduced under the next traffic weight until step 405 is satisfied.

[0143] Furthermore, if the current network traffic is the network traffic that needs to be replicated (network replication traffic), the traffic weight can be subtracted by 2 when the current network replication traffic is replicated, thereby obtaining the next traffic weight. For example, the default traffic weight is 1024, and the current network replication traffic is the first network traffic, so the traffic weight is also 1024. Then, if it is the network traffic of traffic replication, then the weight of each replicated message is subtracted by 2 after processing, that is, 1024-2=1022. And this cycle repeats. Therefore, according to the next traffic weight, step 405 or step 407 is executed.

[0144] 405. When the next traffic weight is zero, all the network replication traffic that has been replicated is released to complete the replication of the network traffic of any PF and VF.

[0145] Based on step 404, when the next traffic weight is zero, all the network replication traffic that has been replicated can be released, thereby completing the replication of the network traffic for any PF and VF.

[0146] In one of the specific embodiments, after the above cycle, when the traffic weight is 0 during a certain traffic replication process, all the network replication traffic that has completed replication can be released, thereby completing the replication process of the network traffic between PF and VF.

[0147] 406. If the network traffic is unicast traffic, the traffic weight of the unicast traffic is cleared to zero, and the unicast traffic is released to execute the traffic replication process of another network traffic.

[0148] Corresponding to step 403, when the network traffic is unicast traffic, the traffic weight of the unicast traffic can be cleared and the unicast traffic can be released, thereby executing the traffic replication process of another network traffic. It should be noted that unicast traffic is used to represent other network traffic except for the network traffic replication.

[0149] In one specific embodiment, when it is found that the network traffic at this time is unicast traffic, the traffic weight of the unicast traffic is directly reduced by 1024 each time. When the subsequent hardware scheduling finds that the weight is 0, the traffic will be released. Therefore, in the subsequent network traffic processing process, the replication process of other network traffic is executed. For details, please refer to Figure 1 In the illustrated embodiment, steps 101 to 104 are shown. That is, the other network traffic is other messages and traffic in other processing flows, which are not specifically limited here.

[0150] 407. When the network replication traffic is the last network traffic in the multicast member table, the traffic weight corresponding to the last network traffic is cleared, and all the network replication traffic that has been replicated is released to complete the replication of the network traffic for any PF and VF.

[0151] Based on step 404, when the network replication traffic is the last network traffic in the multicast member table, the traffic weight corresponding to the last network traffic is cleared, and all the network replication traffic that has been replicated is released, thereby completing the replication of the network traffic for any PF and VF.

[0152] In one specific embodiment, in the multicast member table, when the network traffic that needs to be copied is the last MD, the traffic weight at this time needs to be cleared. Specifically, the weight is subtracted by 1024-((n-1)*2), that is, the last copied network replication traffic, and the remaining traffic weights need to be reduced. Therefore, all network replication traffic can be released after replication is completed, thereby completing the replication of network traffic for any PF and VF.

[0153] Through the traffic processing method disclosed in this embodiment, the original traffic is released only after all traffic is copied, thereby reducing the copying of load traffic. At the same time, the overall processing flow is simpler and more efficient, occupies less hardware resources, and less resource occupation can meet large-scale scenarios and has more flexible configuration.

[0154] It should be understood that, although the various steps in the flowcharts involved in the above-mentioned embodiments are displayed in sequence according to the indication of the arrows, these steps are not necessarily executed in sequence according to the order indicated by the arrows. Unless there is a clear explanation in this article, the execution of these steps does not have a strict order restriction, and these steps can be executed in other orders. Moreover, at least a part of the steps in the flowcharts involved in the above-mentioned embodiments can include multiple steps or multiple stages, and these steps or stages are not necessarily executed at the same time, but can be executed at different times, and the execution order of these steps or stages is not necessarily to be carried out in sequence, but can be executed in turn or alternately with other steps or at least a part of the steps or stages in other steps.

[0155] See also Figure 5 , Figure 5 A schematic diagram of the structure of a traffic processing system disclosed in an embodiment of the present application.

[0156] The division unit 501 is used to divide each physical device PF with hardware functions into multiple virtual instance devices VF based on virtualization technology, and determine the virtual network interface VSI corresponding to any PF and any VF;

[0157] A determination unit 502 is used to determine a target index table according to all VSIs between all PFs and all VFs; wherein the target index table is used to characterize the index relationship between all PFs and all VFs;

[0158] A definition unit 503 is used to define the valid bit value of the index valid bit and the replication bit value of the traffic replication bit of the target index relationship between any PF and any VF in the target index table, so as to obtain a multicast member table corresponding to all PFs and all VFs; wherein the index valid bit is used to identify the valid rule of the target index relationship, and the traffic replication bit is used to characterize the replication rule of the network traffic of the VSI associated with the target index relationship;

[0159] The replication unit 504 is used to replicate the network traffic for any PF and VF according to the valid bit value and the replication bit value in the multicast member table to complete the replication process of the network traffic.

[0160] Exemplarily, the PF includes a register, and the system further includes: an input unit 505;

[0161] The determining unit 502 is specifically used to determine the leaf node of the network traffic marked as the valid bit in the valid bit value and the replication bit in the replication bit value in the multicast member table, and mark the leaf node identifier of the leaf node; wherein the leaf node identifier is used to indicate that the network traffic is the traffic to be replicated;

[0162] An input unit 505, configured to input the network traffic identified as traffic to be copied into a low priority scheduling queue of the register to copy the network traffic when the hardware performance of the register reaches a preset performance condition;

[0163] The input unit 505 is further configured to input the network traffic identified as traffic to be copied into a high priority scheduling queue of the register to copy the network traffic when the hardware performance of the register does not meet the preset performance condition.

[0164] Exemplarily, the system further includes: an acquisition unit 506;

[0165] The acquisition unit 506 is used to obtain an action field table corresponding to the PF according to the flow table rule when the PF hits the flow table rule of the network communication protocol flow table; wherein the action field table includes an action capability field and an action record field, the action capability field is used to describe the field for performing the action in the network communication protocol flow table, and the action record field is used to describe the field for recording the action in the action capability field;

[0166] The determination unit 502 is further used to traverse the action capability field and determine the execution action bit value corresponding to the execution action in the network communication protocol flow table;

[0167] The determination unit 502 is further configured to determine execution action data corresponding to the execution action in the action record field according to the execution action bit value, so as to set the valid bit value of the index valid bit and the replication bit value of the traffic replication bit according to the execution action data.

[0168] Exemplarily, the system further includes: a configuration unit 507;

[0169] The determination unit 502 is further used to determine the starting position of the action capability field;

[0170] Configuration unit 507 is used to configure a clear instruction field corresponding to a clear action interface at a starting position, so as to clear the field corresponding to the execution action in the action capability field and the field corresponding to the record action in the action record field according to the clear instruction field; wherein the clear instruction field is used to clear the offset length of the field for executing the action and the action length of the field for recording the action.

[0171] Exemplarily, the system further includes: a cache unit 508, a determination unit 509, and a release unit 510;

[0172] A cache unit 508, configured to cache the network traffic into a hardware buffer of the PF after the network traffic enters the PF;

[0173] The acquisition unit 506 is specifically used to parse the network traffic in the hardware buffer to obtain traffic information corresponding to the network traffic; wherein the traffic information at least includes a traffic weight and a traffic reference index of the network traffic, and the traffic reference index is associated with an index relationship of the target index table;

[0174] A judging unit 509, configured to judge whether the network traffic is network traffic for traffic replication according to the traffic reference index;

[0175] The acquisition unit 506 is further configured to, when the network traffic is network replication traffic, reduce the traffic weight of the network replication traffic by a preset weight value when any network replication traffic is replicated, obtain the next traffic weight, and execute the step of reducing the traffic weight of the network replication traffic by a preset weight value when any network replication traffic is replicated; wherein the network replication traffic is used to characterize the network traffic for which traffic replication is performed;

[0176] The release unit 510 is used to release all the network replication traffic that has been replicated when the next traffic weight is zero, so as to complete the replication of the network traffic of any PF and VF.

[0177] Exemplarily, the system further includes:

[0178] The release unit 510 is also used to clear the traffic weight corresponding to the last network traffic to zero when the network replication traffic is the last network traffic in the multicast member table, and release all the network replication traffic that has been replicated to complete the replication of the network traffic for any PF and VF.

[0179] Exemplarily, the method further comprises:

[0180] The release unit 510 is also used to clear the traffic weight of the unicast traffic and release the unicast traffic to execute the traffic replication process of another network traffic when the network traffic is unicast traffic; wherein the unicast traffic is used to characterize other network traffic except the network traffic replication.

[0181] See below Figure 6 , a schematic diagram of the structure of a flow processing device disclosed in an embodiment of the present application includes:

[0182] CPU 601, memory 605, input / output interface 604, wired or wireless network interface 603 and power supply 602;

[0183] The memory 605 is a temporary storage memory or a permanent storage memory;

[0184] The CPU 601 is configured to communicate with the memory 605 and execute the instructions in the memory 605 to perform the aforementioned Figures 1 to 4 The traffic processing method in any of the illustrated embodiments.

[0185] The embodiment of the present application also provides a chip system, characterized in that the chip system includes at least one processor and a communication interface, the communication interface and the at least one processor are interconnected through a line, and the at least one processor is used to run a computer program or instruction to execute the aforementioned Figures 1 to 4 The traffic processing method in any of the illustrated embodiments.

[0186] The embodiment of the present application also provides a computer-readable storage medium, the computer-readable storage medium includes instructions, when the instructions are executed on a computer, the computer executes the aforementioned Figures 1 to 4 The traffic processing method in any of the illustrated embodiments.

[0187] The present application also provides a computer program product including instructions, which, when executed on a computer, enables the computer to execute the aforementioned Figures 1 to 4 The traffic processing method in any of the illustrated embodiments.

[0188] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0189] In the several embodiments provided in the present application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be an indirect coupling or communication connection through some interfaces, devices or units, which can be electrical, mechanical or other forms.

[0190] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0191] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0192] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application is essentially or the part that contributes to the prior art or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions to enable a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, read-only memory), random access memory (RAM, random access memory), disk or optical disk and other media that can store program code.

Claims

1. A traffic processing method, characterized in that: Applied to a network device, the network device comprises at least two physical devices PF having hardware functions, the method comprising: Based on virtualization technology, each of the physical devices PF with hardware functions is divided into multiple virtual instance devices VF, and a virtual network interface VSI corresponding to any PF and any VF is determined; Determine a target index table according to all VSIs between all PFs and all VFs; wherein the target index table is used to characterize the index relationship between all PFs and all VFs; The valid bit value of the index valid bit and the replication bit value of the traffic replication bit of the target index relationship between any PF and any VF are defined in the target index table to obtain a multicast member table corresponding to all PFs and all VFs; wherein the index valid bit is used to identify the valid rule of the target index relationship, and the traffic replication bit is used to characterize the replication rule of the network traffic of the VSI associated with the target index relationship; In the multicast member table, the network traffic for any of the PFs and the VFs is replicated according to the valid bit value and the replication bit value to complete the network traffic replication process.

2. The flow processing method according to claim 1, characterized in that: The PF includes a register, and the replicating the network traffic for any of the PF and the VF according to the valid bit value and the replica bit value includes: Determine the leaf node of the network traffic in the multicast member table, which is marked as a valid bit in the valid bit value and as a copy bit in the copy bit value, and mark the leaf node identifier of the leaf node; wherein the leaf node identifier is used to indicate that the network traffic is traffic to be copied; If the hardware performance of the register reaches a preset performance condition, the network traffic identified as traffic to be copied is input into a low priority scheduling queue of the register to copy the network traffic; If the hardware performance of the register does not meet the preset performance condition, the network traffic identified as traffic to be copied is input into the high priority scheduling queue of the register to copy the network traffic.

3. The flow processing method according to claim 1, characterized in that: Before defining the valid bit value of the index valid bit and the replication bit value of the traffic replication bit of the target index relationship between any PF and any VF in the target index table, the method further includes: If the PF hits the flow table rule of the network communication protocol flow table, an action field table corresponding to the PF is obtained according to the flow table rule; wherein the action field table includes an action capability field and an action record field, the action capability field is used to describe the field for performing an action in the network communication protocol flow table, and the action record field is used to describe the field for recording an action in the action capability field; Traversing the action capability field to determine an execution action bit value corresponding to the execution action in the network communication protocol flow table; The execution action data corresponding to the execution action is determined in the action record field according to the execution action bit value, so as to set the valid bit value of the index valid bit and the copy bit value of the traffic copy bit according to the execution action data.

4. The flow processing method according to claim 3, characterized in that: The network communication protocol flow table includes at least a clear action interface, and the method further includes: Determining the starting position of the action capability field; A clear instruction field corresponding to the clear action interface is configured at the starting position to clear the field corresponding to the execution action in the action capability field and the field corresponding to the record action in the action record field according to the clear instruction field; wherein the clear instruction field is used to clear the offset length of the field for executing the action and the action length of the field for recording the action.

5. The flow processing method according to claim 1, characterized in that: The network traffic replication process includes: When the network traffic enters the PF, the network traffic is cached in a hardware buffer of the PF; Parsing the network traffic in the hardware buffer to obtain traffic information corresponding to the network traffic; wherein the traffic information at least includes a traffic weight and a traffic reference index of the network traffic, and the traffic reference index is associated with an index relationship of the target index table; Determining, according to the traffic reference index, whether the network traffic is network traffic for traffic replication; If the network traffic is network replication traffic, when any network replication traffic is replicated, the traffic weight of the network replication traffic is reduced by a preset weight value to obtain a next traffic weight, and the step of reducing the traffic weight of the network replication traffic by a preset weight value when any network replication traffic is replicated is performed; wherein the network replication traffic is used to characterize the network traffic for which traffic replication is performed; When the next traffic weight is zero, all the network replication traffic that has completed replication is released to complete the replication of the network traffic for any of the PF and the VF.

6. The flow processing method according to claim 5, characterized in that: After obtaining the next traffic weight, the method further includes: When the network replication traffic is the last network traffic in the multicast member table, the traffic weight corresponding to the last network traffic is cleared, and all the network replication traffic that has been replicated is released to complete the replication of the network traffic for any of the PF and the VF.

7. The flow processing method according to claim 5, characterized in that: The method further comprises: If the network traffic is the unicast traffic, the traffic weight of the unicast traffic is cleared to zero, and the unicast traffic is released to execute the traffic replication process of another network traffic; wherein the unicast traffic is used to characterize other network traffic except the network traffic replication.

8. A flow processing system, characterized in that: The network device includes at least two physical devices PF with hardware functions, and the system includes: A division unit, used for dividing each of the physical devices PF having hardware functions into a plurality of virtual instance devices VF based on virtualization technology, and determining a virtual network interface VSI corresponding to any PF and any VF; A determination unit, configured to determine a target index table according to all VSIs between all PFs and all VFs; wherein the target index table is used to characterize the index relationship between all PFs and all VFs; A definition unit, used to define the valid bit value of the index valid bit and the replication bit value of the traffic replication bit of the target index relationship between any PF and any VF in the target index table, so as to obtain a multicast member table corresponding to all PFs and all VFs; wherein the index valid bit is used to identify the valid rule of the target index relationship, and the traffic replication bit is used to characterize the replication rule of the network traffic of the VSI associated with the target index relationship; The replication unit is used to replicate the network traffic for any of the PFs and the VFs in the multicast member table according to the valid bit value and the replication bit value to complete the replication process of the network traffic.

9. A flow processing device, characterized in that: The device comprises: CPU, memory, input and output interfaces, wired or wireless network interfaces, and power supply; The memory is a short-term storage memory or a persistent storage memory; The central processing unit is configured to communicate with the memory and execute instruction operations in the memory to perform the traffic processing method described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium includes instructions, and when the instructions are executed on a computer, the computer executes the traffic processing method according to any one of claims 1 to 7.

Images