WO2024045744A1 - Resource planning method and apparatus, and related device - Google Patents

Abstract

The present application provides a resource planning method, which is used for reasonably planning the amount of resources that can be used by each cloud service so as to ensure the normal operations of the cloud services. The resource planning method comprises: acquiring a first feature, wherein the first feature indicates usage of first resources by a first cloud service within a historical first period of time; determining a second feature according to the first feature, wherein the second feature indicates demand of the first cloud service for the first resources within a future second period of time; and determining a first allocated resource amount according to the second feature, the first allocated resource amount being a resource amount of the first resources allocated to the first cloud service within the second period of time. In addition, the present application further provides a corresponding apparatus, a computing device cluster, a computer-readable storage medium, and a computer program product.

Description

A resource planning method, device and related equipment

This application requires the priority of the Chinese patent application submitted to the State Intellectual Property Office of China on September 2, 2022, with the application number 202211071538.2 and the invention title "A method, device and related equipment for resource planning". All its contents have been approved. This reference is incorporated into this application.

Technical field

The present application relates to the field of data processing technology, and in particular to a resource planning method, device and related equipment.

Background technique

With the development of network technology, more and more users choose to use cloud services. After a user subscribes to a cloud service from a cloud service provider, the user can use the resources provided by the cloud service provider to achieve his or her own purposes. For example, for users who subscribe to cloud storage services, users can store data in the storage space provided by the cloud service provider. For users who subscribe to cloud live broadcast services, users can upload live broadcast data to the live broadcast server provided by the cloud service provider, and then send the live broadcast data to users watching the live broadcast through the downlink bandwidth provided by the cloud service provider.

Currently, cloud service providers can often provide multiple types of cloud services. For example, a cloud service provider can provide both cloud storage services and cloud live streaming services. No matter what type of cloud service you use, you need to use the resources of the cloud service provider. Since the total amount of resources owned by a cloud service provider is often limited, the cloud service provider needs to determine the amount of resources that each type of cloud service can use.

Contents of the invention

In view of this, this application provides a resource planning method to reasonably plan the amount of resources that each cloud service can use to ensure the normal operation of the cloud service. This application also provides corresponding devices, computing device clusters, computer-readable storage media, and computer program products.

In a first aspect, this application provides a resource planning method, which can be applied to a resource planning device. During specific implementation, the resource planning device may first obtain the first characteristic, and the first characteristic is used to indicate the usage of the first resource by the first cloud service in the first historical time period. Next, the resource planning device predicts the demand for the first resource by the first cloud service in the second time period in the future based on the usage of the first resource by the first cloud service in the first historical time period. The demand of the first resource by the first cloud service in the second time period in the future may be called the second characteristic. Then, the resource planning device determines the first allocated resource amount according to the second characteristic, and the first allocated resource amount is the resource amount of the first resource allocated for the first cloud service in the second time period.

In this way, the resource planning device can predict the future demand of the first resource by the first cloud service based on the historical demand of the first resource by the first cloud service, and automatically determine the need for the first resource in the second time period in the future. How many first resources are allocated by a cloud service. In this way, automatic resource planning can be realized and the rationality of cloud service resource planning can be improved. Moreover, since the above process can be completed by the resource planning device without manual participation, the frequency of cloud service resource planning can be increased, thereby improving the flexibility of resource planning.

In some possible implementations, the resource planning device may plan the resource amount of the first resource allocated to the first cloud service in multiple time periods in the future according to the first characteristic. Specifically, the resource planning device may determine the demand for the first resource by the first cloud service in the third time period in the future based on the first characteristic. The demand of the first resource by the first cloud service in the third time period in the future is called the third characteristic. Next, the resource planning module determines the second resource allocation according to the third characteristic The second resource allocation amount is the resource amount of the first resource allocated to the first cloud service in the third time period. The third time period and the second time period are different time periods. Optionally, the sum of the duration of the second time period and the duration of the third time period is less than or equal to the duration of the first time period. In this way, the resource demand of multiple time periods in the future is determined based on the resource demand of the first time period, and the resource amount of the first resource allocated to the first and range can be flexibly adjusted.

In some possible implementations, the second time period and the third time period may be divided according to the demand for the first resource. Specifically, the resource planning device may determine the demand for the first resource by the first cloud service in a longer period of time in the future based on the first characteristic. If the first cloud service has a greater demand for the first resource during a certain period of time and has a smaller demand for the first resource during another period of time, the resource planning device may divide the two periods of time into different time periods. That is to say, the difference between the demand of the first resource by the first cloud service in the second time period and the demand of the first resource by the first cloud service in the third time period is greater than the preset threshold. In this way, the time period is divided according to the difference between the resource requirements, and on the premise of ensuring that the resource amount of the first resource allocated to the first cloud service can meet the needs of the first cloud service, it is avoided that the resource amount allocated to the first cloud service is Repeated adjustment of the resource amount of the first resource.

In some possible implementations, the resource planning device can be used to plan resources that have a greater impact on cloud services. That is to say, if the first cloud service runs based on multiple resources, then the first resource planned by the resource planning device is the resource that has the greatest impact on the first cloud service. That is, if the first cloud service runs based on the first resource and the second resource, then the impact of the resource amount of the first resource on the first cloud service is greater than the impact of the resource amount of the second resource on the first cloud service. .

In some possible implementations, the first resource includes any one or more of the following: network bandwidth resources, Elastic Internet Protocol (Elastic Internet Protocol, EIP) address resources, storage space resources, and computing resources.

In some possible implementations, the resource planning device is further configured to plan the attributes of the first resource allocated to the first cloud service based on the demand of the first cloud service for the first resource. For example, assume that the first resource is a network bandwidth resource. Then when determining the first resource allocation amount, the resource planning device can not only determine the target bandwidth amount based on the first cloud service's demand for network bandwidth resources in the second time period, but also determine the target bandwidth amount based on the cloud service characteristics of the first cloud service. target line. The target line is a line that can satisfy the cloud service characteristics of the first cloud service. Next, the resource planning device determines the first allocated resource amount according to the target line and the target bandwidth amount. In this way, the first cloud service can use the bandwidth of the target line during the second time period, and the maximum bandwidth that the first cloud service can use does not exceed the target bandwidth amount. In this way, the first cloud service's demand for the resource amount of the network bandwidth resource can be satisfied, and the first cloud service's demand for the resource attribute of the network bandwidth resource can also be satisfied.

In some possible implementations, the first cloud service is more cost-sensitive, that is, the first and scope cloud service characteristics indicate that the first cloud service is a cost-sensitive cloud service. Then when determining the target line, the resource planning device can select the candidate line with the lowest cost and the remaining bandwidth quality not less than the target bandwidth amount as the target line from multiple alternative lines. In this way, the operating cost of the first cloud service can be reduced on the basis of ensuring that the first cloud service has sufficient network bandwidth resources.

In a second aspect, this application also provides a resource planning device. The device includes: a feature acquisition module, configured to acquire a first feature. The first feature indicates that the first cloud service has used the first resource in a first historical time period. usage; a feature prediction module, configured to determine a second feature based on the first feature, the second feature indicating the demand for the first resource by the first cloud service in a second time period in the future; a resource allocation module, configured to determine according to the second characteristic Determine a first allocated resource amount, which is the resource amount of the first resource allocated for the first cloud service in the second time period.

In some possible implementations, the feature prediction module is further configured to determine a third feature based on the first feature, and the third feature indicates that the first cloud service will predict the performance of the first cloud service in a third time period in the future. The demand situation of the first resource; the resource allocation module is also used to determine a second allocated resource amount according to the third characteristic, and the second allocated resource amount is the amount of the first cloud resource in the third time period. The resource amount of the first resource allocated by the service.

In some possible implementations, the demand of the first resource by the first cloud service in the second time period is different from the demand of the first resource by the first cloud service in the third time period. The difference between the demand situations of the first resource is greater than the preset threshold.

In some possible implementations, the first cloud service runs based on a first resource and a second resource, and the amount of the first resource has a greater impact on the first cloud service than the second resource. The impact of the amount of resources on the first cloud service.

In some possible implementations, the first resource includes any one or more of the following: network bandwidth resources, EIP address resources, storage space resources, and computing resources.

In some possible implementations, the first resource includes a network bandwidth resource, and the resource allocation module is specifically configured to be based on the demand for the network bandwidth resource by the first cloud service in the second time period. Determine the target bandwidth amount according to the situation; determine the target line according to the cloud service characteristics of the first cloud service; determine the first allocated resource amount according to the target line and the target bandwidth amount, and the first allocated resource amount indicates the The target line allocates network bandwidth resources of the target bandwidth amount to the first cloud service within the second time period.

In some possible implementations, the cloud service characteristics of the first cloud service indicate that the first cloud service is a cost-sensitive cloud service, and the resource allocation module is specifically used to select costs from multiple alternative lines. The candidate line that is the lowest and has a remaining bandwidth greater than the target bandwidth is used as the target line.

In a third aspect, the present application provides a computing device cluster. The computing device includes at least one computing device. The at least one computing device includes at least one processor and at least one memory; the at least one memory is used to store instructions. The at least one processor executes the instruction stored in the at least one memory, so that the computing device cluster executes the resource planning method in the above-mentioned first aspect or any possible implementation of the first aspect. It should be noted that the memory can be integrated into the processor or independent of the processor. The at least one computing device may also include a bus. Among them, the processor is connected to the memory through a bus. The memory may include readable memory and random access memory.

In a fourth aspect, the present application provides a computer-readable storage medium that stores instructions that, when run on at least one computing device, cause the at least one computing device to execute the above-mentioned first aspect. Or the method described in any implementation of the first aspect.

In a fifth aspect, the present application provides a computer program product containing instructions that, when run on at least one computing device, cause the at least one computing device to execute the above-mentioned first aspect or any implementation of the first aspect. the method described.

Based on the implementation methods provided in the above aspects, this application can also be further combined to provide more implementation methods.

Description of drawings

In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings needed to be used in the description of the embodiments will be briefly introduced below. Obviously, the drawings in the following description are only some implementations recorded in the present application. For example, those of ordinary skill in the art can also obtain other drawings based on these drawings.

Figure 1 is a schematic diagram of an exemplary application scenario provided by an embodiment of the present application;

Figure 2 is a schematic flow chart of a resource planning method provided by an embodiment of the present application;

Figure 3 is a schematic structural diagram of a computing device provided by an embodiment of the present application;

Figure 4 is a schematic structural diagram of a computing device cluster provided by an embodiment of the present application.

Detailed ways

The solutions in the embodiments provided in this application will be described below with reference to the drawings in this application.

The terms "first", "second", etc. in the description and claims of this application and the above-mentioned drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that the terms so used are interchangeable under appropriate circumstances, and are merely a way of distinguishing objects with the same attributes in describing the embodiments of the present application.

Cloud service providers can provide one or more types of cloud services to users. Each type of cloud service consumes certain resources. The resources consumed by the cloud service may include, for example, computing processing resources, bandwidth resources, storage resources, etc. It can be understood that the resources that a cloud service provider can provide depend on the equipment and other hardware conditions of the cloud service provider, so the total amount of resources that a cloud service provider can provide is relatively fixed.

Therefore, in order to ensure the service quality of each type of cloud service, cloud service providers can plan the resources they can provide and determine the amount of resources that each type of cloud service can use. In this way, after a user subscribes to a certain type of cloud service, a cloud service can be created based on the resources allocated to that type of cloud service, thereby providing the user with the corresponding cloud service.

In order to ensure the rationality of resource planning, currently cloud service providers can discuss how to allocate resources through business meetings. However, manual resource planning has problems such as low efficiency and unreasonable resource planning. In addition, because business meetings often involve relevant personnel of multiple types of cloud services, the frequency of business meetings is relatively low, and resource allocation cannot be adjusted quickly and flexibly.

Based on this, embodiments of the present application provide a resource planning method, which can be executed by a resource planning device to achieve the purpose of determining resources planned for cloud services to ensure the normal operation of cloud services. Specifically, the resource planning device obtains the first feature. The first characteristic is used to indicate the usage of the first resource by the first cloud service in the first historical time period. The first resource is a resource that needs to be consumed during the operation of the first cloud service, and may include, for example, network bandwidth resources, storage space resources, computing resources, and EIP address resources. According to the first feature, the demand of the first resource by the first cloud service in the second time period in the future can be predicted, and the predicted demand of the first resource by the first cloud service in the second time period is called the second feature. According to the second feature, a first allocated resource amount may be determined, and the first allocated resource amount is the resource amount of the first resource allocated for the first cloud service in the second time period.

In this way, the resource planning device can predict the future demand for resources by the cloud service based on the historical demand for resources by the cloud service, and automatically determine how many resources need to be allocated to the cloud service in the second time period in the future. In this way, automatic resource planning can be realized and the rationality of cloud service resource planning can be improved. And, due to the above process It can be completed by the resource planning device without manual participation, which can increase the frequency of cloud service resource planning, thus improving the flexibility of resource planning.

As an example, the above resource planning device can be deployed in the cloud and used to plan resources that can be used by each cloud service in multiple cloud services. For example, in the application scenario shown in FIG. 1 , the resource planning device 110 may be a computing device or a computing device cluster located in the cloud 100 . Moreover, the cloud 100 can provide the first cloud service and the second cloud service to the outside world. For example, the user 210 can subscribe to the first cloud service through the client 310, and the user 220 can subscribe to the second cloud service through the client 320. Both the first cloud service and the second cloud service are implemented based on the resources provided in the cloud 100 .

In actual application, the client 310 may be, for example, an application running on the user-side device corresponding to the first cloud service, or may be an externally provided web browser on the cloud 100 corresponding to the first cloud service, or the like. Similarly, the client 320 may be, for example, an application running on the user-side device corresponding to the second cloud service, or may be an externally provided web browser on the cloud 100 corresponding to the second cloud service, or the like. The resource planning device 100 may include a feature acquisition module 111, a feature prediction module 112, and a resource allocation module 112. Among them, the feature acquisition module 111 is used to acquire the usage of resources by cloud services in the cloud 100 within a historical time period, and send the usage of resources by the cloud services to the feature prediction module 112 as the first feature. The feature prediction module 112 is configured to predict the resource demand of the cloud service in the future time period based on the first feature, and send the resource demand of the cloud service in the future time period as the second feature to the resource allocation module 113 . The resource allocation module 113 is configured to allocate resources to the cloud service according to the second characteristic. For example, if the first characteristic includes the usage of the first resource by the first cloud service in the first historical time period, the resource allocation module 113 may be used to determine the amount of resources allocated to the first cloud service in the second time period in the future. . In this way, during the second time period, the first cloud service can run based on the allocated resources. That is, if the user 210 subscribes to the first cloud service in the second time period, the resources in the cloud 100 can be used to provide the user 210 with the first cloud service, and the cloud 100 is used to provide the first cloud service in the second time period. The total amount of resources for the service does not exceed the amount of resources determined by the resource allocation module 113.

In actual application, the above resource planning device can be implemented by software, or can be implemented by hardware.

As an example of a software functional unit, the resource planning device may include code running on a computing instance. The computing instance may include at least one of a physical host (computing device), a virtual machine, and a container. Furthermore, the above computing instance may be one or more. For example, the resource planning apparatus may include code running on multiple hosts/virtual machines/containers. It should be noted that multiple hosts/virtual machines/containers used to run the code can be distributed in the same region (region) or in different regions. Furthermore, multiple hosts/virtual machines/containers used to run the code can be distributed in the same availability zone (AZ) or in different AZs. Each AZ includes one data center or multiple AZs. geographically close data centers. Among them, usually a region can include multiple AZs.

Likewise, the multiple hosts/VMs/containers used to run the code can be distributed in the same virtual private cloud (VPC), or across multiple VPCs. Among them, usually a VPC is set up in a region. Cross-region communication between two VPCs in the same region and between VPCs in different regions requires a communication gateway in each VPC, and the interconnection between VPCs is realized through the communication gateway. .

As an example of a hardware functional unit, the resource planning device may include at least one computing device, such as a server. Alternatively, the resource planning device may also be implemented using an application-specific integrated circuit (ASIC) or a programmable logic device (PLD). Current equipment, etc. The above-mentioned PLD can be implemented by a complex programmable logical device (CPLD), a field-programmable gate array (FPGA), a general array logic (GAL), or any combination thereof.

Multiple computing devices included in the resource planning device may be distributed in the same region or in different regions. Multiple computing devices included in the resource planning device may be distributed in the same AZ or in different AZs. Similarly, multiple computing devices included in the resource planning apparatus may be distributed in the same VPC or in multiple VPCs. The plurality of computing devices may be any combination of computing devices such as servers, ASICs, PLDs, CPLDs, FPGAs, and GALs.

Next, various non-limiting specific implementations of the resource planning process are described in detail.

Refer to Figure 2, which is a schematic flow chart of a resource planning method in an embodiment of the present application. This method can be applied to the application scenario shown in Figure 1 above, or can also be applied to other applicable application scenarios. The following description takes the application scenario shown in Figure 1 as an example. Moreover, for the functions of each module, please refer to the relevant descriptions of the following embodiments.

The resource planning method shown in Figure 2 may specifically include:

S201: The feature acquisition module 111 acquires the first feature.

The feature acquisition module 111 running on the cloud 100 can acquire the resource usage of one or more cloud services running on the cloud 100. Specifically, for the first cloud service running on the cloud 100, the feature acquisition module 111 can acquire the usage of the first resource by the first cloud service in the first historical time period. In this embodiment, the usage of the first resource by the first cloud service in the first historical time period is called the first characteristic. The first resources are resources required when the first cloud service is running. For example, the first resource may include any one or more of network bandwidth resources, EIP address resources, storage space resources, and computing resources. The computing resources may be, for example, memory resources or central processing unit (CPU) resources.

The length of the first time period can be set according to needs to provide a period for the resource planning module 110 to re-carry out resource planning. That is to say, the feature acquisition module 111 can periodically acquire the usage of the first resource by the first cloud service within the period, using the duration of the first time period as the period duration. In specific implementation, the first time period may be 24 hours. That is, the resource planning device can obtain the use of the first resource by the first cloud service on the nth day at the end of the nth day, and determine the use of the first resource by the first cloud service on the nth day based on the use of the first resource by the first cloud service on the nth day. + How many resources need to be allocated to the first cloud service for 1 day. where n is a positive integer.

In specific implementation, the first resource may be a resource that has a greater impact on the first cloud service. If fewer first resources are allocated to the first cloud service, the scale of the first cloud service provided by the cloud 100 may be smaller, or the service quality of the first cloud service provided by the cloud 100 may be poor. That is, if the first cloud service runs based on the first resource and the second resource, the impact of the amount of the first resource allocated to the first cloud service on the first cloud service is greater than the impact of the amount of the first resource allocated to the first cloud service. The impact of the resource amount of the second resource on the first cloud service.

The following uses specific examples to illustrate the corresponding relationship between the first resource and the first cloud service.

In the first example, the first cloud service is an object storage service (OBS). When the user 210 uses the object storage cloud service, the user 210 can upload the data to be stored to the cloud 100 through the network, and the data will be stored by the cloud. The storage device in the terminal 100 stores. In this process, the main factor limiting the service quality of the object storage service is the uplink bandwidth resources allocated to the object storage service. The greater the upstream bandwidth allocated to the object storage service, the faster the user 210 can upload data to be stored. Therefore, the first resource corresponding to the object storage service may be an uplink bandwidth resource.

In the second example, the first cloud service is a cloud live broadcast service. When user 210 uses the cloud live broadcast service, user 210 can send live broadcast data to cloud 100, and cloud 100 can send the live broadcast data to users watching the live broadcast. In this process, the main factor limiting the service quality of the cloud live broadcast service is the downlink bandwidth resources allocated to the cloud live broadcast service. The greater the downlink bandwidth allocated to the cloud live broadcast service, the better the quality of live broadcast data transmission, and the smoother the live broadcast will be seen by users watching the live broadcast. Therefore, the first resource corresponding to the cloud live broadcast service may be the downlink bandwidth resource.

In the third example, the first cloud service is a cloud computing service. When user 210 uses cloud computing services, user 210 can send data to cloud 100, and cloud 100 can call the computing resources of cloud 100 to calculate the data, and send the calculated data to user 210. During this process. The main factor that limits the service quality of a cloud computing service is the computing resources allocated to the cloud computing service. The more computing resources allocated to cloud computing services, the faster the data calculation process will be. Therefore, the first resource corresponding to the cloud computing service may be a computing resource.

It will be appreciated that the above three examples are given for ease of understanding. In actual application, the cloud 100 may not provide the cloud services in the above three examples, or may provide other cloud services. Moreover, the corresponding relationship between the first cloud service and the first resource can be flexibly adjusted according to the actual situation.

According to the above introduction, the first characteristic is used to indicate the usage of the first resource by the first cloud service in the first time period. In actual application, the first characteristic may specifically be the average usage of the first resource by the first cloud service in the first time period, or it may be the maximum usage of the first resource by the first cloud service in the first time period. .

Or, in some possible implementations, in order to more flexibly adjust the resource amount of the first resource allocated to the first cloud service, the first characteristic can be used to indicate the pair of the first cloud service in multiple sub-time periods in the first time period. The usage of primary resources. That is to say, the first time period may be divided into multiple sub-time periods, and the first characteristic includes the usage of the first resource by the first cloud service in each sub-time period of the multiple sub-time periods. For this situation, the first characteristic is also used to indicate a change pattern of the resource amount of the first resource used by the first cloud service in the first time period.

For example, if the first time period is the usage of the first resource by the first cloud service in the nth day, then the first characteristic may be used to indicate the usage of the first resource by the first cloud service every hour (or every minute) in the nth day. Resource usage. That is to say, when providing the first cloud service to the user, the feature acquisition module 111 or other modules in the cloud 100 can store the usage of the first resource by the first cloud service every hour (or every minute). It can be understood that the feature acquisition module 111 or other modules in the cloud 100 can also store the usage of the first resource by the first cloud service at other frequencies.

It should be noted that the first cloud service in the embodiment shown in FIG. 2 can be executed by any one or more computing devices in the cloud 100 . When acquiring the first feature, the feature acquisition module 111 may respectively acquire the resources used by the first cloud service on the computing device from each computing device running the first cloud service in the cloud 100 .

After obtaining the first feature, the feature acquisition module 111 may send the first feature to the feature prediction module 112 .

S202: The feature prediction module 112 predicts the second feature according to the first feature.

After receiving the first feature sent by the feature acquisition module 111, the feature prediction module 112 may predict the second feature based on the first feature. The second characteristic is used to indicate the demand of the first resource by the first cloud service in the second time period in the future. That is to say, after the pattern of using the first resource by the first cloud service is obtained through the feature acquisition module 111, the resource The source planning device 110 can predict future demand for the first resource by the first cloud service through the feature prediction module 112, thereby determining the amount of resources allocated to the first cloud service based on the demand for the first resource by the first cloud service.

The method of determining the second feature based on the first feature will be described below with a specific example.

In a first example, the feature prediction module 112 may determine the second feature based on the overall pattern of usage of the first resource by the first cloud service in the first time period.

For example, if the first characteristic is the average or maximum value of the first resource used by the first cloud service in the first time period. Correspondingly, the feature prediction module 112 may determine the above-mentioned average value or maximum value as the demand for the first resource by the first cloud service in the second time period. That is, the demand for the first resource by the first cloud service in the second time period indicated by the second characteristic is equal to the usage of the first resource by the first cloud service during the first time period indicated by the first characteristic. match.

In a second example, the feature prediction module 112 may perform a fine-grained analysis on the usage of the first resource by the first cloud service in the first time period, and determine the second feature based on the analysis results.

For example, if the first feature is used to indicate the change pattern of the usage of the first resource by the first cloud service within the first time period, then the feature prediction module 112 may be based on the change in usage of the first resource by the first cloud service. According to the law, the possible change law of the usage of the first resource of the first cloud service in the future is predicted, and the second characteristic is determined accordingly.

In some possible implementations, the duration of the first time period may be greater than the duration of the second time period. Assume that the duration of the first time period is the first duration, and the duration of the second time period is the second duration. Then when determining the second feature, the feature prediction module can first determine the sub-time period in the first time period corresponding to the second time period based on the time mapping relationship, and calculate the first resource in this sub-time period according to the first cloud service. The change pattern of usage determines the second characteristic.

Wherein, the time mapping relationship is used to indicate the corresponding relationship between time periods in which the change patterns of the usage of the first resource are similar. That is, the feature prediction module 112 can select the change rule corresponding to the second time period from the predicted change rules according to the time mapping relationship.

For example, if the first characteristic indicates the usage of the first resource by the first cloud service within 24 consecutive hours on the nth day, the second time period is the time period between 19:00-20:00 on the n+1th day. , n is a positive integer. Then when determining the second feature, the feature prediction module 112 can first determine from the first feature based on the second time period the first cloud service's response to the first resource during the period from 19:00 to 20:00 in the evening of the nth day. The changing pattern of usage. Then, the feature prediction module 112 predicts the first cloud service's usage of the first resource during the period of 19:00-20:00 on the nth day based on the change pattern of the first cloud service on the n+1th night. The second feature is obtained from the possible change pattern of the usage of the first resource during the period from 19:00 to 20:00.

In some possible implementations, if the first characteristic indicates the usage of the first resource by the first cloud service within 24 consecutive hours on the nth day, then the characteristic prediction module 112 may determine the first characteristic per hour on the n+1th day. The demand for primary resources by cloud services. That is to say, the feature prediction module 112 may also determine a third feature based on the first feature, and the third feature indicates the demand for the first resource by the first cloud service in a third time period in the future. The third time period and the second time period respectively correspond to the sub-time periods in the first time period. For example, in the above example, the third time period may be the time period between 18:00 and 19:00 in the evening of the n+1th day.

In the above introduction, the feature prediction module 112 can predict the demand for the first resource by the first service in multiple time periods in the future based on the first feature. However, if the allocated resource amount corresponding to each time period is calculated in subsequent step S203, and the first resource allocated to the first cloud service is allocated according to the allocated resource amount, it may cause the cloud 100 to allocate the first resource to the first cloud service. Frequent changes in the resource amount of the first resource may affect the resource allocation of the cloud 100 and affect the normal operation of the first cloud service and/or other cloud services.

To this end, in some possible implementations, multiple adjacent time periods may be merged into one time period, thereby reducing the frequency of changes in the resource amount of the first resource of the first cloud service. Specifically, it can be determined whether two adjacent time periods can be merged based on the similarity of the demand for the first resource for the first cloud service in the time period.

For example, if the above-mentioned second time period and the third time period are adjacent, then the feature prediction module 112 can determine whether the first cloud service's demand for the first resource in the second time period is consistent with the first cloud service's demand for the first resource in the third time period. Whether the difference in demand for the first resource is greater than a preset threshold. If the difference between the first cloud service's demand for the first resource in the second time period and the first cloud service's demand for the first resource in the third time period is greater than the preset threshold, it means that the first cloud service is in The demand for the first resource is quite different between the second time period and the third time period, and the second time period and the third time period cannot be combined. If the difference between the first cloud service's demand for the first resource in the second time period and the first cloud service's demand for the first resource in the third time period is not greater than the preset threshold, it means that the first cloud service If the difference in demand for the first resource between the second time period and the third time period is small, the feature prediction module 112 may merge the second time period and the third time period.

When merging the second time period and the third time period, the feature prediction module 112 may use the greater demand for the first resource in the second time period and the third time period as the first cloud in the combined time period. The demand for the primary resource by the service.

for example. It is assumed that the feature prediction module 112 determines the demand for downlink bandwidth of the cloud live broadcast service on the n+1th day based on the first feature. Among them, the demand for downlink bandwidth for the cloud live broadcast service from 0:00-4:00 on the n+1st day is three gigabytes per second (GB/s). The demand for downlink bandwidth from 4:00-8:00 on day n+1 is 3.1GB/s, and the cloud live broadcast service’s demand for downlink bandwidth from 8:00-12:00 on day n+1 is 5GB/s.

The difference between the demand for downlink bandwidth of the cloud live broadcast service from 0:00-4:00 on the n+1 day and the demand for downlink bandwidth by the cloud live broadcast service from 4:00-8:00 on the n+1 day is 0.1GB/s, which is less than the preset threshold 1GB/s, then the feature prediction module 112 can determine the demand for downlink bandwidth of the cloud live broadcast service from 0:00-8:00 on the n+1th day to be 3.1GB/s. The difference between the cloud live broadcast service’s demand for downlink bandwidth from 4:00-8:00 on day n+1 and the cloud live broadcast service’s demand for downlink bandwidth from 8:00-12:00 on day n+1 is 1.9GB/s, which is greater than the preset threshold 1GB/s, then the feature prediction module 112 determines that the time period of 4:00-8:00 on the n+1th day is the same as the time period of 8:00-12:00 on the n+1st day. Time periods cannot be merged.

Optionally, in some possible implementations, the above action of merging time periods may also be performed by the resource allocation unit 112.

After determining the demand for the first resource by the first cloud service within the first time period, the feature prediction module 112 may send the second feature to the resource allocation module 113, so that the resource allocation module 113 provides the first resource for the first cloud service based on the second feature. Cloud services allocate the first resources. In the same way, the feature prediction module 112 can also send the third feature to the resource allocation module 113.

S203: The resource allocation module 113 determines the first allocated resource amount according to the second characteristic.

After receiving the second feature sent by the feature prediction module 112, the resource allocation module 113 may determine the first allocated resource amount according to the second feature. The first allocated resource amount is the resource amount of the first resource allocated to the first cloud service in the second time period. That is to say, during the second time period, the first cloud service uses at most the first resource of the first allocated resource amount in the cloud 100 .

Similarly, if the feature prediction module 112 can also determine the second allocated resource amount based on the third feature. Second allocation of funds The source amount is the resource amount of the first resource allocated to the first cloud service in the third time period. I won’t go into details here.

According to the foregoing introduction, it can be known that the second characteristic is used to indicate the demand of the first resource by the first cloud service in the second time period. Specifically, if the second characteristic is used to indicate the demand for the first resource by the first cloud service within the second time period, the resource allocation module 113 may use the demand as the first allocated resource. Alternatively, the resource allocation module 113 may use the sum of the demand amount and the preset redundancy amount as the first allocated resource amount. In this way, even if the user's subscription to the first cloud service increases during the second time period, the first allocated resource amount determined based on the second characteristic and the preset redundancy amount may allow the cloud 100 to provide more first cloud services.

In some possible implementations, the resource allocation module 113 may determine characteristics of the first allocated resource amount based on characteristics of the first cloud service.

For example, if the first resource is a network bandwidth resource, the first allocated resource amount is used to indicate the network bandwidth that can be used by the first cloud service. The current cloud can often provide multiple different lines, each of which has certain network bandwidth resources. In this way, when determining the first allocated resource amount, the resource allocation module 113 can use multiple lines provided by the cloud 100 as alternative lines. The resource allocation module 113 may determine the target bandwidth amount required by the first cloud service in the second time period according to the second characteristic. Next, the resource allocation module 113 determines the alternative lines used by the first cloud service in the second time period based on the target bandwidth amount, the first cloud service's demand for lines, and the network bandwidth resources of each alternative line. Among them, the alternative line selected by the resource allocation module 113 may also be called a target line. In this way, the first cloud service can use the network bandwidth resources of the target line during the second time period, and the maximum bandwidth used by the first cloud service does not exceed the target bandwidth.

Among the multiple lines provided by the cloud 100, there may be lines with better network performance (such as parameters such as delay, packet loss, and jitter), and there may also be lines with lower bandwidth costs. Lines with good network performance tend to be more expensive. Correspondingly, when selecting a line, the resource allocation module 113 may make a selection based on the characteristics of the first cloud service.

The following uses two examples to introduce the method of selecting lines by the resource allocation module.

In the first example, the first cloud service is a cost-sensitive cloud service, that is, the first cloud service has higher cost requirements. Then, in order to reduce the cost of the first cloud service, the resource allocation module 113 can select the candidate line with the lowest cost and the remaining bandwidth amount greater than the target bandwidth amount as the target line from multiple alternative lines.

for example. It is assumed that the first cloud service is a cloud storage service. Since cloud storage services often use data backup mechanisms when transmitting data, cloud storage services do not have high requirements for network quality. Moreover, the amount of data uploaded to the cloud in cloud storage services is often large. If high-cost lines are used to provide network bandwidth resources for cloud storage services, the cost of cloud storage services will increase significantly. Therefore, cloud storage services can be considered as cost-sensitive cloud services.

Correspondingly, when determining the first allocated resource amount of the cloud storage service, the resource allocation module 113 may first determine the target bandwidth amount required by the cloud storage service within the second time period. Then, the resource allocation module 113 can select the candidate line whose remaining bandwidth resources are not less than the target bandwidth amount and has the lowest cost from multiple candidate lines as the target line. In this way, the obtained first allocated resource amount is used to instruct the cloud storage service to use the network bandwidth resources of the target line within the second time period, and the network bandwidth resources that the cloud storage service can use do not exceed the target bandwidth amount.

In the second example, the first cloud service is a performance-sensitive cloud service, that is, the first cloud service has higher requirements on network performance. Then, in order to improve the service quality of the first cloud service, the resource allocation module 113 selects the candidate line with the best network performance and the remaining bandwidth that is greater than the target bandwidth from multiple candidate lines as the target line.

for example. Assume that the first cloud service is a cloud live broadcast service. Since users who watch live broadcasts have higher requirements for live broadcast quality, cloud live broadcast services often have higher requirements for network parameters such as delay, packet loss, and jitter. Therefore, cloud storage services have higher requirements for network quality. Therefore, the cloud live broadcast service can be regarded as a performance-sensitive cloud service.

Correspondingly, when determining the first allocated resource amount of the cloud live broadcast service, the resource allocation module 113 may first determine the target bandwidth amount required by the cloud live broadcast service in the second time period. Then, the resource allocation module 113 can select an alternative line whose remaining bandwidth resources are not less than the target bandwidth amount and has the best network performance from multiple alternative lines as the target line. In this way, the obtained first allocated resource amount is used to instruct the cloud storage service to use the network bandwidth resources of the target line within the second time period, and the network bandwidth resources that the cloud storage service can use do not exceed the target bandwidth amount.

In some possible implementations, cloud 100 may include multiple clusters of computing devices located in different geographical locations. The multiple alternative routes to cloud 100 may then be routes to different clusters of computing devices. In this way, for the cloud storage service, considering that users using the cloud storage service may not care which computing device cluster the data is stored in, the resource allocation module 113 may use any alternative line as a target line.

However, for cloud live broadcast services, the audience of cloud live broadcast services may be concentrated in specific areas. If the computing device cluster corresponding to the target line belongs to a different geographical area from the audience of the cloud live broadcast service, then cross-regional transmission may affect the service quality of the cloud live broadcast service. To this end, in some possible implementations, the resource allocation module 113 may first select an alternative line that matches the cloud live broadcast service from the alternative lines provided by the cloud 100, and then determine the target line based on network performance parameters.

After determining the first allocated resource amount, the resource allocation module 113 may send the first allocated resource amount to the network device in the cloud 100 for running the first cloud service, or send the first allocated resource amount to the management device in the cloud 100 . In this way, within the second time period, the cloud 100 can be controlled to run the first service with first resources that do not exceed the first allocated resource amount.

In this embodiment, the resource planning module 110 can predict the demand of the first resource by the first cloud service in the second time period based on the usage of the first resource by the first cloud service in the first time period, and then predict According to the demand of the first cloud service for the first resource, the resource amount of the first resource that the first cloud service can use in the second time period is determined. In this way, on the one hand, there is no need to manually hold a meeting to determine the first resource allocated to the first cloud service; on the other hand, prediction is made based on the historical usage of the first resource, and the obtained second characteristics and the first allocated resource amount can satisfy the needs of the first cloud service usage needs, resource planning is more reasonable. In addition, the first resources allocated to the first cloud service can be quickly and flexibly adjusted. Compared with the traditional manual resource planning method, the frequency of resource re-planning can be increased and the changing rules of the cloud service business can be more flexibly adapted.

In the application scenario shown in Figure 1, the cloud 100 can also provide the user 220 with a second cloud service. Correspondingly, the feature acquisition module 111 may also acquire the usage of the third resource by the second cloud service in the first historical time period. The third resource and the first resource may be the same resource or different resources. Optionally, the third resource is the resource that has the highest impact on the second cloud service. The introduction to resources can be found above and will not be repeated here.

In the above embodiment shown in Figure 2, the resource planning device (including the above-mentioned feature acquisition module 111, feature prediction module 112 and resource planning module 113) involved in the resource planning process for cloud services may be configured on a computing device or a computing device. software on a cluster of devices and, by running the software on a computing device or a cluster of computing devices, So that the computing device or the computing device cluster realizes the functions of the above resource planning device. Next, based on the perspective of hardware device implementation, the resource planning devices involved in the resource planning process are introduced in detail.

Figure 3 shows a schematic structural diagram of a computing device. The above-mentioned resource planning device 110 can be deployed on the computing device. The computing device can be a computing device (such as a server) in a cloud environment, or a computing device in an edge environment. Or terminal equipment, etc. can be specifically used to implement the functions of the feature acquisition module 111, the feature prediction module 112 and the resource planning module 113 in the embodiment shown in Figure 2.

As shown in FIG. 3 , computing device 300 includes processor 310 , memory 320 , communication interface 330 , and bus 340 . The processor 310, the memory 320 and the communication interface 330 communicate through the bus 340. The bus 340 may be a peripheral component interconnect (PCI) bus or an extended industry standard architecture (EISA) bus, etc. The bus can be divided into address bus, data bus, control bus, etc. For ease of presentation, only one thick line is used in Figure 3, but it does not mean that there is only one bus or one type of bus. The communication interface 330 is used to communicate with the outside, such as receiving requests from users to subscribe to cloud services.

The processor 310 may be a central processing unit (CPU), an application specific integrated circuit (ASIC), a graphics processing unit (GPU), or one or more integrated circuits. The processor 310 may also be an integrated circuit chip with signal processing capabilities. During the implementation process, the functions of each unit in the resource planning device can be completed by instructions in the form of hardware integrated logic circuits or software in the processor 310 . The processor 310 can also be a general-purpose processor, a digital signal processor (DSP), a field programmable gate array (FPGA) or other programmable logic devices, discrete gates or transistor logic devices, Discrete hardware components can implement or execute the methods, steps and logical block diagrams disclosed in the embodiments of this application. Among them, the general processor can be a microprocessor or the processor can be any conventional processor, etc. The method disclosed in combination with the embodiments of the present application can be directly implemented as a hardware decoding processor to complete the execution, or can be performed using decoding processing. The combination of hardware and software modules in the device is executed. 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 320. The processor 310 reads the information in the memory 320 and completes some or all functions of the resource planning device in combination with its hardware.

The memory 320 may include volatile memory, such as random access memory (RAM). The memory 320 may also include non-volatile memory (non-volatile memory), such as read-only memory (ROM), flash memory, HDD or SSD.

The memory 320 stores executable code, and the processor 310 executes the executable code to perform the method executed by the resource planning device.

Specifically, when the embodiment shown in Figure 2 is implemented, and the feature acquisition module 111, feature prediction module 112 and resource planning module 113 described in the embodiment shown in Figure 2 are implemented by software, execute Figure 2 The software or program code required for the functions of the feature acquisition module 111, feature prediction module 112 and resource planning module 113 in 2 is stored in the memory 320. The interaction between the resource planning device 110 and other devices is realized through the communication interface 330, and the processor is used to execute The instructions in the memory 320 implement the method executed by the resource planning device.

Figure 4 shows a schematic structural diagram of a computing device cluster. The computing device cluster 40 shown in Figure 4 includes multiple computing devices, and the above resource planning device can be deployed in a distributed manner on the multiple computing devices in the computing device cluster 40. Prepare. As shown in Figure 4, the computing device cluster 40 includes multiple computing devices 400. Each computing device 400 includes a memory 420, a processor 410, a communication interface 430, and a bus 440. The memory 420, the processor 410, and the communication interface 430 pass through Bus 440 implements communication connections between each other.

Processor 410 may employ a CPU, GPU, ASIC, or one or more integrated circuits. The processor 410 may also be an integrated circuit chip with signal processing capabilities. During the implementation process, part of the functions of the resource planning device may be implemented by instructions in the form of integrated logic circuits or software in the hardware of the processor 410 . The processor 410 can also be a DSP, FPGA, general-purpose processor, other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, and can implement or execute some of the methods, steps and logic block diagrams disclosed in the embodiments of this application. The general processor may be a microprocessor or the processor may be any conventional processor, etc. The steps of the method disclosed in conjunction with the embodiments of the present application may be directly implemented as a hardware decoding processor, or may be executed using a decoding processor. The combination of hardware and software modules in the code processor is executed. 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 420. In each computing device 400, the processor 410 reads the information in the memory 40 and, combined with its hardware, can complete some functions of resource planning and configuration.

Memory 420 may include ROM, RAM, static storage devices, dynamic storage devices, hard disks (eg, SSD, HDD), etc. The memory 420 may store program codes, for example, part or all of the program code used to implement the feature acquisition module 111 , part or all of the program code used to implement the feature prediction module 112 , part or all of the program code used to implement the resource planning module 113 wait. For each computing device 400, when the program code stored in the memory 420 is executed by the processor 410, the processor 410 executes part of the method executed by the resource planning device based on the communication interface 430. For example, some of the computing devices 400 may be used to execute the above. The method executed by the feature acquisition module 111 and the feature prediction module 112, and another part of the computing device 400 is used to execute the method executed by the resource planning module 113. The memory 420 may also store data, such as intermediate data or result data generated by the processor 410 during execution, such as the above-mentioned first features and second features, etc.

The communication interface 403 in each computing device 400 is used to communicate with the outside, such as interacting with other computing devices 400 and so on.

The bus 440 may be a peripheral component interconnection standard bus or an extended industry standard architecture bus, or the like. For ease of presentation, the bus 440 within each computing device 400 in FIG. 4 is represented by only one thick line, but this does not mean that there is only one bus or one type of bus.

Communication paths are established between the above-mentioned plurality of computing devices 400 through a communication network to implement the function of the resource planning device. Any computing device may be a computing device (eg, a server) in a cloud environment, a computing device in an edge environment, or a terminal device.

In addition, embodiments of the present application also provide a computer-readable storage medium that stores instructions that, when run on one or more computing devices, cause the one or more computing devices to execute The method executed by each unit of the resource planning device in the above embodiment.

In addition, embodiments of the present application also provide a computer program product. When the computer program product is executed by one or more computing devices, the one or more computing devices execute any one of the foregoing resource planning methods. The computer program product may be a software installation package. If any of the foregoing resource planning methods is required, the computer program product may be downloaded and executed on the computer.

In addition, it should be noted that the device embodiments described above are only illustrative. The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physically separate. The physical unit can be located in one place, or it can be distributed across multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment. In addition, in the drawings of the device embodiments provided in this application, the connection relationship between units indicates that there are communication connections between them, which can be specifically implemented as one or more communication buses or signal lines.

Through the above description of the embodiments, those skilled in the art can clearly understand that the present application can be implemented by software plus necessary general hardware. Of course, it can also be implemented by dedicated hardware including dedicated integrated circuits, dedicated CPUs, dedicated memories, Special components, etc. to achieve. In general, all functions performed by computer programs can be easily implemented with corresponding hardware. Moreover, the specific hardware structures used to implement the same function can also be diverse, such as analog circuits, digital circuits or special-purpose circuits. circuit etc. However, for this application, software program implementation is a better implementation in most cases. Based on this understanding, the technical solution of the present application can be embodied in the form of a software product in essence or that contributes to the existing technology. The computer software product is stored in a readable storage medium, such as a computer floppy disk. , U disk, mobile hard disk, ROM, RAM, magnetic disk or optical disk, etc., including several instructions to cause a computer device (which can be a personal computer, training device, or network device, etc.) to execute the steps described in various embodiments of this application. method.

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 the embodiments of the present application are generated in whole or in part. The computer may 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, for example, the computer instructions may be transferred from a website, computer, training device, or data The center transmits to another website site, computer, training equipment or data center through wired (such as coaxial cable, optical fiber, digital subscriber line (DSL)) or wireless (such as infrared, wireless, microwave, etc.) means. The computer-readable storage medium may be any available medium that a computer can store, or a data storage device such as a training device or a data center integrated with one or more available media. 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 (Solid State Disk, SSD)), etc.

Claims (17)

1. A resource planning method, characterized in that the method includes:

   Obtain the first characteristic, which indicates the usage of the first resource by the first cloud service in the first historical time period;

   Determine a second characteristic based on the first characteristic, the second characteristic indicating the demand for the first resource by the first cloud service in a second time period in the future;

   The first allocated resource amount is determined according to the second characteristic, and the first allocated resource amount is the resource amount of the first resource allocated to the first cloud service in the second time period.
2. The method of claim 1, further comprising:

   Determine a third characteristic based on the first characteristic, the third characteristic indicating the demand for the first resource by the first cloud service in a third time period in the future;

   A second allocated resource amount is determined according to the third characteristic, and the second allocated resource amount is the resource amount of the first resource allocated to the first cloud service in the third time period.
3. The method according to claim 2, characterized in that, before determining the second allocated resource amount, the method further includes:

   Determine the difference between the demand of the first cloud service for the first resource in the second time period and the demand of the first resource by the first cloud service in the third time period. The difference between them is greater than the preset threshold.
4. The method according to any one of claims 1 to 3, characterized in that the first cloud service runs based on a first resource and a second resource, and the resource amount of the first resource has a significant impact on the first cloud service. The impact of the service is greater than the impact of the resource amount of the second resource on the first cloud service.
5. The method according to any one of claims 1-4, characterized in that the first resource includes any one or more of the following:

   Network bandwidth resources, elastic Internet protocol EIP address resources, storage space resources and computing resources.
6. The method according to claim 5, wherein the first resource includes a network bandwidth resource, and determining the first allocated resource amount according to the second characteristic includes:

   Determine the target bandwidth amount according to the demand of the first cloud service for the network bandwidth resource in the second time period;

   Determine the target line according to the cloud service characteristics of the first cloud service;

   The first allocated resource amount is determined according to the target line and the target bandwidth amount, and the first allocated resource amount indicates that the target line allocates the first cloud service to the first cloud service within the second time period. The target bandwidth amount of network bandwidth resources.
7. The method of claim 6, wherein the cloud service characteristics of the first cloud service indicate that the first cloud service is a cost-sensitive cloud service, and the cloud service characteristics according to the first cloud service Determining target routes includes:

   An alternative line with the lowest cost and a remaining bandwidth greater than the target bandwidth is selected from multiple alternative lines as the target line.
8. A resource planning device, characterized in that the device includes:

   A feature acquisition module, configured to acquire a first feature, which indicates that the first cloud service has The use of primary resources;

   A feature prediction module, configured to determine a second feature based on the first feature, the second feature indicating the demand for the first resource by the first cloud service in a second time period in the future;

   A resource allocation module, configured to determine a first allocated resource amount according to the second characteristic, where the first allocated resource amount is the resource amount of the first resource allocated to the first cloud service in the second time period.
9. The device according to claim 8, characterized in that:

   The feature prediction module is further configured to determine a third feature based on the first feature, the third feature indicating the demand for the first resource by the first cloud service in a third time period in the future;

   The resource allocation module is further configured to determine a second allocated resource amount based on the third characteristic, where the second allocated resource amount is the first resource allocated for the first cloud service in the third time period. Resources.
10. The device according to claim 9, wherein the first cloud service's demand for the first resource in the second time period is different from the first cloud service's demand for the first resource in the third time period. The difference between the demand conditions for the first resource within the segment is greater than a preset threshold.
11. The device according to any one of claims 8-10, characterized in that the first cloud service runs based on a first resource and a second resource, and the resource amount of the first resource has a significant impact on the first cloud service. The impact of the service is greater than the impact of the resource amount of the second resource on the first cloud service.
12. The device according to any one of claims 8-11, characterized in that the first resource includes any one or more of the following:

    Network bandwidth resources, elastic Internet protocol EIP address resources, storage space resources and computing resources.
13. The device according to claim 12, wherein the first resource includes a network bandwidth resource,

    The resource allocation module is specifically configured to determine the target bandwidth amount according to the demand of the network bandwidth resource by the first cloud service in the second time period; determine the target bandwidth amount according to the cloud service characteristics of the first cloud service. Target line; determine the first allocated resource amount according to the target line and the target bandwidth amount, the first allocated resource amount indicates that the target line serves the first cloud service within the second time period Allocate network bandwidth resources of the target bandwidth amount.
14. The device according to claim 13, wherein the cloud service characteristics of the first cloud service indicate that the first cloud service is a cost-sensitive cloud service,

    The resource allocation module is specifically configured to select an alternative line with the lowest cost and a remaining bandwidth amount greater than the target bandwidth amount from multiple alternative lines as the target line.
15. A computing device cluster, characterized in that the computing device cluster includes at least one computing device, and each computing device includes a processor and a memory:

    The memory is used to store instructions;

    The processor is configured to cause the computing device cluster to execute the method according to any one of claims 1-7 according to the instructions.
16. A computer-readable storage medium, characterized in that instructions are stored in the computer-readable storage medium, and when run on a computing device, the computing device causes the computing device to execute as described in any one of claims 1-7 Methods.
17. A computer program product containing instructions that, when run on a computing device, cause the computing device to perform the method of any one of claims 1-7.