A flow scheduling method, device and related equipment

Abstract

The application relates to the technical field of intelligent computing centers, in particular to a traffic scheduling method and device and related equipment. The method is applied to a controller, and the method comprises the following steps: acquiring network topology information and host ARP information of a whole network, wherein the network topology information of the whole network comprises link relationships and interface information of outgoing and incoming interfaces included in each link; defining initial cost values of each link included in the network topology information of the whole network, wherein the size of the initial cost value of a link is inversely proportional to the maximum bandwidth value of the link; determining source and destination information of a to-be-trained task, and determining a target path for transmitting traffic of the to-be-trained task based on the source and destination information of the to-be-trained task, host ARP information and cost values of each link included in the network topology information of the whole network, wherein the target path comprises target links, and the cost values of the target links are less than or equal to preset values.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of intelligent computing centers, in particular to a traffic scheduling method and device and related equipment. BACKGROUND

[0002] In an intelligent computing data center network, the uplink and downlink bandwidth is 1:1, and there are many Equal-Cost Multi-Path routing (ECMP) links in the network. In the case of per-flow forwarding, the switch performs hash routing forwarding according to the five-tuple of the traffic and other calculation factors to achieve load balancing in the network. The traffic of multiple entries is hashed ECMP to the same exit. In the case that there is no traffic on other exits, the load sharing is unbalanced and the network is congested. That is, in the intelligent computing data center networking, the switch device hash or load is uneven, which causes network traffic congestion and cannot be solved. SUMMARY

[0003] The present application provides a traffic scheduling method, device and related equipment.

[0004] In a first aspect, the present application provides a traffic scheduling method applied to a controller, comprising:

[0005] obtaining network topology information and host ARP information of the whole network, wherein the network topology information of the whole network includes link relationship and interface information of each link;

[0006] defining an initial cost value of each link included in the network topology information of the whole network, wherein the size of the initial cost value of a link is inversely proportional to the maximum bandwidth value of the link;

[0007] determining source and destination information of a to-be-trained task, and based on the source and destination information of the to-be-trained task, the host ARP information and the cost value of each link included in the network topology information of the whole network, determining a target path for transmitting traffic of the to-be-trained task, wherein the target path includes a target link, and the cost value of the target link is less than or equal to a preset value.

[0008] Optionally, the controller maintains a current cost value of each link; after determining the target path for transmitting the traffic of the to-be-trained task, the method further comprises:

[0009] updating the current cost value of each target link included in the target path.

[0010] Optionally, the step of updating the current cost value of each target link included in the target path comprises:

[0011] adding 1 to the current cost value of each target link included in the target path.

[0012] Optionally, the step of determining the target path for transmitting the to-be-trained task traffic based on the source and destination information of the to-be-trained task, the host ARP information and the cost values of the links included in the whole-network network topology information comprises:

[0013] determining a first access device accessed by the source end and a second access device accessed by the destination end according to the source and destination information and the host ARP information;

[0014] determining at least one path between the first access device and the second access device according to the whole-network network topology information;

[0015] calculating path cost values of the at least one path respectively according to the cost values of the links included in the whole-network network topology information;

[0016] determining a path with a path cost value less than or equal to a preset value as the target path for transmitting the to-be-trained task traffic according to the path cost values of the at least one path.

[0017] Optionally, the method further comprises:

[0018] decrementing the current cost value of each target link included in the target path by 1 when the to-be-trained task training is detected to be completed and the target path is released.

[0019] In a second aspect, the present application provides a traffic scheduling device applied to a controller, the device comprising:

[0020] an acquisition unit configured to acquire whole-network network topology information and host ARP information, wherein the whole-network network topology information comprises link relationships and out and in interface information included in each link;

[0021] a definition unit configured to define initial cost values of each link included in the whole-network network topology information, wherein the size of the initial cost value of a link is inversely proportional to the maximum bandwidth value of the link;

[0022] a determination unit configured to determine source and destination information of a to-be-trained task and determine a target path for transmitting to-be-trained task traffic based on the source and destination information of the to-be-trained task, the host ARP information and the cost values of the links included in the whole-network network topology information, wherein the target path comprises target links, the cost values of the target links are less than or equal to a preset value.

[0023] Optionally, the controller maintains current cost values of each link; and the device further comprises:

[0024] The updating unit is configured to update a current cost value of each target link included in the target path.

[0025] Optionally, when the updating unit updates the current cost value of each target link included in the target path, the updating unit is further configured to:

[0026] increment the current cost value of each target link included in the target path by 1.

[0027] Optionally, when the determining unit determines the source and destination information of the to-be-trained task, and determines the target path for transmitting the to-be-trained task traffic based on the source and destination information of the to-be-trained task, the host ARP information, and the cost value of each link included in the global network topology information, the determining unit is specifically configured to:

[0028] determine a first access device accessed by the source end and a second access device accessed by the destination end according to the source and destination information and the host ARP information;

[0029] determine at least one path between the first access device and the second access device according to the global network topology information;

[0030] calculate a path cost value of the at least one path according to the cost value of each link included in the global network topology information;

[0031] determine a path with a path cost value less than or equal to a preset value as the target path for transmitting the to-be-trained task traffic according to the path cost value of the at least one path.

[0032] Optionally, the apparatus further includes a detecting unit.

[0033] When the detecting unit detects that the to-be-trained task is trained, the updating unit is further configured to decrement the current cost value of each target link included in the target path by 1.

[0034] In a third aspect, an embodiment of the present application provides a traffic scheduling apparatus, which includes:

[0035] a memory configured to store program instructions;

[0036] a processor configured to invoke the program instructions stored in the memory, and execute the steps of the method according to the obtained program instructions.

[0037] In a fourth aspect, an embodiment of the present application further provides a computer readable storage medium, which stores computer executable instructions, and the computer executable instructions are used to make the computer execute the steps of the method according to any one of the first aspect.

[0038] In summary, the flow scheduling method provided by the embodiments of the present application is applied to a controller, and the method comprises: acquiring network topology information of a whole network and host ARP information, wherein the network topology information of the whole network comprises link relationships and interface information of outgoing and incoming interfaces included in each link; defining initial cost values of each link included in the network topology information of the whole network, wherein the size of the initial cost value of a link is inversely proportional to the maximum bandwidth value of the link; determining source and destination information of a to-be-trained task, and based on the source and destination information of the to-be-trained task, the host ARP information, and the cost values of each link included in the network topology information of the whole network, determining a target path for transmitting flow of the to-be-trained task, wherein the target path comprises target links, and the cost values of the target links are less than or equal to a preset value.

[0039] By using the flow scheduling method provided by the embodiments of the present application, the controller maintains the cost values of each link in the network topology of the whole network, and when routing is performed, the total cost values of each path are calculated, the path with the minimum total cost value is selected as the forwarding path of the to-be-scheduled training flow, and network congestion caused by uneven device load is avoided. BRIEF DESCRIPTION OF DRAWINGS

[0040] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the drawings needed in the description of the embodiments of the present application or the prior art will be briefly introduced. Obviously, the drawings in the following description are only some embodiments described in the present application, and other drawings can also be obtained by those skilled in the art according to these drawings of the embodiments of the present application.

[0041] Figure 1 A flow scheduling schematic diagram;

[0042] Figure 2 A detailed flowchart of a flow scheduling method provided by the embodiments of the present application;

[0043] Figure 3 A structural schematic diagram of a flow scheduling device provided by the embodiments of the present application;

[0044] Figure 4 A hardware architecture schematic diagram of a flow scheduling device provided by the embodiments of the present application. DETAILED DESCRIPTION

[0045] The terminology used in the description of the application herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the application. As used in the description of the application and the claims, the singular forms "a", "an" and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise. It also will be understood that the term "and / or" as used herein refers to and encompasses any and all possible combinations of one or more of the associated listed items.

[0046] It is to be understood that the singular forms "a", "an", and "the" include plural referents unless the context clearly dictates otherwise. It is to be understood that the term "and / or" as used herein encompasses all possible combinations of one or more of the associated listed items and that terms describing conditional functions, such as "if," "when," "responding to," or "in response to," encompass both the positive conditions of the function as well as the negative.

[0047] In the intelligent computing data center network, the uplink and downlink bandwidth is 1:1, and there are many equal-cost multi-path (ECMP) links in the network. In the case of per-flow forwarding, the switch device performs hash routing and forwarding according to the five-tuple and other calculation factors of the traffic, so as to realize load balancing in the network. The traffic of multiple entries is hashed ECMP to the same exit. In the case that there is no traffic on other exits, the load sharing is unbalanced and the network is congested. As shown in the following figure, the red link indicates that two groups of training tasks (node1 and node2, node3 and node4) are scheduled to the same spine and the same exit interface. When the traffic is loaded to the same Leaf during training, the Spine to Leaf2 may be loaded to the same path (exit interface), which causes unbalanced traffic load and congestion of the routing on the spine. Figure 1

[0048] That is, in the current intelligent computing data center networking, the switch device hash or load is unbalanced, which causes network traffic congestion and cannot be solved. The embodiments of the application provide a scheme for planning a full-network forwarding path by a controller to solve network congestion caused by unbalanced load.

[0049] For example, referring to Figure 2 As shown in the following figure, a detailed flowchart of a traffic scheduling method provided by the embodiments of the application is shown. The method is applied to a controller, and the method comprises the following steps:

[0050] Step 200: Obtain full-network network topology information and host ARP information, wherein the full-network network topology information comprises link relationship and exit and entry interface information included in each link. ​

[0051] In the embodiments of the present application, the controller is connected to the network devices in the whole network, and the controller can normally acquire the network topology in the whole network and host ARP (Address Resolution Protocol) information through the Link Layer Discovery Protocol (LLDP). Meanwhile, the controller records the link relationship and ingress / egress interface information of each link included in the network topology in the whole network. The controller can perform a baseline processing on the topology, that is, the link information (link relationship, ingress / egress interface information of each link) and host ARP information are saved in the memory. In this way, the controller has the topology relationship and host ARP information with a global perspective.

[0052] In actual application, the controller can acquire the information that a server (node, for example, a network card of the server) accesses a specific access device interface through the host ARP information. The training task traffic refers to the traffic between two nodes (for example, node1->node2). In this way, the access device (specific interface of the access device) accessed by node1 and the specific interface of the access device accessed by node2 can be acquired according to the host ARP information.

[0053] Step 210: defining the initial cost value of each link included in the network topology information in the whole network.

[0054] The initial cost value of a link is inversely proportional to the maximum bandwidth value of the link.

[0055] In the embodiments of the present application, the Spine-Leaf two-layer network architecture is taken as an example for illustration. The network includes Spine1, Spine2, Leaf-1, Leaf-2 and Leaf-3, wherein:

[0056] The uplink interface 1 of Leaf-1 accesses the downlink interface 1 of Spine1, the uplink interface 2 accesses the downlink interface 2 of Spine1, the uplink interface 3 of Leaf-1 accesses the downlink interface 1 of Spine2, and the uplink interface 4 accesses the downlink interface 1 of Spine2.

[0057] The uplink interface 1 of Leaf-2 accesses the downlink interface 3 of Spine1, the uplink interface 2 accesses the downlink interface 4 of Spine1, the uplink interface 3 of Leaf-2 accesses the downlink interface 3 of Spine2, and the uplink interface 4 accesses the downlink interface 4 of Spine2.

[0058] Leaf-3's uplink interface 1 accesses Spine1's downlink interface 5, Leaf-3's uplink interface 2 accesses Spine1's downlink interface 6, Leaf-3's uplink interface 3 accesses Spine2's downlink interface 5, and Leaf-3's uplink interface 4 accesses Spine2's downlink interface 6.

[0059] The connection between each uplink interface of the Leaf device and each downlink interface of the Spine device is defined as a link.

[0060] When the network is initialized, the cost value of each link needs to be defined according to the link quality (such as the maximum available bandwidth) of each link. Preferably, if the maximum available bandwidth of each link is the same, the cost value of each link can be defined as the same value (such as 1). That is, the cost value (current cost value) of each link maintained in the memory of the controller is 1.

[0061] Step 220: determining source and destination information of a to-be-trained task, and determining a target path for transmitting traffic of the to-be-trained task based on the cost value of each link included in the host ARP information and the full-network network topology information according to the source and destination information of the to-be-trained task.

[0062] The target path includes a target link, and the cost value of the target link is less than or equal to a preset value.

[0063] In the embodiments of the present application, when the source and destination information of a to-be-trained task is determined, and a target path for transmitting traffic of the to-be-trained task is determined based on the cost value of each link included in the host ARP information and the full-network network topology information according to the source and destination information of the to-be-trained task, a preferred implementation manner is as follows:

[0064] According to the source and destination information and the host ARP information, a first access device accessed by the source end is determined, and a second access device accessed by the destination end is determined;

[0065] According to the full-network network topology information, at least one path between the first access device and the second access device is determined;

[0066] According to the cost value of each link included in the full-network network topology information, the path cost value of the at least one path is calculated respectively;

[0067] According to the path cost value of the at least one path, a path with a path cost value less than or equal to a preset value is determined as the target path for transmitting the traffic of the to-be-trained task.

[0068] In the embodiments of the present application, before starting a training task, the source and destination of the training task (source and destination information) can be imported by calling the interface of the controller through the platform. In this way, the controller can determine the first access device accessed by the source and the second device accessed by the destination through the host ARP information. Then, the controller determines at least one forwarding path (each link included in each forwarding path) from the first device to the second device through the network topology information. Finally, the controller determines the forwarding path with the minimum total cost value of each link as the target path for transmitting the training task traffic based on the current cost value of each link maintained locally.

[0069] For example, it is assumed that the forwarding path from the first device to the second device includes path 1 and path 2, path 1 includes link 1 and link 3, path 2 includes link 2 and link 4, the total cost value of link 1 and link 3 included in path 1 is 1+1=2, and the total cost value of link 2 and link 4 included in path 2 is 1+3=4. Then, path 1 is determined as the target path for forwarding the training task traffic.

[0070] In actual application, the controller transmits a high-priority forwarding flow table to each network device included in the target path, so that each network device forwards the training task traffic based on the forwarding flow table. After the training of the training task is completed, each network device is instructed to delete the corresponding forwarding flow table.

[0071] In the embodiments of the present application, since the controller maintains the current cost value of each link, after determining the target path for transmitting the training task traffic, the above traffic scheduling method can further include the following steps:

[0072] Updating the current cost value of each target link included in the target path.

[0073] Specifically, when updating the current cost value of each target link included in the target path, a preferred implementation manner is:

[0074] The current cost value of each target link included in the target path is added by 1.

[0075] Next, taking the selection of path 1 (including link 1 and link 3) as the target path for transmitting the training task traffic as an example, the current cost value of link 1 is 1, and the current cost value of link 3 is also 1. At this time, the controller adds 1 to the cost value of link 1 and updates it to 2. Similarly, the controller adds 1 to the current cost value of link 3 and updates it to 2.

[0076] In practical applications, when updating the current cost value of a link, a fixed value can be accumulated, or a custom setting can be made based on the traffic volume of the latest training task. For example, if the traffic of the latest training task is large, the accumulated value when updating the current cost value of the corresponding link will also be large. For instance, if the training task traffic is less than a certain value, it is accumulated by 1; if the training task traffic is greater than a certain value, it is accumulated by 2. In this embodiment, the specific implementation method is not limited.

[0077] Furthermore, in this embodiment of the application, when the training of the task to be trained is detected to be completed and the target path is released, the current cost value of each target link included in the target path is cumulatively reduced by 1.

[0078] That is, after a training task is completed and the path (each link) corresponding to the training task is released, the current cost value of each link needs to be subtracted from the value accumulated when carrying the training task. For example, if the accumulated value is 1, then the value subtracted is also 1.

[0079] Based on the same inventive concept as the above-described embodiments, see, for example, the following: Figure 3 The diagram shown is a structural schematic of a traffic scheduling device provided in an embodiment of this application. This device is applied to a controller and includes:

[0080] The acquisition unit 30 is used to acquire network topology information and host ARP information of the entire network, wherein the network topology information of the entire network includes the relationship between each link and the ingress and egress interface information of each link;

[0081] Definition unit 31 is used to define the initial cost value of each link included in the network topology information of the whole network, wherein the initial cost value of a link is inversely proportional to the maximum bandwidth value of the link;

[0082] The determining unit 32 is used to determine the source and destination information of the task to be trained, and based on the source and destination information of the task to be trained, the host ARP information and the cost values ​​of each link included in the network topology information of the whole network, determine the target path for transmitting the traffic of the task to be trained, wherein the target path includes the target link, and the cost value of the target link is less than or equal to a preset value.

[0083] Optionally, the controller maintains the current cost value for each link; the device further includes:

[0084] An update unit is used to update the current cost value of each target link included in the target path.

[0085] Optionally, when updating the current cost value of each target link included in the target path, the updating unit is further configured to:

[0086] The current cost value of each target link included in the target path is accumulated by 1.

[0087] Optionally, when the source and destination information of the to-be-trained task is determined, and the cost value of each link included in the host ARP information and the whole-network network topology information is determined, the determining unit 32 is specifically configured to:

[0088] According to the source and destination information and the host ARP information, a first access device accessed by the source end is determined, and a second access device accessed by the destination end is determined;

[0089] According to the whole-network network topology information, at least one path between the first access device and the second access device is determined;

[0090] According to the cost value of each link included in the whole-network network topology information, the path cost value of the at least one path is calculated respectively;

[0091] According to the path cost value of the at least one path, a path with a path cost value less than or equal to a preset value is determined as the target path for transmitting the to-be-trained task flow.

[0092] Optionally, the apparatus further includes a detecting unit:

[0093] When the detecting unit detects that the to-be-trained task training is completed, the target path is released, and the updating unit is further configured to accumulate the current cost value of each target link included in the target path by -1.

[0094] The above units can be one or more integrated circuits configured to implement the above method, for example: one or more application specific integrated circuits (ASICs), or one or more digital signal processors (DSPs), or one or more field programmable gate arrays (FPGAs), etc. For another example, when the above certain unit is implemented in the form of a processing element scheduling code, the processing element can be a general-purpose processor, such as a central processing unit (CPU) or other processor that can call program code. For another example, these units can be integrated together in the form of a system on a chip (SOC).

[0095] Further, the flow scheduling device provided by the embodiment of the present application, in terms of hardware, the hardware architecture diagram of the flow scheduling device can refer to Figure 4 As shown in the figure, the flow scheduling device can include a memory 40 and a processor 41,

[0096] The memory 40 is configured to store program instructions, and the processor 41 invokes the program instructions stored in the memory 40 and executes the above-mentioned method embodiments according to the obtained program instructions. The specific implementation and technical effects are similar, and will not be repeated here.

[0097] Optionally, the present application also provides a controller comprising at least one processing element (or chip) for executing the above-mentioned method embodiments.

[0098] Optionally, the present application also provides a program product, for example, a computer readable storage medium, which stores computer executable instructions for causing the computer to execute the above-mentioned method embodiments.

[0099] Here, the machine readable storage medium can be any electronic, magnetic, optical, or other physical storage apparatus, which can contain or store information, such as executable instructions, data, and the like. For example, the machine readable storage medium can be RAM (Random Access Memory), volatile memory, non-volatile memory, flash memory, storage drive (such as hard disk drive), solid state drive, any type of storage disk (such as optical disk, dvd, etc.), or similar storage medium, or combination thereof.

[0100] The system, device, module or unit illustrated in the above embodiments can be specifically implemented by a computer chip or entity, or by a product with certain function. A typical implementation device is a computer, and the specific form of the computer can be a personal computer, a laptop computer, a cellular phone, a camera phone, a smart phone, a personal digital assistant, a media player, a navigation device, an email transceiver device, a game console, a tablet computer, a wearable device, or a combination of any of these devices.

[0101] For the convenience of description, the above device is described as various units respectively described in terms of function. Of course, in the implementation of the present application, the functions of each unit can be implemented in the same or more software and / or hardware.

[0102] Those skilled in the art will appreciate that embodiments of the application can be devised for a method, a system, or a computer program product. Accordingly, the present application can be embodied in the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, embodiments of the present application can take the form of a computer program product on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROMs, optical storage devices, and the like) embodying computer readable program code.

[0103] The present application is described in reference to the flowchart illustrations and / or block diagrams of methods, apparatus (systems) and computer program products according to embodiments of the application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart illustrations and / or block diagrams. Figure 1 one or more functions specified in the flowchart illustrations and / or block diagrams. Figure 1 one or more functions specified in the flowchart illustrations and / or block diagrams.

[0104] Also, these computer program instructions can be stored in a computer readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer readable memory produce an article of manufacture including instructions which implement the functions specified in the flowchart illustrations and / or block diagrams. Figure 1 one or more functions specified in the flowchart illustrations and / or block diagrams. Figure 1 one or more functions specified in the flowchart illustrations and / or block diagrams.

[0105] The computer program instructions can also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart illustrations and / or block diagrams. Figure 1 one or more functions specified in the flowchart illustrations and / or block diagrams. Figure 1 one or more functions specified in the flowchart illustrations and / or block diagrams.

[0106] The above description is embodied in the application only and is not intended to limit the present application. Any modification, equivalent replacement and improvement made within the spirit and principle of the present application shall fall within the scope of the present application.

Claims

1. A method of flow scheduling, the method comprising: The method applied to a controller comprises: obtaining network topology information and host ARP information, wherein the network topology information comprises link relationships and interface information of each link; defining initial cost values of each link included in the network topology information, wherein the initial cost value of a link is inversely proportional to the maximum bandwidth value of the link; determining source and destination information of a to-be-trained task, and determining a target path for transmitting traffic of the to-be-trained task based on the source and destination information of the to-be-trained task, host ARP information and cost values of each link included in the network topology information, wherein the target path comprises target links, and the cost values of the target links are less than or equal to a preset value; the controller maintains current cost values of each link; after the target path for transmitting the traffic of the to-be-trained task is determined, the method further comprises: updating the current cost values of each target link included in the target path.

2. The method of claim 1, wherein, The step of updating the current cost values of each target link included in the target path comprises: adding 1 to the current cost values of each target link included in the target path.

3. The method according to any one of claims 1 to 2, wherein, The step of determining the target path for transmitting the traffic of the to-be-trained task based on the source and destination information of the to-be-trained task, host ARP information and cost values of each link included in the network topology information comprises: determining a first access device accessed by a source end and a second access device accessed by a destination end according to the source and destination information and host ARP information; determining at least one path between the first access device and the second access device according to the network topology information; calculating path cost values of the at least one path respectively according to the cost values of each link included in the network topology information; determining a path with a path cost value less than or equal to a preset value as the target path for transmitting the traffic of the to-be-trained task according to the path cost values of the at least one path.

4. The method of claim 2, wherein, The method further comprises: when the target path is released after it is detected that the to-be-trained task is trained, subtracting 1 from the current cost values of each target link included in the target path.

5. A flow scheduling apparatus characterized by comprising: The device applied to a controller comprises: an obtaining unit configured to obtain network topology information and host ARP information, wherein the network topology information comprises link relationships and interface information of each link; a defining unit configured to define initial cost values of each link included in the network topology information, wherein the initial cost value of a link is inversely proportional to the maximum bandwidth value of the link; a determining unit configured to determine source and destination information of a to-be-trained task, and determine a target path for transmitting traffic of the to-be-trained task based on the source and destination information of the to-be-trained task, host ARP information and cost values of each link included in the network topology information, wherein the target path comprises target links, and the cost values of the target links are less than or equal to a preset value; the controller maintains current cost values of each link; the device further comprises: An updating unit is configured to update a current cost value of each target link included in the target path.

6. The apparatus of claim 5, wherein, When determining a target path for transmitting the to-be-trained task traffic based on source and destination information of the to-be-trained task, host ARP information, and cost values of links included in a whole-network network topology, the determining unit is specifically configured to: determine a first access device accessed by a source end and a second access device accessed by a destination end according to the source and destination information and the host ARP information; determine at least one path between the first access device and the second access device according to the whole-network network topology information; calculate path cost values of the at least one path according to the cost values of the links included in the whole-network network topology information; determine a path with a path cost value less than or equal to a preset value as the target path for transmitting the to-be-trained task traffic according to the path cost values of the at least one path.

7. A flow scheduling apparatus characterized by comprising: The traffic scheduling device comprises: a memory configured to store program instructions; a processor configured to invoke the program instructions stored in the memory and execute the steps of the method according to any one of claims 1-4.

8. A computer-readable storage medium, characterized in that, The computer readable storage medium stores computer executable instructions for causing the computer to execute the steps of the method according to any one of claims 1-4.

Images