WO2024011957A1

WO2024011957A1 - Traffic scheduling method, apparatus, device, and computer readable storage medium

Info

Publication number: WO2024011957A1
Application number: PCT/CN2023/085081
Authority: WO
Inventors: 单卫华; 高天宇
Original assignee: 华为云计算技术有限公司
Priority date: 2022-07-15
Filing date: 2023-03-30
Publication date: 2024-01-18

Abstract

Disclosed in the present application are a traffic scheduling method, an apparatus, a device, and a computer readable storage medium, belonging to the technical field of communications. The method applied to a scheduler comprises: first, acquiring priorities of at least two service traffics on a first node and the bandwidth occupied by the at least two service traffics on the first node, the transmission directions of the at least two service traffics being the same; afterwards, on the basis of the priorities of the at least two service traffics, determining a target service traffic among the at least two service traffics; and then, on the basis of the bandwidth occupied by the at least two service traffics on the first node, adjusting ratios of target service traffic interaction of the first node and the second node with a terminal. In the present application, a scheduled main service traffic, which is the target service traffic, is determined on the basis of the priorities of the service traffics, and the target service traffic is scheduled on the basis of the bandwidth occupied by each service traffic on each node. Therefore, each service traffic is scheduled in a unified and cooperative manner among the nodes, thus achieving a dynamic balance of the bandwidth occupied by each service traffic.

Description

Traffic scheduling method, device, equipment and computer-readable storage medium

This application requests the priority of the Chinese patent application with application number 202210836899.5 and the invention name "Greedy Multiplexing Bandwidth Cost Optimization Method, Device, Equipment and Medium" submitted on July 15, 2022. This application requests priority in 2022 The priority of the Chinese patent application with application number 202211146275.7 and the invention title "Traffic Scheduling Method, Device, Equipment and Computer-Readable Storage Medium" submitted on September 20, 2020, the entire content of which is incorporated into this application by reference.

Technical field

This application relates to the field of communication technology, and in particular to traffic scheduling methods, devices, equipment and computer-readable storage media.

Background technique

With the development of communication technology, information services are becoming more and more common. In communication networks, the information involved in information services is carried through business traffic, so the nodes included in the communication network need to transmit increasing business traffic. Among them, node transmission of business traffic depends on bandwidth. How to schedule business traffic based on bandwidth has become an issue worthy of attention.

Contents of the invention

This application provides a traffic scheduling method, device, equipment and computer-readable storage medium to schedule business traffic based on bandwidth. The technical solution provided by this application includes the following aspects.

In a first aspect, a traffic scheduling method is provided, which method is applied to a scheduler included in a traffic scheduling system. The traffic scheduling system also includes a first node, a second node and a terminal. Wherein, the first node and the second node are used to exchange service traffic with the terminal. In this method, the scheduler first obtains the priorities of at least two service flows on the first node, and the bandwidth occupied by the at least two service flows on the first node, and the transmission direction of the at least two service flows is the same. Then, the scheduler determines the target service flow among the at least two service flows based on the priorities of the at least two service flows. Then, based on the bandwidth occupied by at least two service flows on the first node, the scheduler adjusts the proportion of the target service flows of the first node and the second node interacting with the terminal.

This application distinguishes the priorities of different service flows, and determines the main service flow for scheduling based on the priority of the service flow, that is, the target service flow. In addition, the bandwidth occupied by each business flow is globally monitored, and when global collaborative scheduling is required, the target business flow is scheduled by adjusting the proportion of the first node and the second node interacting with the terminal. Since priority is considered in the scheduling process and global collaborative scheduling is performed, it not only helps improve the bandwidth reuse rate and reduce bandwidth costs, but also makes the scheduling process more flexible.

In a possible implementation, the scheduler determines the target service flow in at least two service flows based on the priorities of at least two service flows, including: the scheduler determines the target service flow in at least two service flows based on the priorities of at least two service flows. Determine at least one service flow with the lowest priority among the service flows; the scheduler determines the target service flow based on at least one service flow with the lowest priority.

This application determines the service flow with lower priority as the target service flow. Since the target service flow is the main scheduled service flow, the actual service quality of the service flow with higher priority can be guaranteed during the scheduling period. As a result, the actual service quality can meet the user's requirements, or in other words, the changes in the actual service quality can be at the user's discretion. Within the tolerable range, the user experience is guaranteed.

In a possible implementation, the scheduler determines the target service flow based on at least one service flow with the lowest priority, including: in response to the presence of at least two service flows with the lowest priority, the scheduler assigns at least two service flows with the lowest priority The business traffic in which the data volume is greater than the data volume threshold is determined as the target business traffic.

This application uses business traffic with a large amount of data as the target business traffic, so that the scheduler can schedule the target business traffic to a greater extent during scheduling. That is, after completing the scheduling, the bandwidth occupied by the target business traffic can be greater. changes, thereby providing a larger scheduling space during scheduling.

In a possible implementation, the priority is determined based on at least one information from a service level agreement (SLA) and a connection type. The SLA is used to indicate service quality, and the priority is positively related to the service quality indicated by the SLA. , the priority is negatively related to the scheduling convenience corresponding to the connection type.

The higher the service quality indicated by the SLA, the higher the service quality requirements that the service traffic needs to meet, and the lower the user's tolerance for the degradation of the service quality of the service traffic. Therefore, the impact on the actual service quality of the service flow during scheduling should be smaller, so the priority of the service flow is higher, thereby reducing the possibility that the scheduler determines the service flow as the target service flow. In addition, the greater the scheduling convenience corresponding to the connection type of the business traffic, the higher the flexibility in scheduling the business traffic. Therefore, the lower the priority of the service flow, the more likely it is to determine the service flow as the target service flow.

In a possible implementation, the scheduler adjusts the proportion of the first node and the second node to interact with the terminal based on the bandwidth occupied by at least two service flows on the first node, including: the scheduler determines at least two The sum of the bandwidths occupied by each business flow on the first node; the scheduler determines the proportion adjustment method based on the sum of bandwidths; the scheduler adjusts the first node and the second node to interact with the terminal according to the proportion adjustment method. Proportion.

This application determines a proportional adjustment method based on the determined sum of bandwidths. This proportional adjustment method is the basis of the proportional adjustment process. The proportional adjustment process is also the process of scheduling target service traffic.

In a possible implementation, the scheduler determines the proportion adjustment method based on the sum of bandwidths, including: in response to the sum of bandwidths meeting the first condition, the scheduler determines that the proportion adjustment method is to reduce the first node and the second node and the terminal The proportion of the target business traffic is interacted such that the bandwidth occupied by the target business traffic on the first node is reduced by the bandwidth scheduling amount, and the bandwidth occupied by the target business traffic on the second node is increased by the bandwidth scheduling amount; the method also includes: increasing the first business traffic The bandwidth occupied on the first node, the first service flow is the service flow other than the target service flow among the at least two service flows, and the increased bandwidth is less than or equal to the bandwidth scheduling amount.

Among them, if the sum of bandwidths meets the first condition, it means that the first node is in the peak period of business traffic and has insufficient bandwidth, so the proportion adjustment method is used to indicate a reduction in the proportion. That is to say, the bandwidth occupied by the target business flow on the first node is reduced, and the bandwidth occupied by the target business flow on the second node is increased, and the bandwidth occupied by the first business flow other than the target business flow on the first node is also increased. Bandwidth increases. Thus, the service quality of each service flow on the first node and the second node is guaranteed, and the bandwidth on both the first node and the second node can be fully utilized.

In a possible implementation, the method further includes: in response to the sum of bandwidths being greater than a first bandwidth threshold, determining that the sum of bandwidths satisfies the first condition; or in response to the bandwidth cost corresponding to the sum of bandwidths being greater than the first cost threshold, Determine that the sum of bandwidths satisfies the first condition.

Whether the sum of bandwidths is greater than the first bandwidth threshold or the bandwidth cost corresponding to the sum of bandwidths is greater than the first cost threshold, it means that the first node is in the peak period of business traffic and has insufficient bandwidth, so it can be determined that the first condition is met.

In a possible implementation, the scheduler determines the proportion adjustment method based on the sum of bandwidths, including: responding to the bandwidth The sum of the widths satisfies the second condition, and the scheduler determines the proportion adjustment method to increase the proportion of the target service traffic of the first node and the second node interacting with the terminal, so that the bandwidth occupied by the target service traffic on the first node increases the bandwidth scheduling amount, and The bandwidth occupied by the target service traffic on the second node is reduced by the bandwidth scheduling amount.

Among them, if the sum of bandwidths meets the second condition, it means that the first node is in the low period of business traffic and has sufficient bandwidth, so the proportion adjustment method is used to indicate an increase in the proportion. That is, the bandwidth occupied by the target business flow on the first node is increased, and the bandwidth occupied by the target business flow on the second node is reduced. Thus, the service quality of each service flow on the first node and the second node is guaranteed, and the bandwidth on the first node and the second node can be fully utilized.

In a possible implementation, the method further includes: in response to the sum of bandwidths being less than or equal to a second bandwidth threshold, determining that the sum of bandwidths satisfies the second condition and the second bandwidth threshold is less than or equal to the first bandwidth threshold; or, responding If the bandwidth cost corresponding to the sum of bandwidths is less than or equal to the second cost threshold, it is determined that the sum of bandwidths satisfies the second condition and the second cost threshold is less than or equal to the first cost threshold.

Whether the sum of bandwidths is less than the second bandwidth threshold, or the bandwidth cost corresponding to the sum of bandwidths is less than the second cost threshold, it means that the first node is in a low period of business traffic and has excess bandwidth, so it can be determined that the second condition is met.

In a possible implementation, the bandwidth scheduling amount is determined based on the sum of bandwidths, the first bandwidth threshold, and the bandwidth usage of the second node.

The relationship between the sum of bandwidths and the first bandwidth cost threshold can reflect the amount of bandwidth that can be scheduled on the first node. The bandwidth usage of the second node can reflect the amount of bandwidth that can be scheduled on the second node. Therefore, the bandwidth scheduling amount determined in this way is relatively accurate, which is conducive to ensuring the service quality of each service flow on the first node and the second node after the scheduling is completed.

In a possible implementation, the bandwidth scheduling amount is determined based on the bandwidth cost corresponding to the sum of bandwidths, the first cost threshold, and the bandwidth cost usage of the second node.

The relationship between the bandwidth cost corresponding to the sum of bandwidths and the first cost threshold can reflect the amount of bandwidth that can be scheduled on the first node. The bandwidth cost usage of the second node can reflect the amount of bandwidth that can be scheduled on the second node. Therefore, the bandwidth scheduling amount determined in this way is relatively accurate, which is beneficial to reducing bandwidth costs and ensuring the service quality of each service flow on the first node and the second node.

In a possible implementation, the scheduler determines the proportion adjustment method based on the sum of bandwidths, including: in response to the sum of bandwidths being different from the bandwidth occupied by the second service traffic on the first node, the scheduler determines the proportion adjustment method, The transmission directions of at least two service flows are different from the transmission direction of the second service flow.

In this implementation, the bandwidth occupied by business traffic in different transmission directions on the first node is different. Therefore, the bandwidth occupied by the target business traffic on the first node can be changed so that the bandwidth occupied by business traffic in different transmission directions is the same. , or narrow the gap between the bandwidth occupied by business traffic in different transmission directions.

In a possible implementation manner, the proportion adjustment method is determined based on the sum of bandwidth, the bandwidth occupied by the second service traffic on the first node, and the bandwidth usage of the second node.

The proportion adjustment method determined in this way is more accurate.

In a possible implementation manner, the scheduler adjusts the proportion of the target service traffic of the first node and the second node interacting with the terminal in a proportional adjustment manner, including: the scheduler determines the scheduling information corresponding to the target node in a proportional adjustment manner, and provides the The target node sends scheduling information, and the scheduling information and the target node are used to adjust the ratio of the first node, the second node, and the terminal to interact with the target service traffic.

That is to say, for different target nodes, the scheduling information corresponding to the target node is adaptively determined, so that The adjustment of the proportion based on the target node and scheduling information is to achieve the scheduling of target business traffic, thereby ensuring the service quality of each business traffic and the user experience.

In one possible implementation, the target node is a content delivery network (CDN) domain name system (DNS) server, and the scheduling information is between the reference domain name and the Internet Protocol (IP) address. The first correspondence relationship, the IP address indicates the first node or the second node;

The first correspondence relationship is used by the CDN DNS server to return the IP address corresponding to the reference domain name to the terminal based on the reference domain name sent by the terminal. The IP address is used by the terminal to send interaction requests to the node indicated by the IP address to interact with target business traffic.

By sending the first corresponding relationship to the CDN DNS server, the CDN DNS server can interact with the terminal based on the first corresponding relationship, so that the terminal determines which node needs to interact with the target business traffic through the interaction process, thereby realizing the scheduling of the target business traffic.

In a possible implementation, the target nodes are the first node and the second node, the scheduling information is the second correspondence between the interaction request and the IP address, the IP address indicates the first node or the second node; the second correspondence The relationship is used for the target node to interact with the target service traffic with the terminal in response to determining that the node indicated by the IP address corresponding to the interaction request is the current node based on the second correspondence relationship after receiving the interaction request sent by the terminal, or in response to the response based on the second corresponding relationship. If the second correspondence determines that the node indicated by the IP address corresponding to the interaction request is not the current node, the IP address corresponding to the interaction request is returned to the terminal. The IP address is used by the terminal to send the interaction request to the node indicated by the IP address to exchange target business traffic.

Sending the second correspondence relationship to the first node and the second node enables both the first node and the second node to determine which node the terminal needs to interact with the target service traffic based on the second correspondence relationship. If the node that needs to exchange target service traffic with the terminal is the node itself, then exchange the target service flow with the terminal. If the node with which the terminal needs to interact with the target service traffic is not the node itself, the terminal is informed with which node it needs to interact with the target service traffic, so that the terminal can interact with the correct node with the target service traffic. As a result, the scheduling of target business traffic is achieved.

In a possible implementation, the target node is a terminal, and the scheduling information is a second correspondence between the interaction request and the IP address, where the IP address indicates the first node or the second node; the second correspondence is used by the terminal to determine the interaction Request the corresponding IP address and send an interaction request to the node indicated by the IP address to exchange target business traffic.

Sending the second correspondence between the interaction request and the IP address to the terminal enables the terminal to determine whether to interact with the first node for target service traffic or with the second node based on the second correspondence, thereby achieving the target service. Traffic scheduling.

In a possible implementation manner, the bandwidth occupied by at least two service flows on the first node is obtained by at least one of detection and prediction.

Through detection, the bandwidth occupied by at least two business flows on the first node can be monitored in real time, so as to promptly determine the proportion adjustment method and schedule it in a timely manner to ensure the service quality of each business flow on the first node and the user experience. feel. Through prediction, the bandwidth occupied by at least two business flows on the first node can be known in advance, the proportion adjustment method can be determined in a more timely manner, and scheduling can be carried out in advance, further ensuring the service quality of each business flow on the first node. quality and user experience.

In a possible implementation, after the scheduler adjusts the ratio of the first node and the second node to interact with the terminal based on the bandwidth occupied by at least two service flows on the first node, the method also includes: responding to The adjusted ratio makes the bandwidth occupied by the target service flow on the first node equal to zero, and the service quality of the first service flow is lower than the service quality indicated by the SLA of the first service flow. The scheduler issues an alarm, and the first service flow is At least two service flows other than the target service flow.

Wherein, if the bandwidth occupied by the target service flow on the first node is zero, it means that the target service flow no longer exists on the first node, and all the bandwidth on the first node is occupied by the first service flow. However, the actual service quality of the first service quality is still lower than the service quality indicated by the SLA of the first service flow, and the bandwidth on the first node is still insufficient. Therefore, the scheduler can use an alarm to prompt that scheduling needs to continue, so as to ensure that the actual service quality of the first service flow meets the service quality indicated by the SLA of the first service flow.

In a possible implementation, before the scheduler adjusts the ratio of the first node and the second node to interact with the terminal based on the bandwidth occupied by at least two service flows on the first node, the method also includes: the scheduler Send the priorities of at least two business flows to the first node. The priorities of the at least two business flows are used by the first node to configure at least two queues and process at least two business flows through the at least two queues. For at least two services For any business flow in the traffic, the priority of any business flow matches the priority of the queue used to process any one business flow.

By configuring at least two queues and making the priority of a service flow match the priority of the queue used to process the service flow, it is possible to detect a sudden increase or decrease in the bandwidth occupied by each service flow on the first node. , ensuring that business traffic with higher priority is processed first.

In a second aspect, a traffic scheduling device is provided. The device is applied to a scheduler included in a traffic scheduling system. The traffic scheduling system also includes a first node, a second node and a terminal. The first node and the second node are used to interact with the terminal. Business traffic, devices include:

The acquisition module is used to obtain the priorities of at least two business flows on the first node, and the bandwidth occupied by the at least two business flows on the first node, and the transmission direction of the at least two business flows is the same;

a determining module, configured to determine the target service flow among the at least two service flows based on the priorities of the at least two service flows;

The adjustment module is configured to adjust the proportion of the target service traffic of the first node and the second node interacting with the terminal based on the bandwidth occupied by at least two service flows on the first node.

In a possible implementation, the determining module is configured to determine at least one service flow with the lowest priority among the at least two service flows based on the priorities of the at least two service flows; based on at least one service flow with the lowest priority Determine target business traffic.

In a possible implementation, the determining module is configured to, in response to the presence of at least two lowest priority service flows, determine as a target the service flow of the at least two lowest priority service flows whose data amount is greater than the data amount threshold. business traffic.

In a possible implementation, the priority is determined based on at least one information of SLA and connection type. SLA is used to indicate service quality. The priority is positively related to the service quality indicated by SLA. The priority is related to the scheduling corresponding to the connection type. Convenience is negatively related.

In a possible implementation, the adjustment module is used to determine the sum of bandwidths occupied by at least two business flows on the first node; determine a proportional adjustment method based on the sum of bandwidths; and adjust the proportional adjustment method according to the proportional adjustment method. The proportion of the target service traffic of the first node and the second node interacting with the terminal.

In a possible implementation, the adjustment module is configured to, in response to the sum of bandwidths meeting the first condition, determine the proportion adjustment method to reduce the proportion of the target service traffic of the first node and the second node interacting with the terminal, so that the target service traffic The bandwidth occupied on the first node reduces the bandwidth scheduling amount, and the bandwidth occupied by the target business traffic on the second node increases the bandwidth scheduling amount;

The device also includes: an adding module for increasing the bandwidth occupied by the first service flow on the first node. The first service flow is the service flow other than the target service flow among the at least two service flows, and the increased bandwidth is less than or equal to equal to the bandwidth adjustment measure.

In a possible implementation, the adjustment module is further configured to, in response to the sum of bandwidths being greater than the first bandwidth threshold, determine that the sum of bandwidths satisfies the first condition; or, in response to the sum of bandwidths, the corresponding bandwidth cost being greater than the first cost. threshold to determine that the sum of bandwidths meets the first condition.

In a possible implementation, the adjustment module is configured to determine the proportion adjustment method in response to the sum of bandwidths meeting the second condition to increase the proportion of the target service traffic of the first node and the second node interacting with the terminal, so that the target service traffic The bandwidth occupied on the first node increases the bandwidth scheduling amount, and the bandwidth occupied by the target service traffic on the second node decreases the bandwidth scheduling amount.

In a possible implementation, the adjustment module is further configured to, in response to the sum of the bandwidths being less than or equal to the second bandwidth threshold, determine that the sum of the bandwidths satisfies the second condition and the second bandwidth threshold is less than or equal to the first bandwidth threshold; or , in response to the bandwidth cost corresponding to the sum of bandwidths being less than or equal to the second cost threshold, it is determined that the sum of bandwidths satisfies the second condition, and the second cost threshold is less than or equal to the first cost threshold.

In a possible implementation, the adjustment module is configured to determine a proportional adjustment method in response to the sum of bandwidths being different from the bandwidth occupied by the second business traffic on the first node, so that the transmission directions of at least two business flows are consistent with the second Business traffic is transmitted in different directions.

In a possible implementation, the adjustment module is used to determine the scheduling information corresponding to the target node in a proportional adjustment manner, and send the scheduling information to the target node. The scheduling information and the target node are used to adjust the relationship between the first node and the second node. The proportion of terminal interaction target business traffic.

In a possible implementation, the target node is a CDN DNS server, and the scheduling information is the first correspondence between the reference domain name and the IP address, where the IP address indicates the first node or the second node; the first correspondence is used for the CDN The DNS server returns the IP address corresponding to the reference domain name to the terminal based on the reference domain name sent by the terminal. The IP address is used by the terminal to send interaction requests to the node indicated by the IP address to exchange target business traffic.

In a possible implementation, the adjustment module is also configured to respond to the adjusted ratio so that the bandwidth occupied by the target service flow on the first node is zero, and the service quality of the first service flow is lower than the first service flow. Based on the service quality indicated by the SLA, an alarm is issued, and the first service flow is the service flow other than the target service flow among at least two service flows.

In a possible implementation, the adjustment module is also configured to send the priorities of at least two business flows to the first node, and the priorities of the at least two business flows are used by the first node to configure at least two queues, through at least The two queues process at least two service flows, and for any one of the at least two service flows, the priority of any one service flow matches the priority of the queue used to process any one service flow.

In a third aspect, a traffic scheduling device is provided. The device includes a memory and a processor; at least one instruction is stored in the memory, and at least one instruction is loaded and executed by the processor, so that the traffic scheduling device realizes the first aspect and the corresponding possibilities. Any traffic scheduling method in the implementation.

In a fourth aspect, a computer-readable storage medium is provided. At least one instruction is stored in the computer-readable storage medium, and the instruction is loaded and executed by the processor, so that the computer implements as in the first aspect and the corresponding possible implementation manners. Any traffic scheduling method.

In the fifth aspect, a computer program or computer program product is provided. The computer program or computer program product includes a computer program or instructions. The computer program or instructions are executed by a processor to enable the computer to implement the first aspect and corresponding possible implementation methods. Any traffic scheduling method.

In a sixth aspect, a chip is provided, including a processor for calling and running instructions stored in the memory, so that the computer installed with the chip executes any flow of the first aspect and corresponding possible implementations. Scheduling method.

In the seventh aspect, another chip is provided, including: an input interface, an output interface, a processor and a memory. The input interface, the output interface, the processor and the memory are connected through an internal connection path. The processor is used to execute the code in the memory. When the code is executed, the computer installed with the chip executes any one of the traffic scheduling methods of the first aspect and corresponding possible implementations.

It should be understood that for the technical effects achieved by the technical solutions of the second to seventh aspects of the present application and the corresponding possible implementations, please refer to the above-mentioned technologies achieved by the first aspect and the corresponding possible implementations. The effect will not be described here.

Description of drawings

Figure 1 is a schematic diagram of a traffic scheduling system provided by an embodiment of the present application;

Figure 2 is a flow chart of a traffic scheduling method provided by an embodiment of the present application;

Figure 3 is a schematic diagram of a reference correspondence relationship provided by an embodiment of the present application;

Figure 4 is a schematic diagram of scheduling target service traffic under a reduced ratio provided by an embodiment of the present application;

Figure 5 is a schematic diagram of scheduling target service traffic at an increased ratio provided by an embodiment of the present application;

Figure 6 is a schematic flowchart of a traffic scheduling method provided by an embodiment of the present application;

Figure 7 is a schematic flowchart of another traffic scheduling method provided by an embodiment of the present application;

Figure 8 is a schematic diagram illustrating different bandwidths occupied by business traffic in different transmission directions provided by an embodiment of the present application;

Figure 9 is a schematic structural diagram of a scheduler provided by an embodiment of the present application;

Figure 10 is a schematic diagram of traffic scheduling in a DNS scheduling scenario provided by an embodiment of the present application;

Figure 11 is a schematic diagram of traffic scheduling in a 302 scheduling scenario provided by the embodiment of the present application;

Figure 12 is a schematic diagram of traffic scheduling in another 302 scheduling scenario provided by the embodiment of the present application;

Figure 13 is a schematic diagram of traffic scheduling in another 302 scheduling scenario provided by the embodiment of the present application;

Figure 14 is a schematic diagram of process scheduling in an application layer scheduling scenario provided by an embodiment of the present application;

Figure 15 is a schematic diagram of processing business traffic through queues provided by an embodiment of the present application;

Figure 16 is a schematic diagram of bandwidth allocation in a related technology provided by the embodiment of the present application;

Figure 17 is a schematic diagram of a traffic scheduling device provided by an embodiment of the present application;

Figure 18 is a schematic diagram of a traffic scheduling device provided by an embodiment of the present application.

Detailed ways

The terms used in the embodiments of the present application are only used to explain specific embodiments of the present application and are not intended to limit the present application.

The embodiment of the present application provides a traffic scheduling system. Refer to Figure 1. The traffic scheduling system includes a first node 11, a second node 12, a scheduler 13 and a terminal 14. The scheduler 13 is connected to the first node 11 and the second node 12 respectively, and the terminal 14 is connected to at least one of the first node 11 and the second node 12 . The first node 11 and the second node 12 may be used to exchange service traffic with the terminal 14 , and the scheduler 13 may be used to implement scheduling of service traffic on the first node 11 and the second node 12 .

Among them, the first node 11 and the second node 12 shown in Figure 1 may form a CDN, and the first node 11 and the second node 12 are also called edge nodes in the CDN. The CDN is an intelligent virtual network based on edge nodes deployed in various places. Based on the CDN, an intelligent virtual network, the terminal 14 can interact with business traffic with edge nodes that are relatively close to each other.

In the process of interaction of service traffic between the terminal 14 and the edge node, the service traffic has a certain transmission direction, and the transmission direction includes downlink and uplink. When the transmission direction of the service traffic is downlink, the edge node delivers the service traffic to the terminal 14, and the service traffic is downlink service traffic. Or, when the transmission direction of the service traffic is uplink, the terminal 14 uploads the service traffic to the edge node, and the service traffic is uplink service traffic.

It should be understood that the first node 11, the second node 12 and the terminal 14 shown in FIG. 1 are only examples and are not used to limit the number. In this embodiment of the present application, the sum of the numbers of the first nodes 11 and the second nodes 12 is at least two, and the number of terminals 14 is at least one.

Optionally, the traffic scheduling system shown in Figure 1 may also include a source server. For example, the origin server can connect to the edge node and exchange business traffic with the edge node through functions such as load balancing and content distribution. In the process of interaction of business traffic between the source server and the edge node, the business traffic also has a certain transmission direction. The business traffic delivered by the source server to the edge node is downlink business traffic, and the business traffic uploaded by the edge node to the source server is uplink business traffic. The scheduler 13 and the source server shown in Figure 1 may be two independent entities. Alternatively, the scheduler 13 can also be integrated with the source server. For example, scheduler 13 may be a component located in the origin server.

This application provides a traffic scheduling method, which can be applied to the scheduler 13 included in the traffic scheduling system shown in Figure 1. As shown in Figure 2, the method includes but is not limited to the following steps 201 to 203.

Step 201: The scheduler obtains the priorities of at least two service flows on the first node and the bandwidth occupied by the at least two service flows on the first node. The transmission directions of the at least two service flows are the same.

The first node can be any edge node in the CDN, and there are at least two service flows with the same transmission direction on the first node, and the transmission direction includes downlink and uplink. For example, the scheduler can, based on different transmission directions, All business traffic on the first node is divided into two groups. The transmission direction of one group of business traffic is all downlink, and the transmission direction of the other group of business traffic is all uplink. Afterwards, the scheduler treats all or part of the service flows included in the group of service flows as at least two service flows with the same transmission direction.

For example, see Figure 3. A group of business traffic with the transmission direction in the downlink direction includes: live broadcast service traffic, dynamic acceleration service traffic, CDN on-demand service traffic and CDN download service traffic, while a set of business traffic with the transmission direction in the uplink direction includes: live broadcast service traffic. traffic, dynamic acceleration business traffic and video access service (VIS) business traffic. Among them, VIS is a real-time, large-capacity, high-concurrency service. VIS has the capabilities of video data collection, video data distribution, and video data transfer.

For at least two service flows with the same transmission direction on the first node, the priorities of the at least two service flows are described.

In an exemplary embodiment, the priority is determined based on at least one information of SLA and connection type, and SLA is used to indicate service quality. Among them, the priority is positively related to the service quality indicated by the SLA, and the priority is negatively related to the scheduling convenience corresponding to the connection type.

The service quality indicated by the SLA of a business flow is the service quality promised by the communication service provider to the communication service users, that is, the service quality that a business flow needs to meet. The form of commitment is, for example, a contract. The higher the service quality indicated by the SLA of a service flow, the higher the communication service user's requirements for the service quality of the service flow and the lower the tolerance for the degradation of the service quality of the service flow. The service flow needs With stronger availability, the provider of the communication server needs to bear more responsibility for the business traffic.

The provider of the communication server is used to provide the resources required to transmit business traffic. For example, the provider of the communication server is an Internet service provider (ISP), and the resources required to transmit business traffic are, for example, bandwidth. Users of communication services are used to consume business traffic. For example, the user of the communication service is the user of the terminal. In an exemplary embodiment, indicators used to define an SLA for business traffic include, but are not limited to, at least one of the following indicators: mean time between failures (MTBF), mean time to repair (mean time to repair, MTTR), delay, jitter, throughput.

The connection type of a business flow is the type of communication connection established between the terminal and the edge node for exchanging the business flow. The connection type includes but is not limited to short connection and long connection. The short connection is established only when the interaction of the business traffic is required, and is disconnected after the interaction of the business traffic is completed. A long connection can be established before the interaction of the business traffic is required and will not be disconnected after the interaction of the business traffic is completed. For a business flow, if a short connection needs to be established between the terminal and the edge node to interact with the business flow, the connection type of the business flow is a short connection. The short connection corresponds to a greater degree of scheduling convenience, that is, it is convenient for scheduling and can Flexible scheduling. Or, if a long connection needs to be established between the terminal and the edge node to exchange the service traffic, the connection type of the service traffic is a long connection. The scheduling convenience corresponding to the long connection is less, that is, it is inconvenient to schedule and cannot be flexibly scheduled.

For example, the priority of a service flow is positively related to the service quality indicated by the SLA of the service flow. The reason is that the higher the service quality indicated by the SLA of a service flow, the higher the priority of exchanging the service flow, and therefore the higher the priority of the service flow. For example, the priority of a service flow is negatively related to the scheduling convenience corresponding to the connection type of the service flow. The reason is that the higher the scheduling convenience corresponding to the connection type of a service flow and the easier it is to schedule the service flow, the lower the priority of the exchange of the service flow can be, and therefore the lower the priority of the service flow. It should be understood that the priorities of different business traffic may be the same or different.

In an exemplary embodiment, based on at least one information of SLA and connection type, at least two service flows are determined. Priority methods include but are not limited to the following determination method A1 and determination method A2.

Determination method A1 is to obtain a reference correspondence relationship, and determine priorities of at least two business flows based on the reference correspondence relationship. The reference correspondence relationship is: a correspondence relationship between at least one information of SLA and connection type and the priority.

Among them, the reference correspondence can be generated randomly or configured based on experience. The method of obtaining the reference correspondence is not limited here, as long as it is ensured that the priority determined based on the reference correspondence is positively related to the service quality indicated by the SLA of the business traffic. It suffices to be negatively correlated with the scheduling convenience corresponding to the connection type.

For example, see FIG. 3 , which shows an exemplary reference correspondence relationship, which is a correspondence relationship between SLA, connection type, and priority. Among them, the connection type of live broadcast service traffic is long connection and the SLA is high. A high SLA indicates the highest service quality. From the reference relationship, it can be seen that the priority is level one, and level one is the highest priority. The connection type of dynamically accelerating business traffic is long connection and the SLA is high. From the reference relationship, we can know that the priority is level one. The connection type of the CDN on-demand service traffic is short connection and the SLA is medium. An SLA of medium indicates the next highest service quality. From this reference correspondence, it can be seen that the priority is Level 2, and Level 2 is the second highest priority. The connection type of CDN download service traffic is short connection and the SLA is low. A low SLA indicates the lowest service quality. From this reference correspondence, it can be seen that the priority is level three, and level three is the lowest priority. The connection type of VIS service traffic is short connection and the SLA is low. From the reference relationship, it can be seen that the priority is level three.

It should be understood that level one is the highest priority and level three is the lowest priority. This is only an example. In the embodiment of the present application, level three is the highest priority and level one is the lowest priority. Moreover, the priority levels including first level, second level, and third level are only examples, and the number of priorities and the way of expressing the priorities are not limited thereto. For example, the priority levels can also be made to include gold, silver, and bronze, and the priorities can also be made to include level A, level B, level C, level D, and so on.

Determination method A2: the scheduler determines the comprehensive scores of at least two service flows based on at least one information of SLA and connection type, and determines the priorities of at least two service flows based on the comprehensive scores of at least two service flows.

The determination method A2 may include the following steps 1 to 3.

Step 1: Determine the first score corresponding to the SLA and the second score corresponding to the connection type, then the comprehensive score of a business flow is determined based on at least one of the first score and the second score.

Since the embodiment of the present application needs to make the priority determined based on the comprehensive score positively correlated with the service quality indicated by the SLA of the business flow and negatively correlated with the scheduling convenience corresponding to the connection type, therefore, when determining the first score corresponding to the SLA, When determining the second score corresponding to the connection type, there are the following two implementation methods.

In one implementation, the higher the comprehensive score, the higher the priority determined based on the comprehensive score. Then, the higher the service quality indicated by the SLA, the larger the first score corresponding to the SLA, and the higher the scheduling convenience corresponding to the connection type, and the smaller the second score corresponding to the connection type. For example, if SLA is high, the first score is 80, if SLA is medium, the first score is 60, and if SLA is low, the first score is 40. For another example, if the connection type is a long connection, the second score is 80, and if the connection type is a short connection, the second score is 60.

In another implementation manner, the lower the comprehensive score, the higher the priority determined based on the comprehensive score. Then, the higher the service quality indicated by the SLA, the smaller the first score corresponding to the SLA, and the higher the scheduling convenience corresponding to the connection type, and the larger the second score corresponding to the connection type. This embodiment will not be described again.

Step 2: Perform a weighted calculation on at least one of the first score and the second score, and determine the obtained calculation result as a comprehensive score of the business flow.

Since the weighting calculation requires the use of weights, the weight of the first score and the weight of the second score need to be determined. The confirmation method is, for example, random generation or configuration based on experience, which is not limited here. The weight of the first score and the weight of the second score The weights can be the same or different. For example, for a business flow, if there is only the first score, the weight of the first score is 1, or if there is only the second score, the weight of the second score is 1, or, If there is both a first score and a second score, the sum of the weight of the first score and the weight of the second score is 1.

For example, the first score corresponding to the SLA of a business flow is 80, and the weight of the first score is 0.6, and the second score corresponding to the connection type of the business flow is 60, and the weight of the second score is 0.4, then The calculation result obtained by performing weighted calculation, that is, the comprehensive score of the business traffic, is equal to 72.

Step 3: Based on the value range of the comprehensive score of the business traffic, determine the priority corresponding to the value range as the priority of the business traffic.

The value range may correspond to the priority one-to-one, and the number of the value range and the priority may be at least one. For example, the higher the comprehensive score, the higher the priority determined based on the comprehensive score. The value range 1 is 61-70, corresponding to priority level three, and the value range 2 is 71-80, corresponding to priority level two. Level, the value range 3 is 81-90, and the corresponding priority is level one. Then, if the comprehensive score of a business flow is 72 calculated in step 2, then since 72 is in the value range 2, the priority corresponding to the value range 2, that is, Level 2, is determined as the priority of the business flow. class.

In addition, for at least two service flows with the same transmission direction on the first node, the bandwidth occupied by the at least two service flows on the first node is described.

The first node provides a certain total bandwidth for at least two service flows on the first node to occupy, and the value of the bandwidth occupied by each of the at least two service flows on the first node is a positive number. The unit of bandwidth is, for example, megabits per second (Mbps), which is not limited here.

In an exemplary embodiment, the bandwidth occupied by at least two service flows on the first node is obtained by at least one of detection and prediction. Since the scheduler obtains the bandwidth occupied by at least two business flows on the first node, the scheduler globally monitors at least two business flows on the first node to facilitate subsequent unified coordination of at least two business flows. Scheduling, the scheduler is also called the main co-scheduler. Therefore, it is possible to avoid conflicts in the scheduling process of different business flows, thereby avoiding a waste of bandwidth or a waste of bandwidth costs corresponding to the bandwidth.

For the detection method, after the scheduler detects the bandwidth occupied by at least two business flows on the first node in the first period, the scheduler can schedule the business flows based on the detected bandwidth. For the scheduling process, please refer to step 203 below, which will not be described again here. For example, the scheduler may receive bandwidth information periodically and proactively reported by the first node, or the scheduler may send an instruction to the first node and receive bandwidth information reported and fed back by the first node according to the instruction. The bandwidth information includes at least two The bandwidth occupied by business traffic on the first node. From this, the scheduler can implement bandwidth detection. For example, in the bandwidth information reported by the first node, the bandwidth occupied by a business flow on the first node includes at least one bandwidth corresponding to a time unit, and the time unit is greater than or equal to the time unit that can be identified by the first node and the scheduler. the smallest unit.

For the prediction method, after the scheduler detects the bandwidth occupied by at least two business flows on the first node in the first period, it predicts the bandwidth occupied by at least two business flows on the first node in the second period based on the detected bandwidth. , the second time period is later than the first time period. For example, the scheduler can obtain a predictive model that outputs predicted bandwidth based on the input bandwidth. The scheduler inputs the detected bandwidth into the prediction model and obtains the predicted bandwidth output by the prediction model. The predicted bandwidth can reflect the time when the business traffic peaks in the second period, the amount of data, and so on. The method of prediction is not limited here. After the predicted bandwidth is obtained, the service traffic can be scheduled based on the predicted bandwidth. For the scheduling process, please also refer to step 203 below, which will not be described again here.

For example, the scheduling of service traffic based on the predicted bandwidth may be based on the detected bandwidth and the predicted bandwidth. In this case, the amount of data used to schedule the bandwidth of the service flow is relatively large, which can make the business flow more efficient. flow The scheduling is more timely and accurate. Alternatively, the service traffic may be scheduled based on the predicted bandwidth, or the service traffic may be scheduled only based on the predicted bandwidth and not based on the detected bandwidth.

It should be understood that for different service flows, the method of obtaining the bandwidth occupied on the first node may be the same or different. For example, the bandwidth occupied by at least two business flows on the first node is obtained through detection, or both are obtained through prediction. For another example, the bandwidth occupied by a part of the at least two business flows on the first node is obtained through detection, and the bandwidth occupied by the other part of the business flow on the first node is obtained through prediction.

Step 202: The scheduler determines the target service flow among the at least two service flows based on the priorities of the at least two service flows.

Since the scheduler obtains the priorities of at least two service flows in step 201, it can determine at least one target service flow from the at least two service flows based on the priorities of the at least two service flows. In the subsequent process of traffic scheduling, the scheduler mainly schedules the target business traffic, which is also called the main greedy traffic. For example, after the scheduler determines the target service flow, it can send prompt information to the terminal. The prompt information is used to prompt that the target service flow will be the main greedy flow and guide users of the terminal to use the target service flow. The scheduler schedules the target business traffic, for example, to adjust the proportion of the target business traffic between the first node and the second node and the terminal, so that the bandwidth occupied by the target business traffic on the first node and the second node changes, as described below. Details will be described in step 203.

Exemplarily, the scheduler determines the target service flow among the at least two service flows based on the priorities of the at least two service flows, including: the scheduler determines the target service flow among the at least two service flows based on the priorities of the at least two service flows. There is at least one service flow with the lowest priority, and the scheduler determines the target service flow based on at least one service flow with the lowest priority.

Wherein, when there is no service flow with the same priority in at least two service flows, the number of the service flow with the lowest priority is one. Or, when there are service flows with the same priority in at least two service flows, the number of the service flows with the lowest priority may be one or multiple. For example, as shown in Figure 3, the lowest priority is level three, so the number of business flows with the lowest priority is two, namely CDN download business traffic and VIS business traffic.

In some implementations, in response to the fact that there is only one service flow with the lowest priority, this one service flow with the lowest priority is the target service flow. That is, the scheduler determines the target service flow based on at least one service flow with the lowest priority, including: the scheduler uses the service flow with the lowest priority as the target service flow.

Since the priority of a service flow is positively related to the service quality indicated by the SLA of the service flow and negatively related to the scheduling convenience corresponding to the connection type of the service flow, the embodiment of the present application targets the service flow with the lowest priority. Business traffic, therefore the target business traffic is a kind of business traffic with lower service quality and higher scheduling convenience indicated by SLA. The scheduler mainly schedules this kind of target service traffic. Even if the actual service quality of the target service flow is reduced due to scheduling, the service quality indicated by the SLA of the target service flow is itself low, so that the actual service quality can still be reduced. Greater than or equal to the service quality indicated by the SLA, thereby ensuring that the actual service quality of the target service flow meets the service quality indicated by the SLA. Moreover, the scheduling of such target business traffic is also more flexible and convenient.

In other embodiments, in response to the presence of at least two lowest priority service flows, the scheduler determines the target service flow based on at least one lowest priority service flow, including: the scheduler assigns the at least two lowest priority service flows to The business traffic whose data volume is greater than the data volume threshold is determined as the target business traffic.

In this implementation, in addition to ensuring that the actual service quality of the target service flow meets the service quality indicated by the SLA and making the scheduling of the target service flow more flexible and convenient, it can also ensure that the target service flow has a relatively large amount of data. Large business traffic, or target business traffic is business traffic that accounts for a relatively large proportion of traffic. due to target business The amount of traffic data is large, so the scheduler can schedule the target business traffic to a greater extent, such as making a large change in the bandwidth occupied by the target business traffic on the first node and the second node, thereby providing sufficient Scheduling space. For example, the business traffic with a data volume greater than the data volume threshold may be the business traffic with the largest data volume.

Of course, in addition to determining the target service flow based on at least one service flow with the lowest priority according to the above description, a service flow with a non-lowest priority and a non-highest priority can also be determined as the target service flow. This method also has a high probability of making the actual service quality of the target service flow meet the service quality indicated by the SLA, and the scheduling of this kind of target service flow is also more flexible and convenient. For example, there are 10 types of business traffic on the first node, with priorities ranging from level one to level ten, with level one being the highest priority and level ten being the lowest priority. In addition to determining the service traffic with priority level ten as the target service flow according to the above description, the service traffic with priority level seven, eight, or nine can also be determined as the target service flow.

The above describes the method of determining target business traffic. In this embodiment of the present application, at least one target service flow may be determined for each transmission direction. For example, among each service flow with the transmission direction being downlink, determine a target service flow with the transmission direction being downlink, and among each service flow with the transmission direction being uplink, determine a target service flow with the transmission direction being uplink.

Step 203: Based on the bandwidth occupied by at least two service flows on the first node, the scheduler adjusts the proportion of target service flows between the first node and the second node and the terminal.

Since in step 201 the scheduler obtains the bandwidth occupied by at least two service flows on the first node, and in step 202 the scheduler determines the target service flow, the scheduler can adjust the first node and the second node based on the obtained bandwidth. The proportion of target business traffic between nodes and terminals. The second node can be any edge node in the CDN. For example, the second node may be an edge node bound to the first node in the CDN, and the binding relationship between the first node and the second node may be configured according to actual requirements, which is not limited here. By adjusting the ratio, the bandwidth occupied by the target business traffic on the first node can be changed, that is, the actual bandwidth occupied by the target business traffic on the first node is limited. In addition, the bandwidth occupied by the target service traffic on the second node also changes.

Among them, if the ratio decreases, it means that the target business traffic of the first node interacting with the terminal decreases, and the bandwidth occupied by the target business traffic on the first node decreases, while the target business traffic of the second node interacting with the terminal increases, and the target business traffic The increase in the bandwidth occupied on the second node is equivalent to scheduling the target service traffic from the first node to the second node. For example, before the ratio is reduced, the bandwidth occupied by the target service traffic on the second node may be zero, that is, before the ratio is reduced, the second node may not exchange service traffic with the terminal. Of course, before reducing the ratio, the bandwidth occupied by the target service traffic on the second node may not be zero, which is not limited here.

For example, before adjusting the ratio, 20 terminals located in location A and 30 terminals located in location B all interact with the first node for target business traffic, while 50 terminals located at location C interact with the second node for target business traffic. , then before adjusting the ratio, the ratio of the target service traffic of the first node and the second node to the terminal is 50:50. After reducing the ratio, the 20 terminals at location A switch to interacting with the second node for target business traffic. After reducing the ratio, the ratio of the first node and the second node interacting with the terminal for target business traffic is 30: 70.

Or, if the ratio increases, it means that the target business traffic of the first node interacting with the terminal increases, and the bandwidth occupied by the target business traffic on the first node increases, while the target business traffic of the second node interacting with the terminal is less, and the target business The bandwidth occupied by the traffic on the second node is reduced, which is equivalent to scheduling part of the target service traffic from the second node to the first node. This situation will not be repeated again.

In an exemplary embodiment, the scheduler adjusts the first node based on the bandwidth occupied by at least two service flows on the first node. The proportion of the target service traffic of the node and the second node interacting with the terminal includes: the scheduler determines the sum of the bandwidths occupied by at least two business flows on the first node; the scheduler determines the proportion adjustment method based on the sum of bandwidths; scheduling The controller adjusts the proportion of the target service traffic of the first node and the second node interacting with the terminal according to the proportional adjustment method.

After the scheduler obtains the bandwidth occupied by at least two business flows on the first node, the scheduler sums the obtained bandwidths to obtain the sum of the bandwidths. Then, the proportion adjustment method can be determined based on the sum of the bandwidths, and the proportion adjustment method can be used to indicate a reduction in the proportion, or can also be used to indicate an increase in the proportion. Then, according to the proportional adjustment method, the proportion of the first node and the second node interacting with the terminal can be adjusted.

For example, the scheduler determines the proportion adjustment method based on the sum of bandwidths, including but not limited to the following determination methods B1 to B3.

Determining method B1, in response to the sum of bandwidths meeting the first condition, the scheduler determines the proportion adjustment method to reduce the proportion of the target service traffic of the first node and the second node interacting with the terminal, so that the bandwidth occupied by the target service traffic on the first node Reduce the bandwidth scheduling amount, and increase the bandwidth scheduling amount for the bandwidth occupied by the target service traffic on the second node.

Among them, the first condition is used to indicate that the first node is in the peak period of business traffic and has insufficient bandwidth. Therefore, the proportion adjustment method is used to indicate reducing the proportion so that the bandwidth occupied by the target business traffic on the first node is reduced to alleviate the problem of the first node. bandwidth pressure.

In some embodiments, in response to the sum of bandwidths being greater than a first bandwidth threshold, it is determined that the sum of bandwidths satisfies the first condition. The first bandwidth threshold can be set based on experience or actual needs, which is not limited here. For example, for a business flow, determine the reference bandwidth required when the actual service quality of the business flow meets the service quality indicated by the SLA of the business flow, that is, the reference bandwidth required to meet user requirements and ensure user experience. For the reference bandwidth, the sum of the first reference bandwidths of at least two service flows is used as the first bandwidth threshold. Then, when the sum of bandwidths is greater than the first bandwidth threshold, it means that there may be service traffic whose actual service quality does not meet the service quality indicated by the SLA, and the user experience is affected, which further means that the first node is in the peak period of service traffic and has insufficient bandwidth.

In other embodiments, in response to the bandwidth cost corresponding to the sum of bandwidths being greater than a first cost threshold, it is determined that the sum of bandwidths satisfies the first condition. The first cost threshold can be set based on experience or actual needs, which is not limited here. For example, the bandwidth cost corresponding to the sum of the first reference bandwidths of the at least two service flows may be used as the first cost threshold.

Whether it is the bandwidth cost corresponding to the sum of bandwidths or the bandwidth cost corresponding to the sum of the first reference bandwidths, they need to be determined based on a certain billing model. The billing model includes but is not limited to 95 percent (percent 95, P95) Billing. Taking the bandwidth cost corresponding to the sum of bandwidths as an example to illustrate the P95 billing process. Among them, according to a certain collection frequency, such as once every 5 minutes, collect the sum of multiple bandwidths of the first node within a period of time, and arrange the sum of multiple bandwidths in descending order to obtain a sequence. In this sequence, the first 5% of the total bandwidth is not billed. Among the remaining 95% of the total bandwidth, the fee corresponding to the largest sum of bandwidth is determined as the first node's cost during that period. The bandwidth cost corresponding to the sum of bandwidths.

Referring to Figure 4, the at least two service flows include the target service flow and the first service flow. The first service flow is the service flow other than the target service flow in the at least two service flows. When the sum of the bandwidths occupied by at least two business flows on the first node satisfies the first condition, the scheduler needs to reduce the proportion of the first node and the second node's interaction with the terminal in a proportional adjustment manner. , the target business flow is scheduled from the first node to the second node, so that the bandwidth occupied by the target business flow on the first node is reduced by the bandwidth scheduling amount, and the bandwidth occupied by the target business flow on the second node is increased by the bandwidth scheduling amount. In Figure 4, the solid line represents the target service flow before scheduling, the dotted line represents the target service flow after scheduling, and the arrow represents the scheduling direction.

Exemplarily, the situation shown in Figure 4 is the situation where the target service traffic is scheduled from the first node to a second node, except that In addition to the situation shown in Figure 4, the target service traffic can also be scheduled from the first node to two or more second nodes. Then the bandwidth occupied by the target service flow on the second node is increased by the bandwidth scheduling amount, which means that the sum of the increases in the bandwidth occupied by the target service flow on more than two second nodes is the bandwidth scheduling amount.

It should be understood that the bandwidth scheduling amount needs to be determined before the scheduler reduces the ratio. In some implementations, the value of the bandwidth scheduling amount is a default value, which can be obtained through configuration and is not limited here. In other implementations, the scheduler obtains the bandwidth scheduling amount through calculation, see the following description.

For example, if the sum of bandwidths is greater than the first bandwidth threshold, it is determined that the sum of bandwidths satisfies the first condition, and the bandwidth scheduling amount is determined based on the sum of bandwidths, the first bandwidth threshold and the bandwidth usage of the second node. Optionally, the bandwidth usage of a second node is used to indicate whether there is available bandwidth on the second node. If the bandwidth usage of a second node indicates that there is bandwidth that can be occupied on the second node, the scheduler uses the second node as an alternative second node.

Next, the scheduler determines the first difference between the sum of bandwidths and the first bandwidth threshold, and uses the first difference as the minimum value of the bandwidth scheduling amount. The minimum value of the bandwidth scheduling amount reflects the need to transfer resources from the first node. How much target business traffic. The scheduler also determines the sum of the bandwidths that can be occupied by each candidate second node, and uses the sum of the bandwidths that can be occupied as the maximum value of the bandwidth scheduling amount. The maximum value of the bandwidth scheduling amount reflects how many target services the second node can accommodate. flow. Then, the scheduler selects a value between the minimum value of the bandwidth scheduling amount and the maximum value of the bandwidth scheduling amount as the bandwidth scheduling amount according to the actual situation. According to this bandwidth scheduling amount reduction ratio, the scheduler can make the sum of the bandwidths occupied by the at least two business flows on the first node less than the first bandwidth threshold after the scheduling is completed.

Optionally, the scheduler calculates the sum of the second reference bandwidth of the bandwidth occupied by each service traffic on the second node, and determines a bandwidth value for charging on the second node according to the charging mode of P95 charging. , this bandwidth value is the dividing point between 5% that does not require charging and 95% that requires charging. When the sum of the second reference bandwidths is less than this bandwidth value, charging is required. When the sum of the second reference bandwidths No charging is required if the bandwidth is greater than this value.

If the sum of the second reference bandwidths is less than the bandwidth value, and the difference between the bandwidth value and the sum of the second reference bandwidths is greater than the first reference threshold, it is considered that there is bandwidth that can be occupied on the second node, and the second node is a candidate The second node, and the bandwidth that can be occupied is the difference between the bandwidth value and the sum of the second reference bandwidth.

Or, if the sum of the second reference bandwidths is greater than the bandwidth value, no matter how much bandwidth is added to the sum of the second reference bandwidths, it does not require charging, so it is considered that there is available bandwidth on the second node. Bandwidth, the second node is an alternative second node, and the bandwidth it can occupy is infinite.

For example, if the bandwidth cost corresponding to the sum of bandwidths is greater than the first cost threshold, it is determined that the sum of bandwidths meets the first condition, and the bandwidth scheduling amount is based on the bandwidth cost corresponding to the sum of bandwidths, the first cost threshold and the second cost threshold. The node's bandwidth cost usage is determined. Optionally, the cost usage of a second node is used to indicate whether there is usable bandwidth cost on the second node. If the cost usage of a second node indicates that there is usable bandwidth cost on the second node, the scheduler will use the second node as an alternative second node, and the bandwidth that can be occupied on an alternative second node is the usable bandwidth cost. The bandwidth cost corresponds to the bandwidth.

The scheduler can determine the second difference between the bandwidth cost corresponding to the sum of bandwidths and the first cost threshold, use the bandwidth corresponding to the second difference as the minimum value of the bandwidth scheduling amount, and calculate the bandwidth that can be occupied on each candidate second node. The sum of the bandwidths is taken as the maximum value of the bandwidth scheduling amount, so that the bandwidth scheduling amount is obtained by selecting between the minimum value of the bandwidth scheduling amount and the maximum value of the bandwidth scheduling amount. According to this bandwidth scheduling reduction ratio, the scheduler can make the bandwidth cost corresponding to the sum of the bandwidths occupied by the at least two business flows on the first node less than the first cost threshold after the scheduling is completed.

In an exemplary embodiment, for the determination method B1, the method may further include: increasing the first service flow at the first node The bandwidth occupied is less than or equal to the bandwidth scheduling amount. According to the description corresponding to Figure 4 above, it can be known that the first service flow is the service flow other than the target service flow among the at least two service flows.

Among them, since the scheduler reduces the bandwidth occupied by the target service flow on the first node, it then increases the bandwidth occupied by the first service flow on the first node, which is equivalent to the target service flow giving up the bandwidth to the first service flow. . Thus, the delay between the first node and the terminal in interacting with the first service flow can be reduced, and the actual service quality of the first service flow can be improved, which is conducive to ensuring that the actual service quality of the first service flow meets the SLA indication of the first service flow. Service quality can improve user experience. In addition, since the increased bandwidth is less than or equal to the bandwidth scheduling amount, after increasing the bandwidth occupied by the first service flow on the first node, the sum of the bandwidth occupied by the target service flow and the bandwidth occupied by the first service flow can be does not exceed the first bandwidth threshold, or the cost corresponding to the sum of bandwidths does not exceed the first cost threshold.

Determining method B2, in response to the sum of bandwidths meeting the second condition, the scheduler determines the proportion adjustment method to increase the proportion of the target service traffic of the first node and the second node interacting with the terminal, so that the bandwidth occupied by the target service traffic on the first node The bandwidth scheduling amount is increased, and the bandwidth occupied by the target service traffic on the second node is reduced by the bandwidth scheduling amount.

Among them, the second condition is used to indicate that the first node is in a low period of business traffic and has excess bandwidth, so the proportion adjustment method is used to indicate increasing the proportion so that the bandwidth occupied by the target business traffic on the first node increases, so that the first node The excess bandwidth can be fully reused. As a result, peak shaving and valley filling are realized on the first node, and the valley represents the excess bandwidth on the first node.

In some embodiments, in response to the sum of bandwidths being less than or equal to a second bandwidth threshold, it is determined that the sum of bandwidths satisfies the second condition, and the second bandwidth threshold is less than or equal to the first bandwidth threshold explained in the above determination manner B1, the The second bandwidth threshold can be set based on experience or actual needs, and is not limited here.

In other embodiments, in response to the bandwidth cost corresponding to the sum of bandwidths being less than or equal to the second cost threshold, it is determined that the sum of bandwidths satisfies the second condition. The method of calculating the bandwidth cost corresponding to the sum of bandwidths can refer to the determination method B1 above. The description in will not be repeated here. The second cost threshold is less than or equal to the first cost threshold described in the above determination method B1. The second cost threshold can be set based on experience or actual needs, and is not limited here.

Referring to Figure 5, at least two service flows include target service flow and first service flow. The first service flow is the service flow other than the target service flow in the at least two service flows. When the sum of the bandwidths occupied by at least two business flows on the first node satisfies the second condition, the scheduler needs to adjust proportionally to increase the proportion of target business flows between the first node and the second node interacting with the terminal. , the target business flow is scheduled from the second node to the first node, so that the bandwidth occupied by the target business flow on the first node is increased by the bandwidth scheduling amount, and the bandwidth occupied by the target business flow on the second node is reduced by the bandwidth scheduling amount. In Figure 5, the solid line represents the target service flow before scheduling, the dotted line represents the target service flow after scheduling, and the arrow represents the scheduling direction.

For example, the situation shown in Figure 5 is a situation where the target service traffic is scheduled from a second node to the first node. In addition to the situation shown in Figure 5, the target service traffic can also be dispatched from two or more third nodes. The two nodes are respectively scheduled to the first node. The bandwidth reduction amount of the bandwidth occupied by the target service flow on the second node means that the sum of the reduction amounts of the bandwidth occupied by the target service flow on more than two second nodes is the bandwidth scheduling amount.

It should be understood that the bandwidth scheduling amount needs to be determined before the scheduler increases the ratio. In some implementations, the value of the bandwidth scheduling amount is a default value, which can be obtained through configuration and is not limited here. In other implementations, the scheduler obtains the bandwidth scheduling amount through calculation, see the following description.

For example, if the sum of bandwidths is less than or equal to the second bandwidth threshold, it is determined that the sum of bandwidths meets the second condition, and the bandwidth scheduling amount is determined based on the sum of bandwidths, the first bandwidth threshold and the bandwidth usage of the second node. . optionally, The bandwidth usage of a second node is used to indicate whether there is target business traffic occupying a certain bandwidth on the second node. If the bandwidth usage of a second node indicates that there is target business traffic occupying a certain bandwidth on the second node, The scheduler uses the second node as an alternative second node.

Next, the scheduler determines the third difference between the first bandwidth threshold and the sum of bandwidths, and uses the third difference as a value index for the bandwidth scheduling amount. This value index reflects how much the first node can accommodate. Target business traffic. The scheduler also determines the sum of the bandwidth occupied by the target service traffic on each candidate second node, and uses the sum of the occupied bandwidth as another value indicator of the bandwidth scheduling amount. This other value indicator reflects the second node's bandwidth. How much target business traffic can be transferred away. Then, the scheduler takes the smaller of the two value indicators as the target value indicator, and selects a value less than or equal to the target value indicator as the bandwidth scheduling amount according to the actual situation. According to this bandwidth scheduling amount increase ratio, the scheduler can make the sum of the bandwidths occupied by the at least two business flows on the first node greater than the second bandwidth threshold and less than the first bandwidth threshold after the scheduling is completed.

For example, if the bandwidth cost corresponding to the sum of bandwidths is less than or equal to the second cost threshold, it is determined that the sum of bandwidths meets the second condition, and the bandwidth scheduling amount is based on the bandwidth cost corresponding to the sum of bandwidths, the first cost threshold and The bandwidth cost usage of the second node is determined. Optionally, the cost usage of a second node is used to indicate whether there is target service traffic using a certain bandwidth cost on the second node. If the cost usage of a second node indicates that there is target service traffic using a certain bandwidth cost on the second node, the scheduler uses the second node as an alternative second node.

The scheduler may determine the fourth difference between the first cost threshold and the bandwidth cost corresponding to the sum of bandwidths, use the bandwidth corresponding to the fourth difference as a value indicator of the bandwidth scheduling amount, and assign the bandwidth cost to each candidate second node The bandwidth corresponding to the sum of the bandwidth corresponding to the bandwidth cost used by the target business traffic is used as another value indicator of the bandwidth scheduling amount. Then, the scheduler uses the smaller of the two value indicators as the target value indicator. , according to the actual situation, select a value less than or equal to the target value index as the bandwidth scheduling amount. According to this bandwidth scheduling increase ratio, the scheduler can make the bandwidth cost corresponding to the sum of the bandwidths occupied by the at least two business flows on the first node greater than the second cost threshold and less than the first cost threshold after the scheduling is completed. Cost threshold.

The above determination method B1 is a way to reduce the ratio, and the above determination method B2 is a way to increase the ratio. It should be understood that the ratio of the first node and the second node to the terminal interactive target service traffic can also remain unchanged. . For example, determine the sum of bandwidths occupied by at least two business flows on the first node. If the sum of bandwidths is between the first bandwidth threshold and the second bandwidth threshold, it means that although there is excess bandwidth on the first node , but there is less spare bandwidth, so the target service traffic can not be scheduled, thus keeping the ratio unchanged. For another example, if the bandwidth cost corresponding to the sum of bandwidths is between the first cost threshold and the second cost threshold, it means that there is excess bandwidth cost on the first node, but the excess bandwidth cost is less, which means that the first node There is also less spare bandwidth, so the ratio can be kept unchanged. Based on this, please refer to Figures 6 and 7, which are schematic flow charts of a traffic scheduling method respectively.

In Figure 6, the first bandwidth threshold and the second bandwidth threshold are planned, and business traffic monitoring is performed on the first node. If the sum of the bandwidths occupied by at least two service flows on the first node is greater than the first bandwidth threshold, the proportion is reduced, that is, the bandwidth occupied by the target service flow on the first node is reduced. If the sum of the bandwidths is not greater than the first bandwidth threshold and is less than or equal to the second bandwidth threshold, the proportion is increased, that is, the bandwidth occupied by the target service traffic on the first node is increased. If the sum of the bandwidths is not greater than the first bandwidth threshold but greater than the second bandwidth threshold, the ratio remains unchanged, and the bandwidth occupied by the target service traffic on the first node does not change. Thus, the scheduling of business traffic can be completed once.

In Figure 7, the first cost threshold and the second cost threshold are planned, and business traffic monitoring is performed on the first node. If the bandwidth cost corresponding to the sum of bandwidths is greater than the first cost threshold, the proportion is reduced, that is, the target business traffic is reduced at the first node. bandwidth occupied. If the bandwidth cost corresponding to the sum of the bandwidths is not greater than the first cost threshold and less than or equal to the second cost threshold, the proportion is increased, that is, the bandwidth occupied by the target business traffic on the first node is increased. If the bandwidth cost corresponding to the sum of bandwidths is not greater than the first cost threshold but greater than the second cost threshold, the ratio remains unchanged, and the bandwidth occupied by the target business traffic on the first node does not change. Thus, the scheduling of business traffic can be completed once.

Determining the method B3, in response to the sum of the bandwidths being different from the bandwidth occupied by the second service traffic on the first node, the scheduler determines the proportion adjustment method.

Wherein, the transmission directions of at least two service flows are different from the transmission direction of the second service flow. For example, the transmission direction of at least two service flows is uplink, and the transmission direction of the second service flow is downlink. For another example, the transmission direction of at least two service flows is downlink, and the transmission direction of the second service flow is uplink. Therefore, when the sum of bandwidths is different from the bandwidth occupied by the second service traffic on the first node, it means that the bandwidth occupied by the service traffic in different transmission directions on the first node is different. For example, see FIG. 8 , which shows a schematic diagram showing that business traffic in different transmission directions occupies different bandwidths.

Therefore, the scheduler can change the bandwidth occupied by the target business traffic on the first node by adjusting the ratio, so that the bandwidth occupied by the business traffic in different transmission directions is the same, or the bandwidth occupied by the business traffic in different transmission directions is narrowed. difference. For example, the proportion adjustment method is determined based on the sum of bandwidth, the bandwidth occupied by the second service traffic on the first node, and the bandwidth usage of the second node.

If the sum of bandwidths is greater than the bandwidth occupied by the second service traffic on the first node, the ratio can be reduced to reduce the bandwidth occupied by the target service traffic on the first node, and the sum of bandwidths will also be reduced. In addition, the scheduler may determine a fifth difference between the sum of bandwidths and the bandwidth occupied by the second service flow on the first node, and use the fifth difference as the bandwidth scheduling amount, thereby reducing the proportion according to the bandwidth scheduling amount.

Alternatively, if the sum of bandwidths is less than the bandwidth occupied by the second service traffic on the first node, the ratio can be increased to increase the bandwidth occupied by the target service traffic on the first node, and the sum of bandwidths will also increase. In addition, the scheduler may determine a sixth difference between the bandwidth occupied by the second service flow on the first node and the sum of the bandwidths, use the sixth difference as the bandwidth scheduling amount, and increase the ratio according to the bandwidth scheduling amount.

Above, the process of the scheduler determining the proportion adjustment method based on the sum of bandwidths has been described through the determination methods B1 to B3. Next, the process of the scheduler adjusting the ratio of the first node and the second node to the terminal to interact with the target service traffic in a proportional adjustment manner will be continued to be described. Exemplarily, the scheduler adjusts the proportion of the target service traffic of the first node and the second node interacting with the terminal in a proportional adjustment manner, including: the scheduler determines the scheduling information corresponding to the target node in a proportional adjustment manner, and sends the scheduling information to the target node. , the scheduling information and the target node are used to adjust the ratio of the first node and the second node to interact with the terminal and target service traffic.

Among them, referring to Figure 9, the scheduler may include a scheduling perceptron, a scheduling decision maker and a scheduling executor. The scheduling sensor is used to sense the bandwidth occupied by at least two service flows on the first node, and send the perceived bandwidth to the scheduling decision maker. The scheduling sensor can also be used to sense the bandwidth occupied by each service flow on the second node and send it to the scheduling decision maker. For the sensing process, please refer to the description in step 201. The scheduling decision maker is used to determine the priority of each service flow based on at least one information of SLA and connection type. For the determination process, please refer to the description in step 201. The scheduling decision maker is also used to determine the target service flow based on the priority of each service flow. For the determination process, please refer to the description in step 202. The scheduling decision maker is also used to determine the proportion adjustment method according to the bandwidth sent by the scheduling sensor. For the determination process, please refer to the above description in step 203. Therefore, the scheduling decision maker can determine the scheduling strategy corresponding to the target node in a proportional adjustment manner, and then send the scheduling strategy to the scheduling executor. For example, the scheduling policy includes but is not limited to: what percentage of the target service traffic is scheduled from one node to another node, or how much bandwidth is scheduled for the target service traffic to be scheduled from one node to another node. After that, schedule execution The scheduler is used to generate scheduling information based on the scheduling policy, and adjust the ratio of the first node and the second node to the terminal to interact with the target service flow based on the scheduling information, thereby achieving scheduling of the target service flow.

In some embodiments, if the target service flow has an autonomous scheduling function, for example, the target service flow corresponds to an autonomous scheduler, the scheduling executor can modify the outer parameters of the autonomous scheduler based on the scheduling information to achieve the target service flow. Scheduling. Or, referring to Figure 9, the scheduling executor can deliver the scheduling information to the target node to achieve scheduling of the target service traffic. For example, for different target nodes, the scheduling information is also different, so the manner of scaling adjustment based on the scheduling information and the target node is also different. In an exemplary embodiment, the adjustment methods include, but are not limited to, the following adjustment methods C1 to C3.

Adjustment method C1 is suitable for DNS scheduling scenarios. In this adjustment method C1, the target node is a CDN DNS server, the scheduling information is the first correspondence between the reference domain name and the IP address, and the IP address indicates the first node or the second node. Among them, the first correspondence relationship is used for the CDN DNS server to return the IP address corresponding to the reference domain name to the terminal based on the reference domain name sent by the terminal. The IP address is used for the terminal to send an interaction request to the node indicated by the IP address to interact with the target business traffic.

The DNS scheduling scenario can be seen in Figure 10. The DNS scheduling scenario includes but is not limited to the terminal, the authoritative DNS server with authoritative DNS installed, the CDN DNS server corresponding to the CDN, the edge node in the CDN, the source server and the scheduler. The terminal includes applications and local domain name system (local DNS, LDNS). The edge node includes but is not limited to a first node and a second node. The IP address of the first node is IP A and the IP address of the second node is IP B. Among them, the terminal is connected to the authoritative DNS server, CDN DNS server and edge node respectively. The edge node is also connected to the source server and scheduler, and the scheduler is also connected to the CDN DNS server.

When a user of the terminal enters a domain name for the application, the application receives the domain name entered by the user of the terminal. The application sends the domain name to LDNS, LDNS determines the IP address corresponding to the domain name, and returns the IP address corresponding to the domain name to the application. The process by which LDNS determines the IP address corresponding to the domain name is also called the process of resolving the domain name. This allows the terminal to send an interaction request to the edge node indicated by the IP address corresponding to the domain name through the application program. The interaction request carries the domain name, thereby allowing the terminal to interact with the edge node indicated by the IP address with target business traffic corresponding to the domain name. For example, the application is a browser, the domain name is a web page, such as xxx.com, and the target business traffic corresponding to the domain name is the content that needs to be displayed on the web page.

In the process of LDNS determining the IP address corresponding to the domain name, LDNS first sends the domain name to the authoritative DNS server. The authoritative DNS server stores a canonical name (CNAME), which is the correspondence between the domain name and the reference domain name. For example, when the domain name is xxx.com, the reference domain name can be xxx.com.dns.cn. After receiving the domain name, the authoritative DNS server queries the CNAME based on the domain name, obtains the reference domain name corresponding to the domain name, and returns the reference domain name to LDNS.

Next, LDNS receives the reference domain name, queries the domain name server (NS) record based on the reference domain name, and obtains the CDN DNS server corresponding to the reference domain name. LDNS sends the reference domain name and the local IP address of the terminal to the CDN DNS server. The local IP address of the terminal indicates the location of the terminal. The CDN DNS server stores the correspondence between the reference domain name and the IP address. After the CDN DNS server receives the reference domain name and the local IP address of the terminal, it queries the correspondence between the reference domain name and the IP address based on the reference domain name, and obtains all IP addresses corresponding to the reference domain name. Each IP address indicates an edge node. Then, the CDN DNS server selects an IP address based on the terminal's local IP address among all IP addresses and returns the selected IP address to LDNS. The edge node indicated by the selected IP address is the edge node closest to the location of the terminal among the edge nodes included in the CDN corresponding to the CDN DNS server. After LDNS receives the IP address sent by the CDN DNS server, it uses the received IP address as The IP address corresponding to the domain name.

After LDNS determines the IP address corresponding to the domain name, in addition to returning the IP address corresponding to the domain name to the application, the LDNS will also cache the correspondence between the domain name and the IP address. If the LDNS sends the reference domain name and the local IP address of the terminal to the CDN DNS server at time M, and caches the correspondence between the domain name and the IP address, then the cache start time of the correspondence is time M, and the cache end time is N time, N time is later than M time. The cache duration of this correspondence is △T1, and △T1 is the difference between time N and time M.

Between time M and time N, if LDNS receives the same domain name sent by the application again, LDNS no longer needs to send the domain name to the authoritative DNS server. Instead, it can directly query the cached correspondence and include the hit correspondence. The IP address is returned to the application as the IP address corresponding to the domain name. After N time, because the corresponding relationship is deleted, even if LDNS receives the same domain name sent by the application again, LDNS needs to send the domain name to the authoritative DNS server again to obtain the IP address corresponding to the domain name and re-cache a corresponding relation.

Before scheduling the target business traffic, the correspondence between the reference domain name and the IP address stored by the CDN DNS server as the target node is the old correspondence. When scheduling target business traffic, since the scheduler sends scheduling information to the CDN DNS server, that is, the first correspondence between the reference domain name and the IP address, the correspondence between the reference domain name and the IP address stored by the CDN DNS server, The old correspondence will be transformed into a new correspondence.

For example, if the CDN DNS server receives the first correspondence sent by the scheduler at time X, then the CDN DNS server will uniformly transform the old correspondence into a new correspondence at time Y to achieve the goal starting from time Y. Scheduling of business traffic. The Y time is later than the X time, and the difference between the Y time and the X time is △T2, and the △T2 is greater than or equal to △T1. There are the following three situations regarding the relationship between the M time when LDNS sends the reference domain name and the terminal's local IP address to the CDN DNS server, and the X time and Y time.

Case 1: Time M is earlier than time X. Since at time M the CDN DNS server has not received the first correspondence sent by the scheduler, or the CDN DNS server does not know that the target business traffic needs to be scheduled, the CDN DNS server determines and returns the used Indicates the IP address of an edge node. This IP address is called the old IP address. Then, LDNS caches the correspondence between the domain name and the old IP address according to the cache duration ΔT1. The cache start time of this correspondence is M time, and the cache end time is N time, that is, this correspondence is It is cleared at time N. Since time M is earlier than time X, and △T2 is greater than or equal to △T1, time N will not be later than time Y.

Case 2: Time M is between time X and time Y. Although the CDN DNS server has received the first correspondence sent by the scheduler at time M, since time Y has not yet been reached, the CDN DNS server still determines and returns the old IP address according to the old correspondence. In addition, the CDN DNS server also returns a special cache duration ΔT3, which is smaller than the cache duration ΔT1. Then, LDNS caches the correspondence between the domain name and the old IP address according to the special cache duration ΔT3. The cache start time of this correspondence is M time, and the cache end time is N' time, that is, this This correspondence is cleared at N' time, and the difference between N' time and M time is △T3. Among them, the CDN DNS server will limit the value of △T3 to ensure that the N’ time is earlier than or equal to the Y time.

Case 3: Time M is later than time Y. Since at time M the CDN DNS server has received the first correspondence sent by the scheduler and has reached time Y, the CDN DNS server determines and returns the IP address used to indicate an edge node according to the new correspondence, and This IP address is called the new IP address. Then, LDNS caches the correspondence between the domain name and the new IP address according to the cache duration ΔT1.

According to these three situations, it can be seen that LDNS caches the correspondence between the domain name and the old IP address before time Y, so After receiving the domain name sent by the application, LDNS returns the old IP address to the application. Then the terminal where the application is located will exchange the target business traffic corresponding to the domain name with the old edge node indicated by the old IP address. LDNS caches the correspondence between the domain name and the new IP address after time Y. Therefore, after receiving the domain name sent by the application, LDNS returns the new IP address to the application. Then the terminal where the application is located is the same as The new IP address indicates the target business traffic corresponding to the new edge node interaction domain name. As a result, the target service flow is scheduled from the old edge node to the new edge node starting from time Y, that is, the target service flow is scheduled starting from time Y.

Referring to Figure 10, the traffic scheduling process in the DNS scheduling scenario is explained by taking reducing the proportion of the target service traffic of the first node and the second node interacting with the terminal, that is, scheduling the target service traffic from the first node to the second node, as an example.

Before scheduling the target business traffic, the old correspondence stored by the CDN DNS server is: the reference domain name corresponding to the domain name corresponding to the target business traffic corresponds to IP A. Then the old IP address obtained by LDNS from the CDN DNS server is IP A, and the terminal sends an interaction request carrying the domain name to the first node indicated by IP A, and the terminal exchanges the target business traffic corresponding to the domain name with the first node. When scheduling the target business traffic, the scheduler sends the first correspondence to the CDN DNS server. The first correspondence is: the reference domain name corresponding to the domain name corresponding to the target business traffic, corresponding to IP B, CDN DNS server This first correspondence is regarded as a new correspondence. After scheduling the target service traffic, the new IP address obtained by LDNS from the CDN DNS server is IP B, so the terminal sends an interaction request carrying the domain name to the second node indicated by IP B, and the terminal interacts with the second node for the target service. traffic, and no longer interacts with the target service traffic with the first node, thereby achieving scheduling of the target business traffic from the first node to the second node.

Adjustment method C2 is suitable for 302 scheduling scenarios. In this adjustment method C2, the target nodes are the first node and the second node, the scheduling information is the second correspondence between the interaction request and the IP address, and the IP address indicates the first node or the second node. Wherein, the second correspondence relationship is used for the target node to interact with the terminal on the target service traffic after receiving the interaction request sent by the terminal and in response to determining that the node indicated by the IP address corresponding to the interaction request is the current node based on the second correspondence relationship, or , in response to determining that the node indicated by the IP address corresponding to the interaction request is not the current node based on the second correspondence relationship, the IP address corresponding to the interaction request is returned to the terminal. The IP address is used by the terminal to send an interaction request to the node indicated by the IP address to interact. Target business traffic.

The 302 scheduling scenario can be seen in Figure 11. The 302 scheduling scenario includes but is not limited to: terminal, authoritative DNS server with authoritative DNS installed, CDN DNS server corresponding to CDN, edge node in CDN, source server and scheduler. The terminal includes applications and local domain name system (local DNS, LDNS). The edge node includes a first node and a second node. The IP address of the first node is IP A and the IP address of the second node is IP B. The total number of nodes included in the CDN where the edge node is located is not limited here. Among them, the terminal is connected to the authoritative DNS server and CDN DNS server respectively. The terminal is also connected to the edge node, and the edge node is also connected to the source server and scheduler.

Before scheduling the target business traffic, the application sends a domain name to LDNS. Based on the domain name, LDNS obtains an old IP address from the CDN DNS server. The old IP address is used to indicate the old edge node. The terminal where the application is located sends the old IP address to the old edge node. The edge node sends an interaction request, and the terminal exchanges target business traffic corresponding to the domain name with the old edge node. For this process, please refer to the instructions in adjustment method C2 and will not be described in detail here.

When scheduling the target business traffic, the scheduler sends the second correspondence between the interaction request and the IP address to each edge node, and each edge node stores the second correspondence respectively, and the interaction request in the second correspondence It is used to exchange the target business traffic corresponding to the domain name. The IP address in the second correspondence relationship is a new IP address used to indicate the new edge node.

After scheduling the target service traffic, because the terminal has not obtained the new IP address of the new edge node, the terminal still Then send an interaction request to the old edge node according to the old IP address. After the old edge node receives the interaction request, it queries the second correspondence relationship based on the interaction request and obtains the new IP address corresponding to the interaction request. Since the new IP address is different from the IP address of the old edge node itself, the old edge node returns the new IP address corresponding to the interaction request to the terminal, as well as a return code that conforms to universal standards. The return code is used to instruct the terminal to access the new IP address. , so that the terminal obtains the new IP address of the new edge node. After that, the terminal can send interaction requests to the new edge node according to the new IP address. After the new edge node receives the interaction request, it queries the second correspondence relationship based on the interaction request and also obtains the new IP address corresponding to the interaction request. Since the new IP address is the same as the IP address of the new edge node itself, the new edge node can interact with the terminal for target service traffic.

Referring to Figure 12, the traffic scheduling process in the 302 scheduling scenario is explained by taking reducing the proportion of the target service traffic of the first node and the second node interacting with the terminal, that is, scheduling the target service traffic from the first node to the second node, as an example. Before scheduling the target service traffic, the terminal sends an interaction request to the old IP address, that is, the first node indicated by IP A, and the terminal exchanges the target service traffic with the first node. When scheduling the target business traffic, the scheduling executor in the scheduler sends a second correspondence to the first node and the second node. The second correspondence is: an interaction request for interacting with the target business traffic, corresponding to IP B. After scheduling the target business traffic, the terminal first sends an interaction request to the first node. The first node queries the second corresponding relationship based on the interaction request to obtain IP B. Since IP B is different from the IP A of the first node, the first node The terminal returns a 302 response, which includes IP B. Then, the terminal sends an interaction request to the second node according to IP B. The second node queries the second correspondence based on the interaction request and also obtains IP B. Since IP B is the same as the IP B of the second node, the second node can interact with the terminal. Target business traffic.

Referring to Figure 13, the traffic scheduling process in the 302 scheduling scenario is explained by taking increasing the ratio of the first node and the second node to interact with the terminal and the target service traffic, that is, scheduling the target service traffic from the second node to the first node, as an example. Before scheduling the target service traffic, the terminal sends an interaction request to the old IP address, that is, the second node indicated by IP B, and the terminal exchanges the target service traffic with the second node. When scheduling the target business traffic, the scheduling executor in the scheduler sends a second correspondence to the second node and the first node. The second correspondence is: an interaction request for interacting with the target business traffic, corresponding to IP A. After scheduling the target business traffic, the terminal first sends an interaction request to the second node. The second node queries the first correspondence based on the interaction request to obtain IP A. Since IP A is different from the second node’s IP B, the second node The terminal returns a 302 response, which includes IP A. Then, the terminal sends an interaction request to the first node according to IP A. The first node queries the first corresponding relationship based on the interaction request and also obtains IP A. Since IP A is the same as the IP A of the first node, the first node can interact with the terminal. Target business traffic.

According to the description in adjustment method C1, the interaction request can carry a domain name. In addition, the schedulers shown in Figures 12 and 13 both include a scheduling perceptron, a scheduling decision maker and a scheduling executor. The content of the interaction between the first node and the second node and the scheduling perceptron, the content of the interaction between the scheduling perceptron and the scheduling decision maker, and the content of the interaction between the scheduling decision maker and the scheduling executor can all be found in the corresponding description in Figure 9 above. Here No further details will be given.

Adjustment method C3 is suitable for application layer scheduling scenarios. The target node is a terminal, and the scheduling information is the second correspondence between the interaction request and the IP address. The IP address indicates the first node or the second node, and the second correspondence is for the terminal. Determine the IP address corresponding to the interaction request, and send the interaction request to the node indicated by the IP address to interact with the target business traffic.

The application layer scheduling scenario can be seen in Figure 14. The application layer scenario includes but is not limited to the terminal, the application layer server corresponding to the CDN, the edge node in the CDN, the origin server and the scheduler. The terminal includes applications and application layer components. The edge node includes but is not limited to a first node and a second node. The IP address of the first node is IP A and the IP address of the second node is IP B. Among them, the terminal is connected to the application layer server, edge node and scheduler respectively, and the edge node is also connected to the source server and scheduler. catch. In addition, the scheduler and the application layer server may be integrated into the same device, or the scheduler and the application layer server may be two different devices connected. The scheduler may synchronize the second correspondence relationship delivered to the terminal to the application layer server.

Before scheduling target business traffic, when a user of the terminal enters a domain name for the application program, the application program receives the domain name input by the user of the terminal. The application generates an interaction request corresponding to the domain name, and the interaction request carries the domain name. The application sends an interaction request to the application layer component. The application layer component determines the IP address corresponding to the interaction request and returns the IP address corresponding to the interaction request to the application. The IP address indicates an edge node. As a result, the terminal can send an interaction request to the edge node indicated by the IP address through the application program, so that the terminal can interact with the edge node indicated by the IP address with target business traffic corresponding to the domain name.

For example, the application layer server includes but is not limited to a hypertext transfer protocol (HTTP) DNS server, and accordingly, the interaction request may be an HTTP request.

In some implementations, the application layer component stores a correspondence between the interaction request and the IP address, and the correspondence is obtained through the interaction between the application layer component and the application layer server. After the application layer component receives the interaction request sent by the application, it can query and hit the corresponding relationship based on the interaction request, obtain the IP address corresponding to the interaction request, and return it to the application.

In other embodiments, the application layer does not store the corresponding relationship between the interaction request and the IP address. After the application layer component receives the interaction request sent by the application program, it cannot hit the corresponding relationship. Therefore, the application layer component needs to communicate with the application layer. The server interacts to obtain the corresponding relationship. During the interaction process, the interaction request and the local IP address of the terminal are sent to the application layer server. The local IP address of the terminal indicates the location of the terminal. The application layer server determines the IP address corresponding to the interaction request based on the received interaction request and the local IP address of the terminal. The edge node indicated by the IP address is the edge node included in the CDN corresponding to the application layer server, and is the same as the location of the terminal. The nearest edge node. Afterwards, the application layer server returns the IP address corresponding to the interaction request to the application layer component, and the application layer component can return the IP address corresponding to the interaction request to the application program. The application layer component also stores the correspondence between the interaction request and the IP address.

According to the above description, before the target business traffic is scheduled, the correspondence between the interaction request and the IP address stored by the application layer component is the old correspondence obtained by the interaction between the application layer component and the application layer server. The old correspondence Including the old IP address, which indicates the old edge node, the terminal will exchange target service traffic with the old edge node.

When scheduling the target business traffic, since the scheduler delivers the second correspondence to the terminal, that is, the second correspondence between the interaction request and the IP address, the application layer component included in the terminal replaces the old correspondence with the second correspondence. Correspondence relationship, a new correspondence relationship is obtained, the new correspondence relationship includes a new IP address, and the new IP address indicates a new edge node. After scheduling the target business traffic, when the terminal sends an interaction request to the application layer component, the application layer component will query the new correspondence based on the interaction request, thereby returning a new IP address to the application, and the terminal will interact with the new edge node. business traffic. For example, the scheduler can flexibly select a manner of delivering the second correspondence relationship to the terminal based on the application layer component. The manner in which the scheduler delivers the second correspondence relationship to the terminal is not limited here.

Referring to Figure 14, the traffic scheduling process in the application layer scheduling scenario is explained by taking reducing the proportion of target service traffic between the first node and the second node and the terminal, that is, scheduling the target service traffic from the first node to the second node, as an example.

Before scheduling the target business traffic, the old correspondence stored by the application layer component is: the interaction request corresponding to the domain name, corresponding to IP A. If the application layer component receives the interaction request sent by the application program, then the IP address returned by the application layer component to the application program is IP A, then the terminal sends an interaction request carrying the interaction request to the first node indicated by IP A, and the terminal interacts with the first node. business traffic. When scheduling the target business traffic, the scheduler sends a second correspondence to the application layer component. The second correspondence is: the interaction request corresponding to the domain name corresponds to IP B. The application layer component sends the second correspondence to IP B. The two correspondences serve as new correspondences. After scheduling the target business traffic, if the application layer component receives the interaction request sent by the application program, the IP address returned by the application layer component to the application program is IP B, so the terminal sends a message carrying the IP address to the second node indicated by IP B. In the interaction request, the terminal exchanges target service traffic with the second node, but no longer interacts with the first node, thereby achieving scheduling of the target service flow from the first node to the second node.

Above, steps 201 to 203 have been described. In addition to these steps, the traffic scheduling method provided by the embodiment of the present application also includes some other steps, see the following description.

In an exemplary embodiment, after the scheduler adjusts the proportion of the first node and the second node to interact with the terminal based on the bandwidth occupied by at least two service flows on the first node, the method further includes: responding to the adjusted The ratio is such that the bandwidth occupied by the target service flow on the first node is zero, and the service quality of the first service flow is lower than the service quality indicated by the SLA of the first service flow, the scheduler issues an alarm, and the first service flow is at least two Business traffic other than the target business traffic among the business traffic.

Referring to Figures 6 and 7, after completing the scheduling of the target service traffic, that is, after adjusting the proportion, if the adjusted proportion makes the bandwidth occupied by the target service traffic on the first node be zero, it means that the bandwidth occupied by the target service traffic on the first node is zero. There is no target service flow. If the actual service quality of the first service flow still cannot meet the service quality indicated by the SLA of the first service flow at this time, an abnormality may exist and an alarm is required.

For example, the abnormal situation is that there is a problem with the distribution of business traffic, or that too much business traffic is allocated on the first node. Then the scheduler issues an alarm and may prompt to redistribute the service traffic, such as scheduling part of the first service traffic on the first node to the second node. For another example, the abnormal situation is that there is a problem with the bandwidth configuration of the first node, or that the first node can provide less bandwidth. Then, when the scheduler issues an alarm, it can prompt the first node to reconfigure the bandwidth, such as increasing the bandwidth that the first node can provide. For another example, the abnormal situation is that there is an unexpected sudden increase in the first business traffic and needs to be adjusted. Correspondingly, when the scheduler issues an alarm, the scheduler may prompt to analyze the cause of the sudden increase in the first service flow, and adjust the first service flow based on the analyzed cause.

In an exemplary embodiment, before the scheduler adjusts the ratio of the first node and the second node to interact with the terminal based on the bandwidth occupied by at least two service flows on the first node, the method further includes: the scheduler sends a request to the first node. A node sends at least two business flows with priorities, and the priorities of at least two business flows are used for the first node to configure at least two queues and process at least two business flows through at least two queues. For at least two business flows, Any business flow, the priority of any business flow matches the priority of the queue used to process any business flow.

On the first node, the bandwidth occupied by each business traffic may increase suddenly. This sudden increase is uncertain and often difficult to predict. Moreover, the scheduling of target business traffic may take some time to take effect. Therefore, in this embodiment of the present application, the business traffic can be processed through the configured queue before scheduling the target business traffic, that is, before adjusting the ratio. This can achieve fast and timely scheduling of business traffic, avoid business traffic with different priorities from seizing each other's bandwidth, and avoid the actual business traffic decline that affects user experience.

For example, when the first node configures at least two queues based on the priorities of at least two service flows, the priorities of the service flows and the priorities of the queues can be in one-to-one correspondence. For example, see Figure 15, when there are 8 different priorities for business traffic, the first node can also configure 8 queues with different priorities, namely queue 1 to queue 8. Correspondingly, for any one of the at least two service flows, the priority of any one service flow matches the priority of the queue used to process any one service flow. That is, if the priority of a business flow is higher, the data packets included in the business flow are added to the queue with a higher priority, so that the data packets included in the business flow are processed with higher priority, for example, through the first The node's port is forwarded with higher priority.

Among them, each queue obtains the tokens in the token bucket in order from high to low priority, and relies on the tokens to process data packets. The token bucket is an internal storage pool of the first node, and the first node fills the token bucket with tokens at a certain rate, and the tokens represent the resources required to process the data packet. When the number of tokens in the token bucket reaches the set capacity of the token bucket, excess tokens can be discarded. Therefore, the higher the priority of a queue, the more priority it can obtain the token, so that the data packets added to the queue can be processed with more priority.

By using the configured queue in conjunction with the above-mentioned steps 201 to 203, when business traffic suddenly increases, the configured queue can be used to respond quickly first, ensuring that high-priority business traffic can occupy sufficient bandwidth, so that high-priority business traffic can occupy sufficient bandwidth. The actual service quality of priority service traffic meets the SLA requirements. After that, the target service traffic is scheduled through steps 201 to 203. For example, the service traffic with lower priority is used as the target service flow, so that the service traffic with lower priority can also occupy sufficient bandwidth through scheduling. Then even for lower-priority service traffic, the actual service quality can be guaranteed, so that the actual service quality of low-priority service traffic can also meet SLA requirements.

In related technologies, the bandwidth that a node can provide is evenly distributed to each service flow on the node, without considering the priorities of different service flows. For example, see Figure 16, the node allocates an average bandwidth of 60Mbps to each business flow. Moreover, each business flow is scheduled independently, and each business flow can only use the allocated bandwidth. That is, the bandwidth corresponding to different business flows cannot be reused with each other. The bandwidth occupied by traffic in different transmission directions on a node is often not the same. There are certain differences, which results in low bandwidth reuse rate, high bandwidth cost, and inflexible scheduling process.

In the embodiment of this application, the priorities of different business flows are clearly distinguished, the target business traffic is selected based on the priority, and the bandwidth occupied by each business flow is globally monitored. When global coordinated scheduling is required, the first node is adjusted The ratio of the target service flow to the second node and the terminal is interacted with to realize the scheduling of the target service flow. Since priority is considered in the scheduling process and global collaborative scheduling is performed, it not only helps improve the bandwidth reuse rate and reduce bandwidth costs, but also makes the scheduling process more flexible.

In addition, the target service traffic in the embodiment of the present application may be service traffic with lower priority and larger data volume. Such target service traffic is suitable for scheduling and can provide sufficient scheduling space. When bandwidth is insufficient, lower-priority service traffic can yield bandwidth to other higher-priority service flows, thereby ensuring that the SLAs of higher-priority service flows are met first. When there is excess bandwidth, lower-priority business traffic can occupy the excess bandwidth, which will not affect the SLA of higher-priority business traffic and ensure that the SLA of lower-priority business traffic is met.

The above describes the traffic scheduling method provided by the embodiment of the present application. Corresponding to the above method, the embodiment of the present application also provides a traffic scheduling device. Wherein, the device is applied to a scheduler included in a traffic scheduling system. The traffic scheduling system also includes a first node, a second node and a terminal. The first node and the second node are used to interact business traffic with the terminal. This device is used to execute the traffic scheduling method executed by the scheduler in Fig. 2 through each module shown in Fig. 17. As shown in Figure 17, the traffic scheduling device provided by the embodiment of the present application includes the following modules.

The acquisition module 1701 is used to obtain the priorities of at least two service flows on the first node, and the bandwidth occupied by the at least two service flows on the first node, and the transmission direction of the at least two service flows is the same;

Determining module 1702, configured to determine the target service flow among the at least two service flows based on the priorities of the at least two service flows;

The adjustment module 1703 is configured to adjust the proportion of the target service traffic of the first node and the second node interacting with the terminal based on the bandwidth occupied by at least two service flows on the first node.

In an exemplary embodiment, the determining module 1702 is configured to determine, based on the priorities of at least two business flows, at least two Determine at least one service flow with the lowest priority among the service flows; determine the target service flow based on at least one service flow with the lowest priority.

In an exemplary embodiment, the determination module 1702 is configured to, in response to the existence of at least two lowest priority service flows, determine the service flow whose data amount is greater than the data amount threshold among the at least two lowest priority service flows as the target service. flow.

In an exemplary embodiment, the priority is determined based on at least one information of SLA and connection type. SLA is used to indicate service quality. The priority is positively related to the service quality indicated by SLA. The priority is related to the scheduling convenience corresponding to the connection type. Negative correlation.

In an exemplary embodiment, the adjustment module 1703 is configured to determine the sum of bandwidths occupied by at least two business flows on the first node; determine a proportional adjustment method based on the sum of bandwidths; and adjust the first proportional adjustment method according to the proportional adjustment method. The proportion of target business traffic between the node and the second node interacting with the terminal.

In an exemplary embodiment, the adjustment module 1703 is configured to, in response to the sum of bandwidths meeting the first condition, determine the proportion adjustment method to reduce the proportion of the target service traffic of the first node and the second node interacting with the terminal, so that the target service flow is within The bandwidth occupied on the first node reduces the bandwidth scheduling amount, and the bandwidth occupied by the target business traffic on the second node increases the bandwidth scheduling amount;

The device also includes: an adding module for increasing the bandwidth occupied by the first service flow on the first node. The first service flow is the service flow other than the target service flow among the at least two service flows, and the increased bandwidth is less than or equal to Equal to the bandwidth scheduling amount.

In an exemplary embodiment, the adjustment module 1703 is further configured to, in response to the sum of bandwidths being greater than the first bandwidth threshold, determine that the sum of bandwidths satisfies the first condition; or, in response to the sum of bandwidths corresponding to a bandwidth cost being greater than the first cost threshold. , determine that the sum of bandwidths satisfies the first condition.

In an exemplary embodiment, the adjustment module 1703 is configured to, in response to the sum of bandwidths meeting the second condition, determine the proportion adjustment method to increase the proportion of the target service traffic of the first node and the second node interacting with the terminal, so that the target service flow is within The bandwidth occupied by the first node increases the bandwidth scheduling amount, and the bandwidth occupied by the target service traffic on the second node decreases the bandwidth scheduling amount.

In an exemplary embodiment, the adjustment module 1703 is further configured to, in response to the sum of the bandwidths being less than or equal to the second bandwidth threshold, determine that the sum of the bandwidths satisfies the second condition and the second bandwidth threshold is less than or equal to the first bandwidth threshold; or, In response to the bandwidth cost corresponding to the sum of bandwidths being less than or equal to the second cost threshold, it is determined that the sum of bandwidths satisfies the second condition and the second cost threshold is less than or equal to the first cost threshold.

In an exemplary embodiment, the bandwidth scheduling amount is determined based on the sum of bandwidths, the first bandwidth threshold, and the bandwidth usage of the second node.

In an exemplary embodiment, the bandwidth scheduling amount is determined based on the bandwidth cost corresponding to the sum of bandwidths, the first cost threshold, and the bandwidth cost usage of the second node.

In an exemplary embodiment, the adjustment module 1703 is configured to determine a proportional adjustment method in response to the sum of bandwidths being different from the bandwidth occupied by the second service flow on the first node. The transmission directions of at least two service flows are consistent with the second service flow. Traffic travels in different directions.

In an exemplary embodiment, the proportion adjustment method is determined based on the sum of bandwidth, the bandwidth occupied by the second service traffic on the first node, and the bandwidth usage of the second node.

In an exemplary embodiment, the adjustment module 1703 is configured to determine the scheduling information corresponding to the target node in a proportional adjustment manner, and send the scheduling information to the target node. The scheduling information and the target node are used to adjust the relationship between the first node and the second node and the terminal node. Proportion of end-to-end interaction target business traffic.

In an exemplary embodiment, the target node is a CDN DNS server, the scheduling information is the first correspondence between the reference domain name and the IP address, the IP address indicates the first node or the second node; the first correspondence is for the CDN DNS server Based on the reference domain name sent by the terminal, the IP address corresponding to the reference domain name is returned to the terminal. The IP address is used by the terminal to send an interaction request to the node indicated by the IP address to exchange target business traffic.

In an exemplary embodiment, the target nodes are the first node and the second node, the scheduling information is the second correspondence between the interaction request and the IP address, the IP address indicates the first node or the second node; the second correspondence is represented by After the target node receives the interaction request sent by the terminal, in response to determining that the node indicated by the IP address corresponding to the interaction request is the current node based on the second correspondence relationship, the target node interacts with the target service flow with the terminal, or in response to determining based on the second correspondence relationship that the node indicated by the IP address corresponding to the interaction request is the current node. If the relationship determines that the node indicated by the IP address corresponding to the interaction request is not the current node, the IP address corresponding to the interaction request is returned to the terminal. The IP address is used by the terminal to send the interaction request to the node indicated by the IP address to exchange target business traffic.

In an exemplary embodiment, the target node is a terminal, and the scheduling information is the second correspondence between the interaction request and the IP address, where the IP address indicates the first node or the second node; the second correspondence is used by the terminal to determine the interaction request correspondence IP address, send an interaction request to the node indicated by the IP address to interact with the target business traffic.

In an exemplary embodiment, the bandwidth occupied by at least two service flows on the first node is obtained by at least one of detection and prediction.

In an exemplary embodiment, the adjustment module 1703 is also configured to respond to the adjusted ratio so that the bandwidth occupied by the target service flow on the first node is zero, and the service quality of the first service flow is lower than that of the first service flow. Based on the service quality indicated by the SLA, an alarm is issued, and the first service flow is the service flow other than the target service flow among at least two service flows.

In an exemplary embodiment, the adjustment module 1703 is also configured to send the priorities of at least two service flows to the first node, and the priorities of the at least two service flows are used by the first node to configure at least two queues, through at least two Each queue processes at least two service flows, and for any one of the at least two service flows, the priority of any one service flow matches the priority of the queue used to process any one service flow.

In the embodiment of this application, the priorities of different business traffic are distinguished, the target business traffic is selected based on the priority, and the bandwidth occupied by each business traffic is globally monitored. When global coordinated scheduling is required, the first node and the third node are adjusted. The two nodes interact with the terminal based on the ratio of target business traffic to achieve scheduling of target business traffic. Since priority is considered in the scheduling process and global collaborative scheduling is performed, it not only helps improve the bandwidth reuse rate and reduce bandwidth costs, but also makes the scheduling process more flexible.

It should be understood that when the device provided in Figure 17 realizes its functions, the division of the above functional modules is only used as an example. In actual applications, the above functions can be allocated to different functional modules according to needs, that is, the equipment The internal structure is divided into different functional modules to complete all or part of the functions described above. In addition, the apparatus and method embodiments provided in the above embodiments belong to the same concept, and the specific implementation process can be found in the method embodiments, which will not be described again here.

Referring to Figure 18, Figure 18 shows a schematic structural diagram of an exemplary traffic scheduling device 1800 of the present application. The traffic scheduling Device 1800 includes at least one processor 1801 and memory 1803 .

The processor 1801 is, for example, a general central processing unit (Central Processing Unit, CPU), a digital signal processor (digital signal processor, DSP), a network processor (network processor, NP), a GPU, or a neural network processor (neural-network processing). units (NPU), data processing unit (Data Processing Unit, DPU), microprocessor or one or more integrated circuits or application-specific integrated circuits (ASIC), programmable logic used to implement the solution of this application device (programmable logic device, PLD), other general-purpose processors or other programmable logic devices, discrete gates, transistor logic devices, discrete hardware components, or any combination thereof. PLD is, for example, a complex programmable logic device (CPLD), a field-programmable gate array (FPGA), a general array logic (GAL), or any combination thereof. A general-purpose processor can be a microprocessor or any conventional processor, etc. It is worth noting that the processor may be a processor that supports advanced RISC machines (ARM) architecture. It may implement or execute the various logical blocks, modules, and circuits described in connection with this disclosure. The processor can also be a combination that implements computing functions, such as a combination of one or more microprocessors, a combination of a DSP and a microprocessor, and so on.

Optionally, the traffic scheduling device 1800 also includes a bus 1802. Bus 1802 is used to transmit information between various components of traffic scheduling device 1800. The bus 1802 may be a peripheral component interconnect (PCI) bus or an extended industry standard architecture (EISA) bus, etc. The bus 1802 can be divided into an address bus, a data bus, a control bus, etc. For ease of presentation, only one line is used in Figure 18, but it does not mean that there is only one bus or one type of bus.

Memory 1803 is, for example, volatile memory or non-volatile memory, or may include both volatile and non-volatile memory. Among them, non-volatile memory can be read-only memory (ROM), programmable ROM (PROM), erasable programmable read-only memory (erasable PROM, EPROM), electrically removable memory. Erase electrically programmable read-only memory (EPROM, EEPROM) or flash memory. Volatile memory can be random access memory (RAM), which is used as an external cache.

By way of illustration, but not limitation, many forms of ROM and RAM are available. For example, ROM is a compact disc read-only memory (CD-ROM). RAM includes but is not limited to static random access memory (static RAM, SRAM), dynamic random access memory (dynamic random access memory, DRAM), synchronous dynamic random access memory (synchronous DRAM, SDRAM), double data rate synchronous dynamic Random access memory (double data rate SDRAM, DDR SDRAM), enhanced synchronous dynamic random access memory (enhanced SDRAM, ESDRAM), synchronous link dynamic random access memory (synchlink DRAM, SLDRAM) and direct memory bus random access memory (direct rambus RAM, DR RAM).

Memory 1803 may also be other types of storage devices that may store static information and instructions. Or it could be other types of dynamic storage devices that can store information and instructions. Or it can be other optical disk storage, optical disc storage (including compressed optical discs, laser discs, optical discs, digital versatile discs, Blu-ray discs, etc.), magnetic disk storage media or other magnetic storage devices, or can be used to carry or store instructions or data structures without limitation, any other medium in the form of the desired program code and capable of being accessed by a computer. The memory 1803 exists independently, for example, and is connected to the processor 1801 through the bus 1802 . Memory 1803 may also be integrated with processor 1801.

In specific implementations, as some embodiments, the processor 1801 may include one or more CPUs, such as CPU0 and CPU1 as shown in FIG. 18 . Each of these processors can be a single-core processor or a multi-core processor device. A processor here may refer to one or more devices, circuits, and/or processing cores for processing data (eg, computer program instructions).

In specific implementations, as some implementations, the traffic scheduling device 1800 may include multiple processors, such as two processors 1801 as shown in Figure 18. Each of these processors can be a single-core processor or a multi-core processor. A processor here may refer to one or more devices, circuits, and/or processing cores for processing data (such as computer program instructions).

In some embodiments, the memory 1803 is used to store program instructions 1804 for executing the solution of the present application, and the processor 1801 can execute the program instructions 1804 stored in the memory 1803. That is to say, the traffic scheduling device 1800 can implement the method provided by the method embodiment, that is, the traffic scheduling method shown in Figure 2, through the processor 1801 and the program instructions 1804 in the memory 1803. Program instructions 1804 may include one or more software modules. Optionally, the processor 1801 itself can also store program instructions for executing the solution of the present application.

During specific implementation, the traffic scheduling device 1800 of the present application may correspond to the scheduler used to perform the above method. The processor 1801 in the traffic scheduling device 1800 reads the instructions in the memory 1803 to enable the traffic scheduling shown in Figure 18 The device 1800 is capable of performing all or part of the steps in the method embodiments.

The traffic scheduling device 1800 may also correspond to the device shown in Figure 17 above. Each functional module in the device shown in Figure 17 is implemented using the software of the traffic scheduling device 1800. In other words, the functional modules included in the device shown in Figure 17 are generated by the processor 1801 of the traffic scheduling device 1800 after reading the program instructions 1804 stored in the memory 1803.

Each step of the method shown in Figure 2 is completed through an integrated logic circuit of hardware or instructions in the form of software in the processor of the traffic scheduling device 1800. The steps of the method embodiments disclosed in this application can be directly implemented by a hardware processor, or executed by a combination of hardware and software modules in the processor. The software module can be located in random access memory, flash memory, read-only memory, programmable read-only memory or electrically erasable programmable memory, registers and other mature storage media in this field. The storage medium is located in the memory, and the processor reads the information in the memory and completes the steps of the above method embodiment in combination with its hardware. To avoid repetition, the details will not be described here.

In an exemplary embodiment, a computer program (product) is provided. The computer program (product) includes: computer program code. When the computer program code is run by a computer, it causes the computer to perform any of the above exemplary traffic scheduling methods.

In an exemplary embodiment, a computer-readable storage medium is provided. The computer-readable storage medium stores programs or instructions. When the program or instructions are run on a computer, the computer executes any of the above exemplary traffic scheduling methods.

In an exemplary embodiment, a chip is provided, including a processor for calling and running instructions stored in the memory, so that a computer installed with the chip executes any of the above exemplary traffic scheduling methods.

In an exemplary embodiment, another chip is provided, including: an input interface, an output interface, a processor, and a memory. The input interface, the output interface, the processor, and the memory are connected through an internal connection path. The processor is used to execute the memory. When the code is executed, the computer installed with the chip executes any of the above exemplary traffic scheduling methods.

In the above embodiments, it may be implemented in whole or in part by software, hardware, firmware, or any combination thereof. When implemented using software, it may be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, the processes or functions described in this application are generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions may be stored in or transmitted from one computer-readable storage medium to another, e.g., the computer instructions may be transferred from a website, computer, server, or data center Transmission to another website, computer, server or data center through wired (such as coaxial cable, optical fiber, digital subscriber line) or wireless (such as infrared, wireless, microwave, etc.) means. The computer-readable storage medium may be any available medium that can be accessed by a computer or a data storage device such as a server, data center, etc. that contains one or more available media integrated therein. The available media may be magnetic media (eg, floppy disk, hard disk, magnetic tape), optical media (eg, DVD), or semiconductor media (eg, Solid State Disk), etc.

In this application, the terms "first", "second" and other words are used to distinguish the same or similar items with basically the same functions and functions. It should be understood that the terms "first", "second" and "nth" There is no logical or sequential dependency, and there is no limit on the number or execution order. It should also be understood that, although the following description uses the terms first, second, etc. to describe various elements, these elements should not be limited by the terms. These terms are only used to distinguish one element from another.

It should also be understood that in each embodiment of the present application, the size of the sequence number of each process does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not be determined by the execution order of the embodiments of the present application. The implementation process constitutes no limitation.

The term "at least one" in this application means one or more, and the term "plurality" in this application means two or more. For example, a plurality of second devices means two or more Secondary device above. The terms "system" and "network" are often used interchangeably in this article.

It is to be understood that the terminology used in the description of the various examples herein is for the purpose of describing the particular example only and is not intended to be limiting. As used in the description of various described examples and the appended claims, the singular forms "a," "an" and "the" are intended to include the plural forms as well, unless the context dictates otherwise. Instruct clearly.

It will also 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. The term "and/or" is an association relationship that describes related objects, indicating that there can be three relationships. For example, A and/or B can mean: A alone exists, A and B exist simultaneously, and B alone exists. situation. In addition, the character "/" in this application generally indicates that the related objects are in an "or" relationship.

It should also be understood that the terms "if" and "if" may be interpreted to mean "when" or "upon" or "in response to determining" or "in response to detecting." Similarly, depending on the context, the phrase "if it is determined..." or "if [stated condition or event] is detected" may be interpreted to mean "when it is determined..." or "in response to the determination... ” or “on detection of [stated condition or event]” or “in response to detection of [stated condition or event].”

The above are only examples of the present application and are not intended to limit the present application. Any modifications, equivalent substitutions, improvements, etc. made within the principles of the present application shall be included in the protection scope of the present application.

Claims

A traffic scheduling method, characterized in that the method is applied to a scheduler included in a traffic scheduling system. The traffic scheduling system also includes a first node, a second node and a terminal. The first node and the second The node is used to exchange service traffic with the terminal, and the method includes:

The scheduler obtains the priorities of at least two service flows on the first node, the bandwidth occupied by the at least two service flows on the first node, and the transmission direction of the at least two service flows. same;

The scheduler determines the target service flow among the at least two service flows based on the priorities of the at least two service flows;

The scheduler adjusts the proportion of the first node and the second node that interact with the terminal with the target service flow based on the bandwidth occupied by the at least two service flows on the first node.
The method according to claim 1, characterized in that the scheduler determines the target service flow among the at least two service flows based on the priorities of the at least two service flows, including:

The scheduler determines at least one service flow with the lowest priority among the at least two service flows based on the priorities of the at least two service flows;

The scheduler determines the target service flow based on the at least one lowest priority service flow.
The method according to claim 2, wherein the scheduler determines the target service flow based on the at least one lowest priority service flow, including:

In response to the presence of at least two service flows with the lowest priority, the scheduler determines, among the at least two service flows with the lowest priority, the service flow whose data amount is greater than the data amount threshold as the target service flow.
The method according to any one of claims 1-3, characterized in that the priority is determined based on at least one information of a service level agreement (SLA) and a connection type, the SLA is used to indicate service quality, and the priority The priority is positively related to the service quality indicated by the SLA, and the priority is negatively related to the scheduling convenience corresponding to the connection type.
The method according to any one of claims 1 to 4, characterized in that the scheduler adjusts the first node and the third node based on the bandwidth occupied by the at least two service flows on the first node. The proportion of the target service traffic that the two nodes interact with the terminal includes:

The scheduler determines the sum of the bandwidths occupied by the at least two service flows on the first node;

The scheduler determines the proportion adjustment method based on the sum of the bandwidths;

The scheduler adjusts the proportion of the first node and the second node interacting with the terminal in the target service flow according to the proportion adjustment method.
The method according to claim 5, characterized in that the scheduler determines the proportion adjustment method based on the sum of the bandwidths, including:

In response to the sum of the bandwidths meeting the first condition, the scheduler determines that the proportion adjustment method is to reduce the proportion of the first node and the second node interacting with the terminal with the target service traffic, so that the Where the target business traffic is located The bandwidth occupied by the first node is reduced by the bandwidth scheduling amount, and the bandwidth occupied by the target service flow on the second node is increased by the bandwidth scheduling amount;

The method also includes:

Increase the bandwidth occupied by the first service flow on the first node, the first service flow is the service flow other than the target service flow in the at least two service flows, and the increased bandwidth is less than or equal to The bandwidth scheduling amount.
The method of claim 6, further comprising:

In response to the sum of bandwidths being greater than a first bandwidth threshold, determining that the sum of bandwidths satisfies the first condition;

Alternatively, in response to a bandwidth cost corresponding to the sum of bandwidths being greater than a first cost threshold, it is determined that the sum of bandwidths satisfies the first condition.
The method according to claim 5, characterized in that the scheduler determines the proportion adjustment method based on the sum of the bandwidths, including:

In response to the sum of the bandwidths meeting the second condition, the scheduler determines that the proportion adjustment method is to increase the proportion of the first node and the second node interacting with the terminal with the target service traffic, so that the The bandwidth occupied by the target service flow on the first node is increased by the bandwidth scheduling amount, and the bandwidth occupied by the target service flow on the second node is reduced by the bandwidth scheduling amount.
The method of claim 8, further comprising:

In response to the sum of bandwidths being less than or equal to a second bandwidth threshold, determining that the sum of bandwidths satisfies the second condition, and the second bandwidth threshold is less than or equal to the first bandwidth threshold;

Alternatively, in response to the bandwidth cost corresponding to the sum of bandwidths being less than or equal to a second cost threshold, it is determined that the sum of bandwidths satisfies the second condition, and the second cost threshold is less than or equal to the first cost threshold.
The method according to claim 7 or 9, characterized in that the bandwidth scheduling amount is determined based on the sum of the bandwidth, the first bandwidth threshold and the bandwidth usage of the second node.
The method according to claim 7 or 9, characterized in that the bandwidth scheduling amount is determined based on the bandwidth cost corresponding to the sum of the bandwidth, the first cost threshold and the bandwidth cost usage of the second node.
The method according to claim 5, characterized in that the scheduler determines the proportion adjustment method based on the sum of the bandwidths, including:

In response to the sum of the bandwidths being different from the bandwidth occupied by the second service flow on the first node, the scheduler determines the proportion adjustment method, and the transmission directions of the at least two service flows are consistent with the second service flow. Business traffic is transmitted in different directions.
The method according to claim 12, characterized in that the proportion adjustment method is determined based on the sum of the bandwidth, the bandwidth occupied by the second service traffic on the first node and the bandwidth usage of the second node. .
The method according to any one of claims 5 to 13, characterized in that the scheduler adjusts the first node and the second node to interact with the terminal in the target service flow according to the proportion adjustment method. proportions, including:

The scheduler determines the scheduling information corresponding to the target node according to the proportion adjustment method, and sends the scheduling information to the target node. The scheduling information and the target node are used to realize the adjustment of the first node and the target node. The proportion of the target service traffic that the second node interacts with the terminal.
The method according to claim 14, characterized in that the target node is a content distribution network CDN Domain Name System DNS server, the scheduling information is the first correspondence between a reference domain name and an Internet Protocol IP address, and the IP The address indicates the first node or the second node;

The first correspondence relationship is used for the CDN DNS server to return the IP address corresponding to the reference domain name to the terminal based on the reference domain name sent by the terminal, and the IP address is used for the node indicated by the terminal to the IP address. Send an interaction request to interact with the target service traffic.
The method of claim 14, wherein the target node is the first node and the second node, and the scheduling information is a second correspondence between an interaction request and an Internet Protocol IP address, The IP address indicates the first node or the second node;

The second correspondence relationship is used for the target node to determine, based on the second correspondence relationship, that the node indicated by the IP address corresponding to the interaction request is the current node after receiving the interaction request sent by the terminal. The terminal interacts with the target service traffic, or in response to determining that the node indicated by the IP address corresponding to the interaction request is not the current node based on the second correspondence relationship, returns the IP address corresponding to the interaction request to the terminal. , the IP address is used by the terminal to send an interaction request to the node indicated by the IP address to interact with the target service traffic.
The method of claim 14, wherein the target node is the terminal, the scheduling information is a second correspondence between an interaction request and an IP address, the IP address indicates the first node or the second node;

The second correspondence relationship is used for the terminal to determine the IP address corresponding to the interaction request, and send the interaction request to the node indicated by the IP address to interact with the target service traffic.
The method according to any one of claims 1 to 17, characterized in that the bandwidth occupied by the at least two service flows on the first node is obtained by at least one of detection and prediction.
The method according to any one of claims 1-18, characterized in that the scheduler adjusts the first node and the third node based on the bandwidth occupied by the at least two service flows on the first node. After the two nodes exchange the proportion of the target service traffic with the terminal, the method further includes:

In response to the adjusted proportion causing the bandwidth occupied by the target service flow on the first node to be zero, and the service quality of the first service flow being lower than the service quality indicated by the SLA of the first service flow, the The scheduler issues an alarm, and the first service flow is the service flow other than the target service flow among the at least two service flows.
The method according to any one of claims 1 to 19, characterized in that the scheduler adjusts the first node and the third node based on the bandwidth occupied by the at least two service flows on the first node. The two nodes interact with the terminal as described Before determining the proportion of target business traffic, the method also includes:

The scheduler sends the priorities of the at least two service flows to the first node, and the priorities of the at least two service flows are used by the first node to configure at least two queues. A queue processes the at least two business flows. For any one of the at least two business flows, the priority of the any one business flow is the same as the priority of the queue used to process the any one business flow. match.
A traffic scheduling device, characterized in that the device is applied to a scheduler included in a traffic scheduling system. The traffic scheduling system also includes a first node, a second node and a terminal. The first node and the second The node is used to exchange service traffic with the terminal, and the device includes:

Obtaining module, configured to obtain the priorities of at least two service flows on the first node, the bandwidth occupied by the at least two service flows on the first node, the transmission of the at least two service flows Same direction;

A determining module, configured to determine the target service flow among the at least two service flows based on the priorities of the at least two service flows;

An adjustment module, configured to adjust the proportion of the first node and the second node interacting with the terminal with the target service flow based on the bandwidth occupied by the at least two service flows on the first node.
The device according to claim 21, characterized in that the determining module is configured to determine at least one service flow with the lowest priority among the at least two service flows based on the priorities of the at least two service flows. ; Determine the target service flow based on the at least one service flow with the lowest priority.
The device according to claim 22, characterized in that the determining module is configured to, in response to the existence of at least two lowest priority service flows, determine whether the data amount in the at least two lowest priority service flows is greater than that of the data. The business traffic volume threshold is determined as the target business traffic volume.
The device according to any one of claims 21-23, characterized in that the priority is determined based on at least one information of a service level agreement (SLA) and a connection type, the SLA is used to indicate service quality, and the priority The priority is positively related to the service quality indicated by the SLA, and the priority is negatively related to the scheduling convenience corresponding to the connection type.
The device according to any one of claims 21-24, characterized in that the adjustment module is used to determine the sum of the bandwidths occupied by the at least two business flows on the first node; based on the The sum of the bandwidths determines the proportion adjustment method; according to the proportion adjustment method, the proportion of the first node and the second node interacting with the terminal and the target service traffic is adjusted.
The device according to claim 25, characterized in that the adjustment module is configured to, in response to the sum of the bandwidths meeting a first condition, determine the proportion adjustment method to reduce the first node and the second The node interacts with the terminal in a proportion of the target service flow such that the bandwidth occupied by the target service flow on the first node is reduced by the bandwidth scheduling amount, and the target service flow occupies on the second node The bandwidth is increased by the bandwidth scheduling amount;

The device further includes: an adding module for increasing the bandwidth occupied by the first service flow on the first node, the first service flow being the at least two service flows in addition to the target service flow. business traffic, the increased The bandwidth is less than or equal to the bandwidth scheduling amount.
The device according to claim 26, wherein the adjustment module is further configured to, in response to the sum of bandwidths being greater than a first bandwidth threshold, determine that the sum of bandwidths satisfies the first condition; or, in response to When the bandwidth cost corresponding to the sum of bandwidths is greater than the first cost threshold, it is determined that the sum of bandwidths satisfies the first condition.
The device according to claim 25, wherein the adjustment module is configured to, in response to the sum of the bandwidths meeting a second condition, determine that the proportion adjustment method is to increase the first node and the second The node interacts with the terminal in a proportion of the target service flow, such that the bandwidth occupied by the target service flow on the first node is increased by the bandwidth scheduling amount, and the bandwidth occupied by the target service flow on the second node is Bandwidth is reduced by the amount of bandwidth scheduled.
The device according to claim 28, wherein the adjustment module is further configured to, in response to the sum of bandwidths being less than or equal to a second bandwidth threshold, determine that the sum of bandwidths satisfies the second condition, so The second bandwidth threshold is less than or equal to the first bandwidth threshold; or, in response to the bandwidth cost corresponding to the sum of bandwidths being less than or equal to the second cost threshold, it is determined that the sum of bandwidths satisfies the second condition, and the third The second cost threshold is less than or equal to the first cost threshold.
The device according to claim 27 or 29, characterized in that the bandwidth scheduling amount is determined based on the sum of the bandwidth, the first bandwidth threshold and the bandwidth usage of the second node.
The device according to claim 27 or 29, characterized in that the bandwidth scheduling amount is determined based on the bandwidth cost corresponding to the sum of bandwidths, the first cost threshold and the bandwidth cost usage of the second node.
The device according to claim 25, wherein the adjustment module is configured to determine the proportion adjustment method in response to the sum of the bandwidths being different from the bandwidth occupied by the second service traffic on the first node. , the transmission directions of the at least two service flows are different from the transmission directions of the second service flow.
The device according to claim 32, wherein the proportion adjustment method is determined based on the sum of the bandwidth, the bandwidth occupied by the second service traffic on the first node and the bandwidth usage of the second node. .
The device according to any one of claims 25-33, characterized in that the adjustment module is used to determine the scheduling information corresponding to the target node according to the proportion adjustment method, and send the scheduling information to the target node, so The scheduling information and the target node are used to adjust the ratio of the first node and the second node to the terminal for interacting with the target service flow.
The device according to claim 34, characterized in that the target node is a content distribution network CDN domain name system DNS server, the scheduling information is the first correspondence between a reference domain name and an Internet Protocol IP address, and the IP The address indicates the first node or the second node; the first correspondence is used for the CDN DNS server to return the IP address corresponding to the reference domain name to the terminal based on the reference domain name sent by the terminal, and the IP The address is used by the terminal to send an interaction request to the node indicated by the IP address to interact with the target service traffic.
The device according to claim 34, wherein the target node is the first node and the second node, and the scheduling information is a second correspondence between an interaction request and an Internet Protocol IP address, The IP address indicates the first node or the second node; the second correspondence relationship is used for the target node to respond to determining the interaction request based on the second correspondence relationship after receiving the interaction request sent by the terminal. If the node indicated by the IP address corresponding to the interaction request is the current node, interact with the target service traffic with the terminal, or in response to determining based on the second correspondence that the node indicated by the IP address corresponding to the interaction request is not The current node returns the IP address corresponding to the interaction request to the terminal. The IP address is used by the terminal to send the interaction request to the node indicated by the IP address to interact with the target service traffic.
The device according to claim 34, wherein the target node is the terminal, the scheduling information is a second correspondence between an interaction request and an IP address, the IP address indicates the first node Or the second node; the second correspondence relationship is used for the terminal to determine the IP address corresponding to the interaction request, and send the interaction request to the node indicated by the IP address to interact with the target service traffic.
The device according to any one of claims 21 to 37, characterized in that the bandwidth occupied by the at least two service flows on the first node is obtained by at least one of detection and prediction.
The device according to any one of claims 21 to 38, characterized in that the adjustment module is further configured to make the bandwidth occupied by the target service flow on the first node be zero in response to the adjusted ratio, If the service quality of the first service flow is lower than the service quality indicated by the SLA of the first service flow, an alarm is issued. The first service flow is the at least two service flows other than the target service flow. business traffic.
The device according to any one of claims 21 to 39, characterized in that the adjustment module is further configured to send the priorities of the at least two business flows to the first node. The priority is used for the first node to configure at least two queues, and the at least two service flows are processed through the at least two queues. For any one of the at least two service flows, the any The priority of a service flow matches the priority of the queue used to process any one of the service flows.
A traffic scheduling device, characterized in that the device includes a memory and a processor; at least one instruction is stored in the memory, and the at least one instruction is loaded and executed by the processor, so that the traffic scheduling device Implement the traffic scheduling method described in any one of claims 1-20.
A computer-readable storage medium, characterized in that at least one instruction is stored in the computer-readable storage medium, and the instruction is loaded and executed by a processor, so that the computer implements any one of claims 1-20. The traffic scheduling method described above.
A computer program product, characterized in that the computer program product includes a computer program or instructions, and the computer program or instructions are executed by a processor to enable the computer to implement the traffic scheduling method described in any one of claims 1-20. Law.