Computing power scheduling method and device and readable storage medium

Abstract

The invention provides a computing power scheduling method and device and a readable storage medium, and relates to the technical field of communication, and the method comprises the steps that the computing power scheduling device receives a computing power request from a computing power demand end, the computing power request is used for requesting to select a computing power node for a service, and the computing power request comprises a first field, the first field is used for indicating an application scene for selecting the computing power node. The computing power scheduling device determines at least one computing power node in a computing power node set according to node selection parameters corresponding to the application scene, and the at least one computing power node is a computing power node supporting providing computing power service for the business. Wherein the node selection parameters comprise one or more of computing power node availability parameters, service end-to-end availability and end-to-end computing power demand parameters of the computing power demand end. The computing power service quality can be improved, and reasonable allocation of computing power resources is realized.

Description

Technical Field

[0001] The present application relates to the field of communication technology, and in particular to a computing power scheduling method, device and readable storage medium. Background Art

[0002] With the vigorous development of the digital economy, the Internet, big data, and artificial intelligence are deeply integrated with all walks of life, and the amount of data and computing generated is exponentially exploding. The processing and analysis of the current massive data requires strong computing power to provide support. The computing power network refers to a new paradigm that effectively allocates basic resources such as computing and storage between cloud-edge-end on the basis of the continuous ubiquitous development of computing power, and collaboratively provides digital technology and application integration services. The computing power network can call ubiquitously distributed computing resources (i.e. computing power resources), obtain dynamic change information of computing resources, and combine network resource information and other factors to select suitable computing power nodes that provide computing power services for the user end.

[0003] However, in the current computing power scheduling, the computing power nodes are mainly selected based on factors such as network latency, cost, and computing power performance. It is possible that the selected computing power nodes cannot meet the needs of the user end, resulting in unreasonable computing power scheduling. Summary of the invention

[0004] The present application provides a computing power scheduling method, device and readable storage medium to solve the problem of unreasonable computing power scheduling existing in the current computing power resource scheduling.

[0005] In a first aspect, a computing power scheduling method is provided, comprising: receiving a computing power request from a computing power demand side, the computing power request being used to request selection of a computing power node for a service, the computing power request comprising a first field, the first field being used to indicate an application scenario for selecting a computing power node; determining at least one computing power node in a computing power node set according to a node selection parameter corresponding to the application scenario, the at least one computing power node being a computing power node that supports providing computing power services for the service, wherein the node selection parameter comprises one or more of a computing power node availability parameter, a service end-to-end parameter, and an end-to-end computing power demand parameter of the computing power demand side.

[0006] Optionally, the computing power node availability parameter is used to characterize the failure rate and / or security of the computing power node, and the computing power node availability parameter includes a node availability parameter and a network availability parameter of the computing power node.

[0007] Optionally, the end-to-end computing power requirement parameter includes one or more of the following:

[0008] The end-to-end latency parameters of the service expected by the computing power demander;

[0009] The expected computing power performance parameters of the computing power demand side;

[0010] The end-to-end availability parameters of the service expected by the computing power demander;

[0011] The computing power scheduling cost parameter expected by the computing power demand side.

[0012] Optionally, if the first field is a first value, the application scenario is a first application scenario, and the first application scenario is to select a computing power node that supports the computing power service of the service for deployment of the service, or in other words, the first application scenario is to select a computing power node that can provide the computing power service of the service for deployment of the service. Alternatively, if the first field is a second value, the application scenario is a second application scenario, and the second application scenario is to select a computing power node from the computing power nodes that have deployed the service to provide the computing power service of the service.

[0013] Optionally, the application scenario is the first application scenario, and the method further includes: deploying the service to the at least one computing power node.

[0014] Optionally, the node selection parameter includes the end-to-end computing power requirement parameter, and the computing power request includes a second field, and the second field is used to indicate the end-to-end computing power requirement parameter.

[0015] Optionally, the end-to-end service parameters include end-to-end service delay parameters and / or scheduling costs of computing power nodes, and / or the node selection parameters also include available computing power performance parameters.

[0016] Optionally, the application scenario is the second application scenario, and the method further includes: scheduling the at least one computing power node to provide computing power services for the business.

[0017] Optionally, the computing power node set is a set of computing power nodes on which the business has been deployed.

[0018] Optionally, the service end-to-end parameters include one or more of the service end-to-end delay parameters of the computing power node and / or the service average end-to-end delay parameters of the computing power node set; and / or, the node selection parameters also include the service availability parameters of the computing power node, the available computing power performance parameters of the computing power node and / or the average computing power performance parameters of the computing power node set.

[0019] Optionally, the determining of at least one computing power node in the computing power node set according to the node selection parameters corresponding to the application scenario includes: determining the comprehensive service index of each computing power node in the computing power node set according to the node selection parameters corresponding to the application scenario; determining the at least one node according to the comprehensive service index of the computing power nodes in the computing power node set, wherein the comprehensive service index is obtained by a weighted combination of multiple indicators, the multiple indicators include an availability index, the availability index is determined according to the availability parameters of the computing power node, and the weight corresponding to the availability index is greater than the weight corresponding to the indicators among the multiple performance indicators except the availability index.

[0020] Optionally, when the application scenario is the first application scenario, the multiple performance indicators also include a latency indicator, a computing power performance indicator and a scheduling cost indicator corresponding to the first application scenario; or, when the application scenario is the second application scenario, the multiple performance indicators also include a latency indicator and a computing power performance indicator corresponding to the second application scenario.

[0021] In a second aspect, a computing power scheduling device is provided, comprising a unit / module for implementing the computing power scheduling method provided in the first aspect and any one of the optional implementation modes of the first aspect. Exemplarily, the computing power scheduling device comprises a transceiver module and a processing module. The transceiver module is used to receive a computing power request from a computing power demand side, the computing power request is used to request the selection of a computing power node for a service, the computing power request comprises a first field, the first field is used to indicate an application scenario for selecting a computing power node; the processing module is used to determine at least one computing power node in a computing power node set according to a node selection parameter corresponding to the application scenario, the computing power node is a computing power node selected to provide computing power services for the service,

[0022] Among them, the node selection parameters include one or more of the computing power node availability parameters, the service end-to-end parameters and the end-to-end computing power demand parameters of the computing power demand end.

[0023] In a third aspect, a computing power scheduling device is provided, comprising a processor, wherein the processor is used to execute the computing power scheduling method provided in the first aspect and any optional implementation of the first aspect.

[0024] The device may further include a memory for storing instructions and data. The memory is coupled to the processor, and when the processor executes the instructions stored in the memory, the method provided in the first aspect may be implemented. The device may further include a communication interface, and the communication interface is used for the device to communicate with other devices. Exemplarily, the communication interface may be a transceiver, circuit, bus, module or other type of communication interface.

[0025] In a fourth aspect, a computer-readable storage medium is provided, comprising instructions, which, when executed on a computer, enable the computer to implement the method provided in the first aspect and any one of the first aspects.

[0026] In a fifth aspect, a computer program product is provided, which includes: a computer program (also referred to as code, or instructions), which, when executed, enables a computer to execute the computing power scheduling method provided in the first aspect and any one of the optional implementations of the first aspect.

[0027] The present application provides a computing power scheduling method, device and readable storage medium. The computing power scheduling node selects computing power nodes for specific application scenarios of services requested by the computing power demand side, selects computing power nodes based on different node selection parameters for different application scenarios, and can consider the availability of nodes, end-to-end requirements, and can select suitable computing power nodes, improve the quality of computing power services, and realize the reasonable allocation of computing power resources. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] Figure 1 It is a schematic diagram of the architecture of the computing power network system applicable to this application;

[0029] Figure 2 It is a schematic flow chart of the computing power scheduling method provided by this application;

[0030] Figure 3 is another schematic flow chart of the computing power scheduling method provided by this application;

[0031] Figure 4 It is a schematic block diagram of the computing power scheduling device provided by this application;

[0032] Figure 5 It is another schematic block diagram of the computing power scheduling device provided in this application. DETAILED DESCRIPTION

[0033] Exemplary embodiments will be described in detail herein, examples of which are shown in the accompanying drawings. When the following description refers to the drawings, the same numbers in different drawings represent the same or similar elements unless otherwise indicated. The implementations described in the following exemplary embodiments do not represent all implementations consistent with the present application. Instead, they are merely examples of devices and methods consistent with some aspects of the present application as detailed in the appended claims.

[0034] Figure 1 Schematic diagram of the computing network system applicable to the embodiment of the present application. Figure 1As shown, the computing power network system may include a computing power scheduling node 101, and the computing power scheduling node 101 may be a computing power scheduling device, or the computing power scheduling node may be a computing power scheduling device or module configured on the computing power scheduling device (such as the module may be a computing power scheduling system software module), and the computing power scheduling node has the functions of computing power resource perception and computing power resource scheduling. The computing power network system may also include multiple computing power nodes 102, and the computing power nodes 102 have computing power resources and can provide computing power services. The computing power scheduling node 101 can obtain the dynamic change information of the computing power resources of the computing power node 102, respond to the computing power request of the computing power demand end 103, and select a suitable computing power node to provide computing power services for the business of the computing power demand end 103 through the computing power scheduling method provided in the embodiment of the present application. It should be understood that Figure 1 This is only an architectural example of a computing power network system, and this application is not limited to this. The solution provided in this application can also be applied to other system architectures, and the system architecture can also include other devices.

[0035] The computing power demand end provided in the embodiment of the present application can be a terminal device, which can also be called a user device. The computing power demand end can include but is not limited to mobile phones, tablet computers, wearable devices (such as smart watches, bracelets, pedometers, etc.), vehicle-mounted devices (such as devices configured in cars, electric vehicles, trains, high-speed railways, etc.), virtual reality (VR) devices, augmented reality (AR) devices, and smart home devices.

[0036] In the current computing power scheduling, the computing power nodes are selected mainly considering factors such as network latency, cost, and computing power performance. There may be a situation where the selected computing power nodes cannot meet the needs of the user end, and there is a problem of unreasonable computing power scheduling. In response to the above problems, this application proposes a specific application scenario in which the computing power scheduling node selects computing power nodes for the business requested by the computing power demand end, and selects computing power nodes for the business based on the node selection parameters corresponding to the application scenario. The node selection parameters include the availability parameters of the computing power node and / or the end-to-end computing power demand parameters of the computing power demand end. It can allocate appropriate computing power nodes to the business based on the application scenario, thereby improving the computing power service quality and realizing the reasonable allocation of computing power resources.

[0037] For ease of understanding, the following Figure 1 The architecture shown is used to describe the method provided in the embodiment of the present application. Figure 2 It is a schematic flow chart of the computing power allocation method provided in the embodiment of the present application. The following is an example of a computing power scheduling node executing the method, and the computing power scheduling node can be a computing power scheduling device or a device (such as a chip or chip module, etc.) or module (such as a software module, etc.) configured in the computing power scheduling device. Figure 2The computing power allocation method shown includes but is not limited to the following steps 201 and 202:

[0038] 201. A computing power scheduling node receives a computing power request from a computing power demand side. The computing power request is used to request selection of a computing power node for a service. The computing power request includes a first field, and the first field is used to indicate an application scenario for selecting a computing power node.

[0039] The computing power request sent by the computing power demand end to the computing power scheduling node includes a first field, which notifies the computing power scheduling node of the application scenario of selecting a computing power node for the business, so that the computing power scheduling node can select a computing power node based on the node selection parameters corresponding to the application scenario.

[0040] The first field can indicate different application scenarios through different values. Exemplarily, if the first field is the first value, the application scenario is the first application scenario, and the first application scenario is to select a computing power node that supports the computing power service of the business for the deployment of the business. In other words, the first application scenario is to select a computing power node that can provide the computing power service of the business for the deployment of the business. The computing power node that can provide the computing power service of the business refers to a computing power node that can provide the computing power demand side with computing and storage capabilities of the business. If the first field is the second value, the application scenario is the second application scenario, and the second application scenario is to select a computing power node from the computing power nodes that have deployed the business to provide the computing power service of the business.

[0041] For example, the first field may be 1 bit, the first value is 0, and the second value is 1, that is, the 1 bit is 0, indicating that the application scenario is the first application scenario, and the 1 bit is 1, indicating that the application scenario is the second application scenario. Or vice versa, the first value may be 1, and the second value may be 0. However, the present application is not limited thereto, and the application scenario in which the computing power demand end requests to select a computing power node may include multiple application scenarios, and when more than two application scenarios are included, the first field may include multiple bits.

[0042] Exemplarily, the computing power request may be carried in an Application-aware Networking (APN) protocol message, and the first field may be a field in the message header of the APN protocol message. After receiving the APN protocol message from the computing power demand side, the computing power scheduling node parses the APN protocol message to obtain the computing power request, and determines the application scenario based on the first field.

[0043] In one implementation, if the application scenario indicated by the first field is the first application scenario, a computing power node that can provide computing power services for the business is selected for deployment of the business. The computing power request also includes a second field, which is used to indicate the end-to-end computing power demand parameters of the computing power demand end. The node selection parameters corresponding to the first application scenario include the end-to-end computing power demand parameters, and the computing power scheduling node refers to the end-to-end computing power demand parameters when selecting computing power nodes. However, the present application is not limited to this, and the computing power scheduling node can obtain the end-to-end computing power demand parameters in other ways, such as through other messages.

[0044] 202. The computing power scheduling node determines at least one computing power node in the computing power node set according to the node selection parameters corresponding to the application scenario. The computing power node is the computing power node selected to provide computing power services for the business. The node selection parameters include one or more of computing power node availability parameters, business end-to-end parameters, and end-to-end computing power demand parameters of the computing power demand end.

[0045] The computing power scheduling node selects a computing power node for the service according to the application scenario indicated by the first field in the computing power request and the node selection parameters corresponding to the application scenario. If the first field indicates the first application scenario, the computing power scheduling node selects a computing power node suitable for the first application scenario according to the node selection parameters corresponding to the first application scenario. If the first field indicates the second application scenario, the computing power scheduling node selects a computing power node suitable for the second application scenario according to the node selection parameters corresponding to the second application scenario. The node selection parameters corresponding to the first application scenario are different from the node selection parameters corresponding to the second application scenario.

[0046] Specifically, the computing power scheduling node uses a node selection algorithm based on the node selection parameters corresponding to the application scenario to select the computing power node suitable for the application scenario. The node selection algorithm uses different node selection parameters, which can be called different node selection algorithms. Therefore, the computing power scheduling node selects computing power nodes based on different node selection parameters corresponding to different application scenarios, which can also be understood as the computing power scheduling node using different node selection algorithms corresponding to different application scenarios to select computing power nodes. When the computing power scheduling node selects computing power nodes for the first application scenario and the second application scenario, the node selection algorithms used are different.

[0047] The following describes how the computing power scheduling node selects a computing power node based on corresponding node selection parameters when the first field indicates a first application scenario and a second application scenario, respectively.

[0048] If the first field indicates a first application scenario, the computing power scheduling node determines at least one computing power node in the computing power node set according to the node selection parameters corresponding to the first application scenario. The node selection parameters corresponding to the first application scenario include computing power node availability parameters and end-to-end computing power demand parameters of the computing power demand end.

[0049] The computing power node availability parameter is used to characterize the failure rate and / or safety of the computing power node.

[0050] In one implementation, the computing power node availability parameter may include node availability and network availability parameters of the computing power node.

[0051] The node availability of a computing power node can characterize the failure rate and / or security of the computing power node itself. For example, the node availability can be determined by factors such as the stability of the computing power node's infrastructure (such as electricity, wiring, etc.), the node's failure rate, fault recovery capability, and security. Among them, security can refer to the security level of software / hardware, the security of the infrastructure (such as the security level of the computer room where the computing power node is located), etc. The node availability of a computing power node can reflect the reliability of the computing power node in providing computing power services.

[0052] The network availability of a computing node can characterize the failure rate and / or security of the computing node's communication network. For example, network availability can be determined by factors such as the packet loss rate, transmission failure rate, and / or transmission rate when the computing node transmits messages through the network. The network availability of a computing node can reflect the communication reliability of the computing node.

[0053] The computing power scheduling node may obtain the computing power node availability parameter from one or more devices in the cloud management device, network management device, or service monitoring device, or the computing power scheduling node may obtain the determining factors of the computing power node availability parameter from one or more devices in the cloud management device, network management device, or service monitoring device, and determine the computing power node availability parameter of the computing power node based on the obtained parameter determining factors. This application does not limit this.

[0054] The computing power node availability parameter can reflect the ability of a computing power node to provide end-to-end computing power services through node availability and network availability, that is, the ability to request computing power services from the computing power demand side to the computing power node accurately feeding back the calculation results to the computing power demand side.

[0055] The node selection parameters corresponding to the first application scenario also include an end-to-end computing power requirement parameter, which may include but is not limited to one or more of the following:

[0056] The end-to-end availability parameters of the service expected by the computing power demand side;

[0057] The end-to-end latency parameters of the services expected by the computing power demand side;

[0058] The expected computing power performance parameters of the computing power demand side;

[0059] The computing power scheduling cost parameters expected by the computing power demand side.

[0060] The end-to-end computing power demand parameter may include the end-to-end availability parameter of the service expected by the computing power demand side. Or it may be called the end-to-end availability parameter of the service acceptable to the computing power demand side. For example, the end-to-end availability of the service acceptable to the computing power demand side may characterize the failure rate and / or security of the computing power node acceptable to the computing power demand side, such as one or more of the node failure rate, node security, node network failure rate, and node network security acceptable to the computing power demand side. The end-to-end availability of the service acceptable to the computing power demand side may be determined by the service demand.

[0061] The computing power scheduling node can determine the availability index of the computing power node based on the computing power node availability parameter and the end-to-end availability parameter of the service expected by the computing power demand side. The availability index of the computing power node can indicate whether the end-to-end availability of the computing power node meets the end-to-end availability required by the computing power demand side, thereby selecting a computing power node that is more suitable for providing computing power services for the service of the computing power demand side. The end-to-end availability parameter expected by the computing power demand side can also be called the service end-to-end demand availability.

[0062] Exemplarily, the availability index of the computing power node corresponding to the first application scenario is It can be obtained based on the following algorithm:

[0063]

[0064] The availability parameter can be in the form of a percentage or a true fraction, which is not limited in this application. The larger the value, the higher the probability that the end-to-end availability of the computing power node can meet the end-to-end availability expected by the computing power demand side.

[0065] When selecting computing nodes for the first application scenario (selecting computing nodes that can provide computing services for the business for business deployment), the computing power scheduling node refers to the availability index of the computing power nodes, and determines the extent to which the availability of the computing power nodes in the computing power node set can meet the end-to-end availability required by the computing power demand side. This can make up for the deficiency that the end-to-end availability of the computing power nodes is not considered when selecting computing power nodes. It can select computing power nodes that match the availability requirements of the computing power demand side, ensure the end-to-end availability of computing power services, improve the quality of computing power services, and achieve reasonable allocation of computing power resources.

[0066] The end-to-end computing power demand parameters may also include the end-to-end delay parameters of the services expected by the computing power demand side. The end-to-end delay parameters of the services expected by the computing power demand side may also be referred to as the end-to-end delay parameters of the services acceptable to the computing power demand side. And, the node selection parameters corresponding to the first application scenario also include service end-to-end parameters, which include service end-to-end delay parameters, which are determined based on the service delay parameters of the computing power node and the network delay parameters of the computing power node.

[0067] The service delay parameter of the computing power node is used to characterize the average response delay of the computing power node to the computing power service that needs to be deployed. The network delay parameter of the computing power node is used to characterize the network transmission delay required by the computing power node. The sum of the service delay parameter and the network delay parameter of the computing power node is the end-to-end delay of the computing power service provided by the computing power node, that is, the delay from the computing power demand side sending a request, the computing power node providing the computing power service and feeding back the calculation results to the computing power demand side.

[0068] The computing power scheduling node can determine the end-to-end delay parameters of the computing power node's business based on the computing power node's service delay parameters and network delay parameters. The computing power scheduling node can determine the latency index of the computing power node with the end-to-end latency of the business expected by the computing power demand side as a reference. The latency index of the computing power node can indicate whether the end-to-end latency of the computing power service provided by the computing power node can meet the end-to-end latency expected by the computing power demand side, thereby selecting a computing power node that is more suitable for providing computing power services for the business of the computing power demand side.

[0069] For example, the latency indicator of the computing power node corresponding to the first application scenario is It can be obtained based on the following algorithm:

[0070]

[0071] The smaller the end-to-end delay of computing power nodes providing computing power services is compared with the expected end-to-end delay of computing power demand side, the lower the delay index. The larger the value, the higher the quality of the computing service that the computing power node can provide.

[0072] Latency Indicators The larger the value, the higher the probability that the end-to-end latency of the computing power node in providing computing power services can meet the end-to-end availability expected by the computing power demand side.

[0073] Compared with the current computing power scheduling that only considers network latency, there may be a situation where the latency requirements of the computing power demand side cannot be met. When selecting computing power nodes for the first application scenario (selecting computing power nodes that can provide computing power services for the business for business deployment), the computing power scheduling node refers to the latency index of computing power services provided by the computing power nodes to determine whether the end-to-end latency of computing power services provided by the computing power nodes in the computing power node set can meet the end-to-end latency requirements of the computing power demand side. The computing power nodes that meet the latency requirements of the computing power demand side can be selected to improve the quality of computing power services and achieve reasonable allocation of computing power resources.

[0074] The end-to-end computing power demand parameter may also include a computing power performance parameter expected by the computing power demand side. The computing power performance parameter expected by the computing power demand side is used to characterize the computing power performance or processing capability that the computing power demand side expects the computing power node to have. The node selection parameter corresponding to the first application scenario also includes an available computing power performance parameter of the computing power node, which is used to characterize the computing power performance or processing capability of the computing power node.

[0075] The computing power scheduling node can determine the performance indicators of the computing power nodes based on the available computing power performance parameters of the computing power nodes and the computing power performance parameters expected by the computing power demand side as a reference. The performance indicators of the computing power nodes can characterize whether the computing power performance of the computing power nodes can meet the computing power performance required by the computing power demand side, thereby selecting a computing power node that is more suitable for providing computing power services for the business of the computing power demand side.

[0076] Exemplarily, the performance index of the computing power node corresponding to the first application scenario is It can be obtained based on the following algorithm:

[0077]

[0078] The greater the performance of the computing power node is compared to the expected computing power performance of the computing power demand side, the performance index The larger the value, the higher the quality of the computing service that the computing power node can provide.

[0079] The end-to-end computing power demand parameters may also include the computing power scheduling cost parameters expected by the computing power demand side. The computing power scheduling cost parameters expected by the computing power demand side may characterize the transmission bandwidth cost and / or time cost of the computing power service acceptable to the computing power demand side. The node selection parameters corresponding to the first application scenario also include end-to-end service parameters, which include the scheduling cost of the computing power node, and the scheduling cost is used to characterize the end-to-end transmission bandwidth cost and / or time cost required to schedule the computing power node to provide computing power services. Based on the scheduling parameters of the computing power node, the computing power scheduling node can determine the cost index of the computing power node with reference to the computing power scheduling cost expected by the computing power demand side. The cost index of the computing power node can characterize whether the cost overhead of the computing power node providing computing power services can meet the cost overhead acceptable to the computing power demand side, thereby selecting a computing power node that is more suitable for providing computing power services for the business of the computing power demand side.

[0080] Exemplarily, the cost indicator of the computing power node corresponding to the first application scenario is It can be obtained based on the following algorithm:

[0081]

[0082] The smaller the scheduling cost of the computing power node is compared with the expected scheduling cost of the computing power demand side, the better the performance index is. The larger the value, the higher the quality of the computing service that the computing power node can provide.

[0083] In one implementation, the computing power scheduling node can determine multiple indicators of each computing power node in the computing power node set based on the node selection parameters corresponding to the first application scenario, and then determine the comprehensive service indicator of each computing power node based on the multiple indicators. The comprehensive service indicator is determined by a weighted combination of the multiple indicators.

[0084] Among them, the multiple indicators of the computing power node include the availability indicator, and the weight of the availability indicator is greater than the weight of other indicators. Whether the computing power node is available is the basic condition to ensure that the computing power node provides computing power services. Therefore, the computing power scheduling node uses the availability indicator of the computing power node as the main indicator to select the computing power node.

[0085] The multiple indicators for reference in selecting computing power nodes in the first application scenario may also include one or more of the above-mentioned latency indicators, performance indicators and cost indicators.

[0086] Exemplarily, the multiple indicators include availability indicators, latency indicators, performance indicators and cost indicators. The comprehensive service indicator S1 of the computing power node corresponding to the first application scenario can be obtained based on the following algorithm:

[0087]

[0088] Specifically, the comprehensive service index S1 of the computing power node corresponding to the first application scenario can be specifically expressed as:

[0089]

[0090] Among them, the weight β of the availability index in the comprehensive service index S1 of the computing power node is greater than the weights α, γ and δ of other indexes. For example, α=γ=δ=20%, β=40%. However, the present application is not limited to this.

[0091] For example, the computing power node set is a set of computing power nodes that can provide computing power services for the computing power demand side. The computing power scheduling node determines, based on the computing power request, that the application scenario for selecting computing power nodes for the service requested by the computing power demand side is to deploy the service, that is, the first application scenario. The computing power scheduling node determines the comprehensive service index of each computing power node in the computing power node set based on the node selection parameters corresponding to the first application scenario, or in other words, the computing power scheduling node uses the node selection algorithm corresponding to the first application scenario to determine the comprehensive service index of each computing power node in the computing power node set, and the node selection algorithm corresponding to the first application scenario specifically calculates the comprehensive service index of the computing power node based on the node selection parameters corresponding to the first application scenario.

[0092] The computing power scheduling node can determine at least one computing power node with the highest comprehensive service index among the computing power nodes in the computing power node set based on the comprehensive service index of the computing power node, as the selected computing power node. The specific number of selected computing power nodes can be determined based on the number of computing power nodes required by the business or computing power demand side, or determined by the computing power scheduling node, and this application does not limit this.

[0093] After the computing power scheduling node selects at least one computing power node corresponding to the first application scenario based on the computing power request, the computing power scheduling node deploys the service to the at least one computing power node.

[0094] Exemplarily, the computing power demand end is a terminal, and the terminal sends a computing power request to the computing power scheduling node, requesting the deployment of the computing power service corresponding to the AR business, that is, the first field in the computing power request indicates the first application scenario. The computing power request also includes a second field, which indicates the end-to-end demand parameters of the terminal. In response to the computing power request, the computing power scheduling node uses the node selection algorithm corresponding to the first application scenario based on the computing power node availability parameters corresponding to the first application scenario and the end-to-end demand parameters of the terminal (which may also include other parameters in the foregoing text), and determines the comprehensive indicators of the computing power nodes in the computing power node set, thereby selecting a suitable computing power node for the AR business based on the comprehensive indicators, and deploying the computing power service of the AR business to the selected computing power node, so that the selected computing power node can be subsequently scheduled to perform the computing power service of the AR business.

[0095] After the computing power scheduling node deploys the computing power service of the business to the at least one computing power node, the node selection parameters of the computing power node will change, such as processing power, etc. The computing power scheduling node can update the node selection parameters of the computing power node for the next node selection, such as node selection for other businesses and / or other computing power demand ends.

[0096] The above describes that the first field indicates the first application scenario, and the computing power scheduling node determines at least one computing power node in the computing power node set according to the node selection parameter corresponding to the first application scenario. The following describes that the computing power scheduling node selects a computing power node when the first field indicates the second application scenario.

[0097] If the first field indicates the second application scenario, a computing power node is selected from the computing power nodes where the service has been deployed to provide computing power services for the service, and the service is the service indicated by the computing power request. The computing power scheduling node determines at least one computing power node in the computing power node set according to the node selection parameters corresponding to the second application scenario. The node selection parameters corresponding to the second application scenario include computing power node availability parameters.

[0098] In the second application scenario, the computing power node set is the node where the service has been deployed. The computing power nodes in the computing power node set may be computing power nodes where the service has been deployed based on computing power requests from one or more computing power demand ends. After receiving the computing power request indicating the second application scenario, the computing power scheduling node may select a computing power node from the computing power nodes where the service has been deployed to provide computing power services for the computing power demand end of the computing power request.

[0099] The node selection parameters corresponding to the second application scenario include computing power node availability parameters and service end-to-end parameters. The computing power node availability parameters include node availability parameters and network availability parameters of the computing power node. The introduction of specific node availability parameters and network availability parameters can be referred to the previous description, which will not be repeated here.

[0100] The node selection parameter may also include a service availability parameter of the computing power node. The service availability parameter is used to characterize the availability of the computing power service provided by the computing power node for the service. Since the computing power nodes in the computing power node set are nodes that have deployed the service, the computing power scheduling node can obtain the availability of the computing power service provided by the computing power node for the service, such as the service availability parameter can be determined by the actual packet loss rate, service failure ratio and other failure rate factors and / or security factors. The computing power scheduling node can obtain the computing power node service availability parameter from one or more devices in the cloud management device, network management device or service monitoring device, or the computing power scheduling node can obtain the determining factors of the service availability parameter of the computing power node from one or more devices in the cloud management device, network management device or service monitoring device, and determine the service availability parameter of the computing power node based on the obtained parameter determining factors. The service availability parameter may belong to the computing power node availability parameter, that is, in the second application scenario, the computing power node availability parameter includes the node availability parameter, the network availability parameter and the service availability parameter. The computing power node availability parameter may also be referred to as the service end-to-end availability parameter of the computing power node.

[0101] The computing power scheduling node is based on the computing power node availability parameters, such as the node availability parameters, network availability parameters and service availability parameters of the computing power nodes, which can determine the availability indicators of the computing power nodes corresponding to the second application scenario. By referring to the actual availability indicators of the computing power nodes, the computing power nodes providing computing power services are selected, and computing power nodes with higher service quality can be selected.

[0102] For example, the availability index of the computing power node corresponding to the second application scenario is It can be obtained based on the following algorithm:

[0103]

[0104] Among them, the higher the computing power node availability parameter, the higher the availability index. The larger the value, the higher the quality of the computing service that the computing power node can provide.

[0105] The service end-to-end parameters corresponding to the second application scenario may also include the service end-to-end delay parameters of the computing power node. As described above, the service end-to-end delay parameters of the computing power node can be determined by the service delay parameters and the network delay parameters. The introduction of the specific service delay parameters and the network delay parameters can refer to the description above and will not be repeated here. The computing power scheduling node can determine the service end-to-end delay parameters of the computing power node based on the service delay parameters and the network delay parameters of the computing power node. The service end-to-end parameters corresponding to the second application scenario also include the average delay parameters of the computing power nodes in the computing power node set. The average delay parameter is used to characterize the average value of the service end-to-end delay of the computing power nodes in the computing power node set. The computing power scheduling node can determine the delay index corresponding to the second application scenario based on the service end-to-end delay parameters of the computing power node and the average delay parameters of the computing power node set. The computing power scheduling node refers to the delay index and selects the computing power node that provides computing power services, and can select the computing power node with a smaller end-to-end delay in the computing power node set.

[0106] For example, the latency indicator of the computing power node corresponding to the second application scenario is It can be obtained based on the following algorithm:

[0107]

[0108] Among them, the smaller the end-to-end delay of the computing power node is compared with the average delay, the higher the delay index The larger the value, the higher the quality of the computing service that the computing power node can provide.

[0109] The node selection parameters corresponding to the second application scenario may also include the available computing power performance parameters of the computing power nodes and the average available performance parameters of the computing power node set. The available computing power performance parameters of the computing power nodes can refer to the previous description and will not be repeated here. The average available performance parameter is the average value of the computing power performance (or processing capacity) of the computing power nodes in the computing power node set. The computing power scheduling node can determine the performance index corresponding to the second application scenario based on the available performance parameters of the computing power nodes and the average available performance parameters of the computing power node set. The computing power scheduling node refers to the latency index to select the computing power node that provides computing power services, and can select the computing power node with high performance in the computing power node set.

[0110] Exemplarily, the performance index of the computing power node corresponding to the second application scenario is It can be obtained based on the following algorithm:

[0111]

[0112] The greater the available computing power performance of the computing power node is compared to the average available computing power performance, the higher the latency indicator The larger the value, the higher the quality of the computing service that the computing power node can provide.

[0113] In one implementation, the computing power scheduling node can determine multiple indicators of each computing power node in the computing power node set based on the node selection parameters corresponding to the second application scenario, and then determine the comprehensive service indicator of each computing power node (i.e., the comprehensive service indicator corresponding to the second application scenario) based on the multiple indicators. The comprehensive service indicator is determined by a weighted combination of the multiple indicators.

[0114] Among them, for the second application scenario, the multiple indicators of the computing power node include an availability indicator, and the weight of the availability indicator is greater than the weight of other indicators. Whether the computing power node is available is the basic condition to ensure that the computing power node provides computing power services. Therefore, the computing power scheduling node uses the availability indicator of the computing power node as the main indicator to select the computing power node.

[0115] The multiple indicators for reference in selecting computing power nodes in the second application scenario may also include one or more of the above-mentioned latency indicators, performance indicators and cost indicators.

[0116] Exemplarily, the multiple indicators include availability indicators, latency indicators, and performance indicators. The comprehensive service indicator S2 of the computing power node corresponding to the second application scenario can be obtained based on the following algorithm:

[0117]

[0118] Specifically, the comprehensive service index S2 of the computing power node corresponding to the second application scenario can be specifically expressed as:

[0119]

[0120] Among them, the weight β of the availability index in the comprehensive service index S2 of the computing power node corresponding to the second application scenario is greater than the weights α and γ of other indexes. For example, α=γ=30%, β=40%. However, the present application is not limited to this.

[0121] After the computing power scheduling node determines, based on the computing power request, that the application scenario requested by the computing power demand side to select a computing power node for the business is to provide computing power services for the business (i.e., the second application scenario), the computing power scheduling node determines the comprehensive service indicators of each computing power node in the computing power node set based on the node selection parameters corresponding to the second application scenario. In other words, the computing power scheduling node adopts the node selection algorithm corresponding to the second application scenario to determine the comprehensive service indicators of each computing power node in the computing power node set. The node selection algorithm corresponding to the second application scenario specifically calculates the comprehensive service indicators of the computing power nodes based on the node selection parameters corresponding to the second application scenario.

[0122] The computing power scheduling node can determine at least one computing power node with the highest comprehensive service index among the computing power nodes in the computing power node set based on the comprehensive service index of the computing power node, as the selected computing power node. The specific number of selected computing power nodes can be determined based on the number of computing power nodes that provide computing power services for the business required by the business or computing power demand side, or determined by the computing power scheduling node, and this application does not limit this.

[0123] After the computing power scheduling node selects at least one computing power node corresponding to the first application scenario based on the computing power request, the computing power scheduling node schedules the at least one computing power node to provide computing power services for the business. The at least one computing power node provides computing power services for the business and feeds back the calculation results to the computing power demand end.

[0124] Exemplarily, the computing power demand side is a terminal, which sends a computing power request to the computing power scheduling node, requesting the selection of a computing power node to perform the computing power service corresponding to the AR business, that is, the first field in the computing power request indicates the second application scenario. The computing power scheduling node uses the node selection algorithm corresponding to the second application scenario based on the node selection parameters corresponding to the second application scenario to determine the comprehensive indicators of the computing power nodes in the computing power node set, thereby selecting a suitable computing power node for the AR business based on the comprehensive indicators, and scheduling the selected computing power node to provide the computing power service of the AR business. The computing power node performs the computing power service of the AR business and feeds back the calculation results to the computing power demand side.

[0125] After the computing power scheduling node schedules at least one computing power node to provide computing power services for the business, the node selection parameters of the computing power node will change. The computing power scheduling node can update the node selection parameters of the computing power node for the next node selection, such as node selection performed for other businesses and / or other computing power demand ends.

[0126] The specific process of computing power scheduling nodes processing computing power requests can be referred to Figure 3As shown, in step 1, after the computing power node receives the computing power request from the computing power demand side, the computing power scheduling node executes step 2 to parse the computing power request, obtains the first field, and determines the application scenario indicated by the first field in step 3. If the first field indicates the first application scenario, the computing power scheduling node executes step 4, executes the node selection algorithm corresponding to the first application scenario, and determines at least one computing power node. Before step 4, the computing power scheduling node can obtain the node selection parameters of the computing power node set corresponding to the first application scenario. Specifically, the computing power scheduling node can execute the corresponding node selection algorithm based on the node selection parameters corresponding to the first application scenario, obtain the comprehensive service index S1 of the computing power node in the computing power node set, and thus select at least one computing power node based on the comprehensive service index S1. If the first field indicates the second application scenario, the computing power scheduling node executes step 5, executes the node selection algorithm corresponding to the second application scenario, and determines at least one computing power node. Before step 4, the computing power scheduling node can obtain the node selection parameters of the computing power node set corresponding to the first application scenario. Specifically, the computing power scheduling node can execute the corresponding node selection algorithm based on the node selection parameters corresponding to the second application scenario, obtain the comprehensive service index S2 of the computing power nodes in the computing power node set, and then select at least one computing power node based on the comprehensive service index S2. After the computing power scheduling node selects the computing power node for the service, it executes step 6 to publish the selected computing power node. For the first application scenario, the computing power scheduling node deploys the service to the selected computing power node. For the second application scenario, the computing power scheduling node schedules the selected computing power node to provide computing power services for the service.

[0127] The above-mentioned scheme provided by the present application is that the computing power scheduling node selects computing power nodes for specific application scenarios of the business requested by the computing power demand side, selects computing power nodes based on different node selection parameters for different application scenarios, and can consider the availability and end-to-end requirements of the nodes, and can select suitable computing power nodes, which can improve the quality of computing power services and realize the reasonable allocation of computing power resources.

[0128] Figure 4 It is a schematic block diagram of a computing power scheduling device 400 provided in an embodiment of the present application. The computing power scheduling device may include a processing module 401 and a transceiver module 402. Among them, the transceiver module 402 is used to receive a computing power request from a computing power demand end, the computing power request is used to request the selection of a computing power node for a service, and the computing power request includes a first field, and the first field is used to indicate an application scenario for selecting a computing power node; the processing module 401 determines at least one computing power node in a computing power node set according to a node selection parameter corresponding to the application scenario, and the at least one computing power node is a computing power node that supports providing computing power services for the service, wherein the node selection parameter includes one or more of a computing power node availability parameter, a service end-to-end parameter, and an end-to-end computing power demand parameter of the computing power demand end.

[0129] Optionally, the computing power node availability parameter is used to characterize the failure rate and / or security of the computing power node, and the computing power node availability parameter includes a node availability parameter and a network availability parameter of the computing power node.

[0130] Optionally, the end-to-end computing power requirement parameter includes one or more of the following:

[0131] The end-to-end latency parameters of the service expected by the computing power demander;

[0132] The expected computing power performance parameters of the computing power demand side;

[0133] The end-to-end availability parameters of the service expected by the computing power demander;

[0134] The computing power scheduling cost parameter expected by the computing power demand side.

[0135] Optionally, if the first field is a first value, the application scenario is a first application scenario, and the first application scenario is to select a computing power node that supports the computing power service for providing the business for deployment of the business; or, if the first field is a second value, the application scenario is a second application scenario, and the second application scenario is to select a computing power node from the computing power nodes that have deployed the business to provide the computing power service for the business.

[0136] Optionally, the application scenario is the first application scenario, and the transceiver module 402 is also used to deploy the service to the at least one computing power node.

[0137] Optionally, the node selection parameter includes the end-to-end computing power requirement parameter, and the computing power request includes a second field, and the second field is used to indicate the end-to-end computing power requirement parameter.

[0138] Optionally, the end-to-end service parameter includes an end-to-end service delay parameter, and / or the node selection parameter also includes an available computing power performance parameter.

[0139] Optionally, the application scenario is the second application scenario, and the processing module 401 is further used to control the transceiver module 402 to schedule the at least one computing power node to provide computing power services for the business.

[0140] Optionally, the computing power node set is a set of computing power nodes on which the service has been deployed.

[0141] Optionally, the service end-to-end parameters include one or more of the service end-to-end delay parameters of the computing power node and / or the service average end-to-end delay parameters of the computing power node set; and / or, the node selection parameters also include the service availability parameters of the computing power node, the available computing power performance parameters of the computing power node and / or the average computing power performance parameters of the computing power node set.

[0142] Optionally, the processing module 401 is specifically used to determine the comprehensive service index of each computing node in the computing node set according to the node selection parameter corresponding to the application scenario, and determine the at least one node according to the comprehensive service index of the computing nodes in the computing node set. The comprehensive service index is obtained by weighted combination of multiple indicators, and the multiple indicators include an availability index, which is determined according to the availability parameter of the computing node, and the weight corresponding to the availability index is greater than the weight corresponding to the indicator other than the availability index in the multiple performance indicators.

[0143] Optionally, when the application scenario is the first application scenario, the multiple performance indicators also include a latency indicator, a computing power performance indicator and a scheduling cost indicator corresponding to the first application scenario; or, when the application scenario is the second application scenario, the multiple performance indicators also include a latency indicator and a computing power performance indicator corresponding to the second application scenario.

[0144] Figure 5 is a schematic block diagram of a computing power scheduling device 500 provided in an embodiment of the present application. The device can be used to implement the function of the computing power scheduling device in the above method. Among them, the device can be a chip system. In the embodiment of the present application, the chip system can be composed of a chip, or it can include a chip and other discrete devices.

[0145] like Figure 5 As shown, the device 500 may include at least one processor 501, which is used to implement the function of the computing power scheduling device in the method provided in the embodiment of the present application.

[0146] The apparatus 500 may also include a communication interface 503 for communicating with other devices via a transmission medium, so that the device in the apparatus 500 can communicate with other devices. The communication interface 503 may be, for example, a transceiver, an interface, a bus, a circuit, or a device capable of implementing transceiver functions. The processor 501 may use the communication interface 503 to send and receive data and / or information, and to implement Figure 2 The terminal device identification method in the corresponding embodiment.

[0147] Exemplarily, the communication interface 503 is used to receive a computing power request from a computing power demand side, the computing power request is used to request the selection of a computing power node for a service, and the computing power request includes a first field, the first field is used to indicate an application scenario for selecting a computing power node; the processor 501 can be used to determine at least one computing power node in the computing power node set according to the node selection parameter corresponding to the application scenario, and the computing power node is a computing power node selected for the service to provide computing power services. Among them, the node selection parameter includes a computing power node availability parameter and / or an end-to-end computing power demand parameter of the computing power demand side.

[0148] The device 500 may also include at least one memory 502 for storing program instructions and / or data. The memory 502 is coupled to the processor 501. The coupling in the embodiment of the present application is an indirect coupling or communication connection between devices, units or modules, which may be electrical, mechanical or other forms, and is used for information exchange between devices, units or modules. The processor 501 may operate in conjunction with the memory 502. The processor 501 may execute program instructions stored in the memory 502. At least one of the at least one memory may be included in the processor.

[0149] The specific connection medium between the processor 501, the memory 502 and the communication interface 503 is not limited in the embodiment of the present application. Figure 5 In the embodiment, the processor 501, the memory 502 and the communication interface 503 are connected via a bus 504. The bus 504 is Figure 5 The connection between other components is only for illustration and is not intended to be limiting. The bus can be divided into an address bus, a data bus, a control bus, etc. For ease of representation, Figure 5 Only one thick line is used in the diagram, but this does not mean that there is only one bus or only one type of bus.

[0150] The present application also provides a computer program product, which includes: a computer program (also referred to as code or instruction), which, when executed, causes a computer to execute Figure 2 A method according to any one of the embodiments shown.

[0151] The present application also provides a computer-readable storage medium, which stores a computer program (also referred to as code or instruction). When the computer program is executed, the computer executes Figure 2 A method according to any one of the embodiments shown.

[0152] It should be understood that the processor in the embodiment of the present application can be an integrated circuit chip with signal processing capabilities. In the implementation process, each step of the above method embodiment can be completed by an integrated logic circuit of hardware in the processor or an instruction in the form of software.

[0153] The terms "unit", "module" and the like used in this specification may be used to represent a computer-related entity, hardware, firmware, a combination of hardware and software, software, or software in execution.

[0154] It will be appreciated by those skilled in the art that the various illustrative logical blocks and steps described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or in combination with computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application. In the several embodiments provided in this application, it should be understood that the disclosed devices, equipment and methods can be implemented in other ways. For example, the device embodiments described above are only schematic, for example, the division of the above-mentioned units is only a logical function division, and there may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point, the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0155] The units described above as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0156] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0157] Those skilled in the art will readily appreciate other embodiments of the present application after considering the specification and practicing the invention disclosed herein. The present application is intended to cover any modification, use or adaptation of the present application, which follows the general principles of the present application and includes common knowledge or customary techniques in the art that are not disclosed in the present application. The specification and examples are intended to be exemplary only, and the true scope and spirit of the present application are indicated by the following claims.

[0158] It should be understood that the present application is not limited to the precise structures that have been described above and shown in the drawings, and that various modifications and changes may be made without departing from the scope thereof. The scope of the present application is limited only by the appended claims.

Claims

1. A computing power scheduling method, characterized in that: include: Receive a computing power request from a computing power demand side, where the computing power request is used to request selection of a computing power node for a service, and the computing power request includes a first field, where the first field is used to indicate an application scenario for selecting a computing power node; According to the node selection parameters corresponding to the application scenario, at least one computing node is determined in the computing node set, and the at least one computing node is a computing node that supports providing computing services for the business. Among them, the node selection parameters include one or more of the computing power node availability parameters, the service end-to-end parameters and the end-to-end computing power demand parameters of the computing power demand end.

2. The method according to claim 1, characterized in that The computing power node availability parameter is used to characterize the failure rate and / or safety of the computing power node, and the computing power node availability parameter includes a node availability parameter and a network availability parameter of the computing power node.

3. The method according to claim 1 or 2, characterized in that: The end-to-end computing power requirement parameters include one or more of the following: The end-to-end delay parameter of the service expected by the computing power demander; The computing power performance parameters expected by the computing power demand side; The end-to-end availability parameters of the service expected by the computing power demander; The computing power scheduling cost parameters expected by the computing power demand side.

4. The method according to claim 3, characterized in that If the first field is a first value, the application scenario is a first application scenario, and the first application scenario is to select a computing power node that supports the computing power service for providing the business for deployment of the business; or, If the first field is a second value, the application scenario is a second application scenario, and the second application scenario is to select a computing power node from among the computing power nodes where the service has been deployed to provide computing power services for the service.

5. The method according to claim 4, characterized in that The application scenario is the first application scenario, and the method further includes: Deploy the service to the at least one computing power node.

6. The method according to claim 5, characterized in that The node selection parameter includes the end-to-end computing power requirement parameter, and the computing power request includes a second field, where the second field is used to indicate the end-to-end computing power requirement parameter.

7. The method according to claim 6, characterized in that The end-to-end service parameters include end-to-end service delay parameters, and / or the node selection parameters also include available computing power performance parameters.

8. The method according to claim 4, characterized in that The application scenario is the second application scenario, and the method further includes: Scheduling the at least one computing power node to provide computing power services for the business.

9. The method according to claim 8, characterized in that The computing power node set is a set of computing power nodes on which the service has been deployed.

10. The method according to claim 9, characterized in that The service end-to-end parameters include one or more of the service end-to-end delay parameters of the computing power node and / or the service average end-to-end delay parameters of the computing power node set; and / or, the node selection parameters also include the service availability parameters of the computing power node, the available computing power performance parameters of the computing power node and / or the average computing power performance parameters of the computing power node set.

11. The method according to any one of claims 4 to 10, characterized in that The determining at least one computing power node in the computing power node set according to the node selection parameter corresponding to the application scenario includes: Determine the comprehensive service index of each computing power node in the computing power node set according to the node selection parameters corresponding to the application scenario; Determine the at least one node according to the comprehensive service indicators of the computing nodes in the computing node set, Among them, the comprehensive service indicator is obtained by weighted combination of multiple indicators, and the multiple indicators include an availability indicator, and the availability indicator is determined based on the availability parameter of the computing power node. The weight corresponding to the availability indicator is greater than the weight corresponding to the indicators in the multiple performance indicators except the availability indicator.

12. The method according to claim 11, characterized in that When the application scenario is the first application scenario, the multiple performance indicators further include a latency indicator, a computing power performance indicator, and a scheduling cost indicator corresponding to the first application scenario; or, When the application scenario is the second application scenario, the multiple performance indicators also include a latency indicator and a computing power performance indicator corresponding to the second application scenario.

13. A computing power scheduling device, characterized in that: include: A transceiver module, used to receive a computing power request from a computing power demand end, wherein the computing power request is used to request selection of a computing power node for a service, and the computing power request includes a first field, and the first field is used to indicate an application scenario for selecting a computing power node; A processing module is used to determine at least one computing node in the computing node set according to the node selection parameter corresponding to the application scenario, where the computing node is a computing node selected to provide computing services for the business. Among them, the node selection parameters include one or more of computing power node availability parameters, service end-to-end parameters and end-to-end computing power demand parameters of the computing power demand end.

14. A computing power scheduling device, characterized in that: The system comprises a processor coupled to a memory, wherein the processor is used to schedule and execute program instructions in the memory to implement the method according to any one of claims 1 to 12.

15. A computer-readable storage medium, characterized in that: The storage medium stores a computer program, and when the computer program is executed by a processor, the method according to any one of claims 1 to 12 is implemented.

Images