Scheduling method and apparatus, gateway, storage medium, and computer program product

Abstract

The present application discloses a scheduling method and apparatus, a gateway, a storage medium, and a computer program product. The method comprises: receiving first requests sent by a plurality of first devices, wherein the first requests are used for requesting to provide services for the first devices, the first requests comprise first information and second information, the first information is used for identifying computing power services requested by the first devices, and the second information represents computing power service requirements of the first devices; according to service types, on the basis of the plurality of pieces of first information, aggregating the plurality of first requests to obtain a first aggregation result; and when the first aggregation result represents that the types of the computing power services requested by the plurality of first devices are the same, on the basis of the plurality of pieces of second information, selecting, for each first device among the plurality of first devices, a computing power node for providing the computing power service.

Description

Scheduling methods, devices, gateways, storage media, and computer program products

[0001] Cross-references to related applications

[0002] This application claims priority to Chinese Patent Application No. 202410946542.1, filed on July 15, 2024, the entire contents of which are incorporated herein by reference. Technical Field

[0003] This application relates to the field of communication technology, and in particular to a scheduling method, apparatus, gateway, storage medium, and computer program product. Background Technology

[0004] In related technologies, after the ingress gateway in a computing power router receives a service request sent by a user, it can select a suitable computing power node for each service request, and the computing power node can provide computing power services to the user corresponding to the service request.

[0005] However, when the ingress gateway needs to schedule service requests from multiple users, and the computing power service types (which can also be understood as the service types of computing power services) corresponding to the multiple service requests are the same, there is currently no effective solution for how to select computing power nodes for the services requested by multiple users to achieve optimal scheduling. Summary of the Invention

[0006] To address the related technical issues, embodiments of this application provide a scheduling method, apparatus, gateway, storage medium, and computer program product.

[0007] The technical solution of this application embodiment is implemented as follows:

[0008] This application provides a scheduling method applied to a first gateway, including:

[0009] Receive a first request sent by a plurality of first devices, the first request being used to request services to the first devices, the first request including first information and second information, the first information being used to identify the computing power service requested by the first devices, and the second information representing the computing power service needs of the first devices;

[0010] Based on the service type and multiple pieces of first information, multiple first requests are aggregated to obtain the first aggregation result;

[0011] If the first aggregation result indicates that the types of computing power services requested by the multiple first devices are the same, a computing power node is selected to provide computing power services for each of the multiple first devices based on multiple second pieces of information.

[0012] In the above scheme, selecting a computing power node to provide computing power services for each of the multiple first devices includes:

[0013] Using third information and multiple second information, a computing power node is selected for each of the multiple first devices to provide computing power services. The third information includes the status information of multiple computing power nodes.

[0014] In the above scheme, selecting a computing power node to provide computing power services for each of the multiple first devices includes:

[0015] Using the plurality of second information, a fourth piece of information is determined, wherein the fourth piece of information represents the priority of the plurality of first devices;

[0016] Using the fourth and third information, a computing power node is selected to provide computing power services for each of the multiple first devices.

[0017] In the above scheme, the status information of the computing power node includes the load information of the computing power node, and the step of selecting a computing power node to provide computing power services for each of the multiple first devices includes:

[0018] Using the load information and multiple second information, one or more computing power nodes with load information greater than a first threshold are selected to provide computing power services for each of the multiple first devices.

[0019] In the above scheme, the status information of the computing power node includes the number of times the computing power node provides services, and the step of selecting a computing power node to provide computing power services for each of the multiple first devices includes:

[0020] Using the number of service provision times and multiple pieces of second information, one or more computing power nodes with a number of service provision times less than a second threshold are selected to provide computing power services for each of the multiple first devices.

[0021] or,

[0022] The priority of the computing power nodes is determined by the number of times the service is provided; the priority of the computing power nodes is used to select a computing power node to provide computing power services for each of the multiple first devices.

[0023] The method in the above scheme further includes:

[0024] Obtain the third information.

[0025] In the above scheme, obtaining the third information includes:

[0026] Receive status information sent by multiple computing power nodes;

[0027] The latest received status information is used as the third information.

[0028] The method in the above scheme further includes:

[0029] The system receives a second request from one or more second devices. The second request is used to request services to the second devices. The second request includes fifth information and sixth information. The fifth information is used to identify the computing power service requested by the second devices, and the sixth information represents the computing power service requirements of the second devices. The service type corresponding to the second request is different from the service type corresponding to the first request.

[0030] According to the service type, and based on multiple first pieces of information and one or more fifth pieces of information, multiple first requests and one or more second requests are aggregated to obtain a second aggregation result; the second aggregation result contains at least two service types;

[0031] For each of the at least two service types, based on the device's computing power service requirements, select computing power nodes to provide computing power services to the device; where,

[0032] When selecting computing nodes to provide computing power services for a device, the method further includes:

[0033] The seventh piece of information is used to select computing nodes that provide computing power services to the device, whereby the seventh piece of information represents the correspondence between service type and service priority.

[0034] This application embodiment also provides a scheduling device, disposed in a gateway, comprising:

[0035] A receiving unit is configured to receive a first request sent by a plurality of first devices. The first request is used to request services to be provided to the first devices. The first request includes first information and second information. The first information is used to identify the computing power service requested by the first devices, and the second information represents the computing power service needs of the first devices.

[0036] The aggregation unit is used to aggregate multiple first requests according to service type and based on multiple first pieces of information to obtain a first aggregation result;

[0037] The scheduling unit is used to select a computing power node to provide computing power services for each of the multiple first devices based on multiple second information, when the first aggregation result indicates that the types of computing power services requested by the multiple first devices are the same.

[0038] This application embodiment also provides a gateway, including: a processor and a communication interface; wherein,

[0039] The communication interface is used to receive first requests sent by multiple first devices. The first requests are used to request services to the first devices. The first requests include first information and second information. The first information is used to identify the computing power service requested by the first devices, and the second information represents the computing power service needs of the first devices.

[0040] The processor is configured to aggregate multiple first requests according to service type and based on multiple first pieces of information to obtain a first aggregation result; and when the first aggregation result indicates that the computing power service types requested by the multiple first devices are the same, to select a computing power node to provide computing power services for each of the multiple first devices based on multiple second pieces of information.

[0041] This application also provides a gateway, including: a processor and a memory for storing computer programs capable of running on the processor.

[0042] The processor is used to execute any of the above methods when running the computer program.

[0043] This application also provides a storage medium storing a computer program thereon, which, when executed by a processor, implements the operation of any of the above methods.

[0044] This application also provides a computer program product, including a computer program that, when executed by a processor, implements the operation of any of the above methods.

[0045] The scheduling method, apparatus, gateway, storage medium, and computer program product provided in this application embodiment include a first gateway receiving first requests sent by multiple first devices. Each first request requests services to be provided to a first device. The first request includes first information and second information. The first information identifies the computing power service requested by the first device, and the second information characterizes the computing power service requirements of the first device. Based on the service type and the first information, the multiple first requests are aggregated to obtain a first aggregation result. If the first aggregation result indicates that the computing power service types requested by the multiple first devices are the same, a computing power node is selected for each of the multiple first devices to provide computing power services based on the second information. The solution provided in this application embodiment aggregates the received first requests from multiple first devices according to service type. When the aggregation result indicates that the computing power service types corresponding to the multiple first requests are the same, the first gateway comprehensively considers the computing power service requirements corresponding to the multiple first requests when selecting a computing power node for each of the multiple first devices, thus achieving optimal scheduling. Attached Figure Description

[0046] Figure 1 is a flowchart illustrating the scheduling method according to an embodiment of this application;

[0047] Figure 2 is a flowchart illustrating the aggregation request process in an embodiment of this application;

[0048] Figure 3 shows a schematic diagram of the extended Border Gateway Protocol (BGP) message structure according to an embodiment of this application.

[0049] Figure 4 is a flowchart illustrating the concurrent computing power routing method for application examples in this application;

[0050] Figure 5 is a schematic diagram of the scheduling device structure according to an embodiment of this application;

[0051] Figure 6 is a schematic diagram of the gateway structure according to an embodiment of this application. Detailed Implementation

[0052] The present application will now be described in further detail with reference to the accompanying drawings and embodiments.

[0053] In related technologies, the gateway (specifically, the ingress gateway) in a computing power router (also known as a computing power router or a computing network joint router) can receive service requests sent by users (which can also be understood as computing power service requests, business requests, or computing power business requests), select appropriate computing power nodes for the services requested by users, and schedule the service requests sent by users to the selected computing power nodes according to appropriate network paths. The selected computing power nodes then provide computing power services to the users corresponding to the service requests.

[0054] In practical applications, the ingress gateway needs to know the computing power information (which can also be understood as computing power node information or computing power node status information) of each computing power node, as well as the network information between the computing power nodes and the ingress gateway. This allows the ingress gateway to select the optimal computing power node for a user's requested service, taking into account the computing power and network information of each node. Based on this, the computing power nodes need to announce (or report, send, or transmit) their own computing power information to the ingress gateway. Correspondingly, the ingress gateway can use the received computing power information to maintain the computing power and network information of each computing power node, enabling it to select the optimal computing power node for the user's requested service. In this process, in order to reduce the amount of information notification, computing power nodes usually use periodic and / or triggering methods to notify the ingress gateway of computing power information. The periodic method means that the computing power node sends computing power information to the ingress gateway according to a preset period, while the triggering method means that the computing power node triggers the sending of computing power information to the ingress gateway when specific conditions are met.

[0055] After obtaining the computing power and network information of the computing nodes, the ingress gateway, upon receiving a service request from a user, can utilize the corresponding computing power service requirements of the request, along with the maintained computing power and network information of the computing nodes, to select a computing node that can meet the computing power service requirements (or, in other words, select a computing power node whose computing power information matches the computing power service requirements) to provide computing power services to the user corresponding to the service request. In other words, the ingress gateway typically performs scheduling on a single service request basis (or, in other words, at the service request level).

[0056] However, in practical applications, when the ingress gateway receives a large number of service requests simultaneously, or when its processing performance is limited, or when it needs to meet the load balancing requirements of computing nodes, the ingress gateway may need to select appropriate computing nodes for services requested by multiple users simultaneously. This involves scheduling service requests from multiple users to achieve more efficient routing. This situation can also be referred to as request concurrency or batch processing of concurrent requests.

[0057] For example, when computing nodes announce their computing power information to the ingress gateway using a periodic and / or trigger-based method, the ingress gateway can only schedule based on the most recent (updated or synchronized) computing power information received between two receptions. Therefore, if the time interval between the computing nodes synchronizing their computing power information with the ingress gateway is long, the computing power information maintained by the ingress gateway may have a significant error compared to the actual computing power information in the computing nodes. This makes it difficult for the ingress gateway to schedule service requests to the optimal computing power node. In this case, the ingress gateway may need to uniformly schedule all service requests received since the last scheduling each time it receives computing power information to ensure the accuracy of the computing power information used for scheduling. Therefore, if multiple user service requests are received between two schedulings, the ingress gateway needs to schedule the service requests of multiple users simultaneously.

[0058] When the ingress gateway schedules service requests from multiple users simultaneously, if each user's service request corresponds to a different computing power service type, and different service types are provided by different computing power nodes, the ingress gateway can determine the computing power node that matches the computing power service type requested by each user. In other words, it can determine which computing power nodes or nodes can provide the computing power service requested by the user, and schedule the service request to the determined computing power node to achieve optimal scheduling.

[0059] However, if multiple users' service requests correspond to the same type of computing power service, that is, when multiple users request computing power services of the same type, the ingress gateway selects computing power nodes to provide computing power services for these users (i.e., when selecting computing power nodes for the computing power services requested by users). If only the computing power service requirements corresponding to a single service request are considered, and the optimal scheduling scheme (such as scheduling to a computing power node with a lower load) is selected for that service request, it may be difficult to satisfy other service requests, that is, it is difficult to achieve optimal scheduling.

[0060] For example, suppose three users need to access service A and initiate service requests at the same ingress gateway. This means the ingress gateway needs to schedule three service requests simultaneously: service request 1, service request 2, and service request 3. Regarding service A, based on the computing power information maintained by the ingress gateway, we know that the computing power nodes capable of providing service A include computing power node 1 and computing power node 2. Computing power node 1 and computing power node 2 have the same maximum load, and computing power node 1's current load is 30%, while computing power node 2's current load is 50%. If the computing power service requirements for each service request are: service request 1 requires 50% of the load of one computing power node, service request 2 requires 30% of the load of one computing power node, and service request 3 requires... If a computing node is operating at 30% load, and we only consider a single service request, scheduling the requests in the order of service request 1, service request 2, and service request 3, we can schedule service request 1 to the less loaded computing node 1. After scheduling, computing node 1's load will be 80%, and computing node 2's load will be 50%. Then, we can schedule service request 2 to the less loaded computing node 2. After scheduling, both computing nodes 1 and 2 will be at 80% load, which would result in the computing power requirement for service request 3 not being met. However, if the ingress gateway schedules service request 1 to computing node 2 and service requests 2 and 3 to computing node 1, it can simultaneously meet the computing power requirements for all service requests. Therefore, when the ingress gateway schedules multiple service requests, if we only consider the computing power requirement for a single service request, we cannot achieve optimal scheduling.

[0061] Based on this, in various embodiments of this application, the first gateway (such as the ingress gateway) aggregates the first requests sent by multiple first devices according to service type. When the aggregation result indicates that the computing power service types corresponding to multiple first requests are the same, the first gateway selects a computing power node to provide computing power services for each of the multiple first devices, taking into account the computing power service requirements corresponding to multiple first requests. In this way, optimal scheduling can be achieved.

[0062] This application provides a scheduling method applied to a first gateway, as shown in Figure 1. The method includes:

[0063] Operation 101: Receive a first request sent by a plurality of first devices. The first request is used to request services to be provided to the first devices. The first request includes first information and second information. The first information is used to identify the computing power service requested by the first devices, and the second information represents the computing power service needs of the first devices.

[0064] Operation 102: According to the service type and based on multiple pieces of first information, aggregate multiple first requests to obtain the first aggregation result;

[0065] Operation 103: When the first aggregation result indicates that the types of computing power services requested by the plurality of first devices are the same, based on multiple second pieces of information, a computing power node is selected for each of the plurality of first devices to provide computing power services, and the computing power node is used to provide computing power services at least.

[0066] In practical applications, the first gateway may specifically include the entry gateway of a computing power router. This embodiment does not limit the name of the first gateway, as long as its function is implemented. The first device can also be understood as a user equipment (UE), which may include mobile user equipment (such as a terminal, also called a terminal device) and fixed user equipment (such as a server). This embodiment does not limit the name of the first device, as long as its function is implemented.

[0067] In practical applications, when each of the first devices determines that it needs to request computing power nodes to provide computing power services, it can send the first request to the first gateway to request the first gateway to select a suitable computing power node (i.e., select a computing power node for the computing power service of the first device) to provide computing power services to the first device. The first request can also be called a service request, computing power service request, business request, or computing power business request; this application embodiment does not limit the name of the first request. The first request includes first information and second information. The first information can also be called service identification information, service type information, or business type information, etc.; this application embodiment does not limit the name of the first information. The second information can also be called service demand information, computing power service demand information, etc.; this application embodiment does not limit the name of the second information.

[0068] For each first request received from a first device, the first gateway can use the first information contained in the first request to determine the service type corresponding to the first request, which can also be understood as determining the computing power service type corresponding to the first device, or determining which type of computing power service the first device requests. Simultaneously, the first gateway can use the second information to determine (or estimate, predict, or forecast) the amount of computing power resources required for the computing power service corresponding to the first request.

[0069] After determining the service type corresponding to each first request, in operation 102, the first gateway can aggregate the first requests sent by multiple first devices according to the service type to obtain a first aggregation result, thereby using the first aggregation result to determine whether the computing power service types requested by multiple first devices are the same.

[0070] For example, as shown in Figure 2, suppose the first gateway receives three first requests sent by three first devices (as shown in the table on the left of Figure 2). The three first devices are device 1, device 2, and device 3, and the service type corresponding to the first requests sent by device 1, device 2, and device 3 is service A. At this time, the first gateway can aggregate the three first requests and obtain the aggregation result (which can be presented in the form of an information table, as shown in the table on the right of Figure 2).

[0071] In practical applications, when the first aggregation result indicates that the computing power service type requested by the multiple first devices is the same, that is, when multiple first devices request to provide computing power services of the same type, in operation 103, when the first gateway selects a computing power node to provide computing power services for each of the multiple first devices, it comprehensively considers the computing power service needs of the multiple first devices, thus achieving optimal scheduling.

[0072] In practical applications, optimal scheduling can be reflected in several aspects, such as providing computing power services to as many first devices as possible, optimizing the global energy consumption of the computing network system, and ensuring optimal fairness among computing nodes. Based on this, when scheduling the first request, the first gateway can select computing nodes to provide computing power services to each of the multiple first devices according to different strategies (which can also be understood as selection strategies), so as to achieve optimal scheduling for different aspects.

[0073] Specifically, when the goal is to provide computing power services to as many first devices as possible, when the first device selects a computing power node to provide computing power services for each of the multiple first devices, it can comprehensively consider the status information of the multiple computing power nodes and the computing power service requirements of each of the multiple first devices, so that the scheduling scheme can satisfy the first request of each first device as much as possible.

[0074] Based on this, in one embodiment, selecting a computing power node to provide computing power services for each of the plurality of first devices includes:

[0075] Using third information and multiple second information, a computing power node is selected for each of the multiple first devices to provide computing power services. The third information includes the status information of multiple computing power nodes.

[0076] Here, the third information may specifically include the identification information of each computing node among the plurality of computing nodes, the status information of each computing node, the types of computing services that each computing node can provide, and the network information between each computing node and the first gateway. The status information may also be referred to as computing power information or computing node information, and may specifically include load information (such as load occupancy rate), the number of times service has been provided, etc.

[0077] In practical applications, the first gateway needs to obtain the third information in advance. That is, in one embodiment, the method may further include:

[0078] Obtain the third information.

[0079] In practical applications, each computing node can send status information to the first gateway according to a preset period, and / or, when a computing node meets specific conditions, it sends status information to the first gateway. Accordingly, the first gateway can use the latest status information received from each computing node to determine the third information.

[0080] Based on this, in one embodiment, obtaining the third information includes:

[0081] Receive status information sent by multiple computing power nodes;

[0082] The latest received status information is used as the third information.

[0083] In practical applications, the first gateway can store the latest status information sent by each computing node locally as the third information, thereby maintaining (or updating / synchronizing) the status information of each computing node within the first gateway. Specifically, the third information may include a status information table aggregated based on the types of computing services that the computing nodes can provide.

[0084] After acquiring and storing the third information, the first gateway can read the status information of the computing power nodes related to the computing power service type requested by the plurality of first devices, i.e., the status information of the computing power nodes capable of providing the computing power service of that type. The first gateway can then use the read status information and the plurality of second information to select a computing power node to provide computing power services for each of the plurality of first devices, in order to satisfy the first requests of the plurality of first devices as much as possible. Specifically, the first gateway can use the first aggregation result and the status information table corresponding to the third information to perform matching and mapping, thereby determining which computing power nodes(s) in the status information table can provide the computing power service requested by the plurality of first devices, and selecting a computing power node to provide computing power services for each of the plurality of first devices.

[0085] In practical applications, the first gateway may not be able to simultaneously satisfy the first requests of all first devices. For example, when the available computing power resources of the computing power nodes are limited, the first gateway can only select computing power nodes to provide computing power services for some of the multiple first devices. Only after the computing power nodes complete their tasks (such as providing computing power services) and their computing power resources are released can it select computing power nodes to provide computing power services for the remaining first devices. In this case, the first gateway can determine which first devices to prioritize for receiving computing power services based on the priority order of the multiple first devices, thereby prioritizing the selection of computing power nodes to provide computing power services to the first devices that need to provide services as quickly as possible, achieving optimal scheduling order.

[0086] Based on this, in one embodiment, selecting a computing power node to provide computing power services for each of the plurality of first devices includes:

[0087] Using the plurality of second information, a fourth piece of information is determined, wherein the fourth piece of information represents the priority of the plurality of first devices;

[0088] Using the fourth and third information, a computing power node is selected to provide computing power services for each of the multiple first devices.

[0089] Here, the first device priority refers to the priority of selecting computing power nodes to provide computing power services for the first device, and can also be understood as the priority of scheduling the first request sent by the first device.

[0090] In practical applications, the first gateway can use the second information to determine other demand information for each first device, such as whether the first device needs service as soon as possible or whether the first device can accept delayed service, and can record the determined other demand information in the aggregation result.

[0091] After determining the other requirements information for each first device, the first gateway can prioritize the first devices based on this information to obtain the fourth information. Specifically, if the other requirements information of a first device indicates that the first device can receive delayed services, the priority of the first device is reduced; if the other requirements information of a first device indicates that the first device needs to provide services as soon as possible, the priority of the first device is increased.

[0092] For example, based on the above example, as shown in Table 1, assuming that the first gateway determines that device 1 can accept delayed service based on the second information corresponding to device 1; at the same time, the first gateway determines that device 2 needs to be served as soon as possible based on the second information corresponding to device 2; at the same time, the first gateway determines that device 3 has no other needs. At this time, the first gateway can determine the fourth information, which represents the priority order of the three first devices in descending order as device 2, device 3, and device 1.

[0093] Table 1

[0094] After determining the fourth piece of information, the first gateway can select computing power nodes to provide computing power services for each first device in descending order of priority. In this way, computing power nodes that provide computing power services to first devices that need to be served as quickly as possible can be selected first, meaning that the optimal scheduling order is achieved by considering the other needs of each first device.

[0095] Accordingly, when the goal is to achieve optimal global energy consumption of the computing power network system, when the first device selects computing power nodes to provide computing power services for each of the multiple first devices, it can centrally select computing power nodes for multiple first devices. In this way, the centrally selected computing power nodes provide computing power services for all first devices, which can reduce the number of computing power nodes that need to provide computing power services (which can also be understood as being in a working state). This allows more computing power nodes to enter a dormant state without providing computing power services, thereby reducing energy consumption and achieving energy-saving optimization of the computing power network system.

[0096] Specifically, the first gateway can obtain the third information and use it to determine the load information of each computing node; the first gateway can then use the load information of each computing node and the second information corresponding to the multiple first devices to select a computing node to provide computing power services for each of the multiple first devices. The load information may specifically include the amount of computing power resources used and / or the load occupancy rate of the computing node.

[0097] Based on this, in one embodiment, the status information of the computing power node includes the load information of the computing power node, and the step of selecting a computing power node to provide computing power services for each of the plurality of first devices includes:

[0098] Using the load information and multiple second information, one or more (one or more can also be understood as at least one) computing power nodes whose load information is greater than the first threshold are selected to provide computing power services for each of the multiple first devices.

[0099] In practical application, the first gateway can identify one or more computing nodes with load information exceeding a first preset threshold as heavily loaded computing nodes (which can also be understood as computing nodes with high load occupancy or low available computing resources). It then determines whether the remaining computing resources (available computing resources or available computing resources) among these heavily loaded nodes can meet the computing service needs of the multiple first devices. If the needs are met, the first gateway selects one or more heavily loaded computing nodes to provide computing services to each of the multiple first devices. In this way, lightly loaded computing nodes (i.e., computing nodes with load information less than or equal to the first preset threshold) can enter a sleep mode after completing their current task (such as providing computing services), thereby saving energy. The first preset threshold can be set according to actual needs.

[0100] For example, as shown in Table 2, assuming that the three computing nodes Node 1, Node 2, and Node 3 can all provide computing power services of service type A, where the load information of Node 1 is 60%, the load information of Node 2 is 10%, the load information of Node 3 is 20%, and the preset first threshold is 50%, then the first gateway can determine that Node 1 is a heavily loaded computing power node, and select Node 1 as much as possible to provide computing power services to multiple first devices, so that Node 2 and Node 3 can enter sleep mode after completing their tasks.

[0101] Table 2

[0102] In practical applications, when a lightly loaded computing node enters sleep mode, it can notify the first gateway of its sleep status via extended BGP. Simultaneously, during subsequent scheduling, the first gateway can determine (or predict) whether to wake up the sleeping computing node based on the request distribution, and select the woken-up computing node to provide computing power services to the device corresponding to the request. Here, the request distribution can specifically be related to parameters such as the number of requests and the service type corresponding to the requests.

[0103] As can be seen from the above description, centrally scheduling multiple first requests to heavily loaded computing nodes through the first gateway can improve the overall resource efficiency of the computing network system and achieve optimal global energy consumption. Furthermore, compared to existing technologies that migrate ongoing tasks from some computing nodes to other computing nodes so that these nodes can enter a dormant state to reduce energy consumption, this approach reduces the number of computing nodes needing to provide computing services during the request scheduling phase. This avoids interrupting ongoing business processes on computing nodes and provides a better service experience.

[0104] Accordingly, in order to achieve optimal fairness among computing power nodes, when the first device selects a computing power node to provide computing power services for each of the multiple first devices, it can give priority to selecting computing power nodes that provide services less frequently. In this way, the number of times each computing power node provides computing power services can be balanced, thereby making more effective use of the resources of the entire computing power network system and improving resource utilization efficiency.

[0105] Specifically, when multiple computing power nodes are capable of providing computing power services to the multiple first devices, the first gateway can utilize the third information to determine the number of times each computing power node provides services, and select a computing power node to provide computing power services to each of the multiple first devices based on the number of times each computing power node provides services. The number of times services are provided can specifically include the number of times services are provided within a preset time range, which can be set according to actual needs.

[0106] Based on this, in one embodiment, the status information of the computing power node includes the number of times the computing power node provides services, and the step of selecting a computing power node to provide computing power services for each of the plurality of first devices includes:

[0107] Using the number of service provision times and multiple pieces of second information, one or more computing power nodes with a number of service provision times less than a second threshold are selected to provide computing power services for each of the multiple first devices.

[0108] In practical applications, the first gateway can determine the fairness value (or fairness parameter) of each computing node by utilizing the number of times each computing node provides services. The fairness value characterizes the number of service provision times. For example, assuming three computing nodes capable of providing computing services to the plurality of first devices are node 1, node 2, and node 3, and node 1 provides services 70 times, node 2 provides services 20 times, and node 3 provides services 10 times, the first gateway can determine the fairness value of each computing node by using the ratio of the number of service provision times of each computing node to the total number of service provision times. That is, the fairness value of node 1 is 0.7, the fairness value of node 2 is 0.2, and the fairness value of node 3 is 0.1.

[0109] In practical applications, when multiple gateways are scheduling requests, since the fairness value of a computing power node is a unified attribute of the computing power node (i.e., each computing power node corresponds to a fairness value, and the fairness value of each computing power node is unique), the first gateway can synchronize (or share, update, etc.) the fairness value of the computing power node with other gateways among the multiple gateways.

[0110] Specifically, the first gateway can periodically synchronize the fairness value of computing power nodes with the other gateways based on extended BGP. For example, as shown in Figure 5, the Path Attributes field in the BGP message is configured with a triple for each computing power node. The triple includes an attribute type (Attr.TYPE) field, an attribute length (Attr.Length) field, and an attribute value (Attr.Value) field. The Attr.TYPE field can specifically include an attribute flag (Attr.Flags) field and an attribute type code (Attr.Type Code) field. Here, the Attr.Flags field is 7 bits long, with the lower four bits being unused bits (U in Figure 5). The default value for unused bits is 0. Therefore, the fifth bit of the Attr.Flags field can be set to 1 to indicate that the value in Attr.Value is the fairness value of that computing power node, and the fairness value of that computing power node is then filled into the Attr.Value field.

[0111] In practical applications, the first gateway can designate one or more computing power nodes with fewer service provision times than a second threshold (which can also be understood as a fairness value less than the second threshold when the service provision times are represented by a fairness value) as computing power nodes with fewer service provision times, and select one or more computing power nodes with fewer service provision times to provide computing power services to each of the plurality of first devices. This increases the number of times computing power nodes with fewer service provision times provide computing power services, thereby more effectively utilizing the resources of the entire computing power network system and improving resource utilization efficiency. The second preset threshold can be set according to actual needs.

[0112] Of course, the first gateway can prioritize each computing node based on the number of times it provides services, and obtain a ranking result. The computing node with fewer service provision times in the ranking result has a higher priority. The first gateway can then use the ranking result to select computing nodes to provide computing services to each of the multiple first devices in descending order of priority.

[0113] Based on this, in one embodiment, selecting a computing power node to provide computing power services for each of the plurality of first devices includes:

[0114] The priority of the computing power nodes is determined by the number of times the service is provided; the priority of the computing power nodes is used to select a computing power node to provide computing power services for each of the multiple first devices.

[0115] In practical applications, the first gateway may also receive requests whose service type is different from the service type corresponding to the first request. In this case, the first gateway can aggregate all requests according to the service type and based on the computing power service identifier of all received requests to obtain an aggregation result. The first gateway can then select a suitable computing power node to provide computing power services of that service type to one or more devices corresponding to each service type in the aggregation result.

[0116] Based on this, in one embodiment, the method may further include:

[0117] The system receives a second request from one or more second devices. The second request is used to request services to the second devices. The second request includes fifth information and sixth information. The fifth information is used to identify the computing power service requested by the second devices, and the sixth information represents the computing power service requirements of the second devices. The service type corresponding to the second request is different from the service type corresponding to the first request.

[0118] According to the service type, and based on multiple first pieces of information and one or more fifth pieces of information, multiple first requests and one or more second requests are aggregated to obtain a second aggregation result; the second aggregation result contains at least two service types;

[0119] For each of the at least two service types, based on the device's computing power service requirements, select computing power nodes to provide computing power services to the device; where,

[0120] When selecting computing power nodes to provide computing power services for a device, it is necessary to combine the seventh information to select computing power nodes to provide computing power services for the device. The seventh information represents the correspondence between service type and service priority.

[0121] Here, the seventh piece of information can also be referred to as service priority information, which can be predefined (or pre-configured) in the first gateway. The first gateway can use the seventh piece of information to determine the priority order of computing power services of different service types. The computing power service of a service type with a higher priority order needs to be provided first (i.e., priority scheduling of requests corresponding to that service type, or priority provision of computing power services to devices corresponding to that service type).

[0122] In practical applications, the first gateway receives requests for two service types (i.e., the first request and the second request) and aggregates the received requests based on the service type. Thus, the first gateway can determine which devices correspond to each service type of computing power service (i.e., which devices requested the computing power service of that service type) based on the aggregation result. Therefore, the first gateway can select computing power nodes to provide computing power services for the devices corresponding to each service type.

[0123] In practical applications, when the first gateway cannot simultaneously meet the requests of all devices, the first gateway can select computing power nodes to provide computing power services for the devices corresponding to each service type according to the priority order of computing power services of different service types. In this way, it can ensure that services are provided to terminals corresponding to service types with high service priority.

[0124] For example, as shown in Table 3, assume the first gateway receives requests from devices 1, 2, and 3, where devices 1 and 2 are the first devices, requesting computing power services of service type A; and device 3 is the second device, requesting computing power services of service type B. In this case, based on the seventh information, the first gateway can determine that service B is a high-priority service and needs to be provided first. Simultaneously, using the request from device 1, the first gateway determines that device 1 can accept delayed service; and using the request from device 2, the first gateway determines that device 2 needs to provide service as soon as possible. Therefore, the first gateway can determine to first select a computing power node to provide computing power services for the device corresponding to service B (i.e., device 3), and then select a computing power node to provide computing power services for the devices corresponding to service A (i.e., devices 1 and 2). Therefore, the order of selecting computing power nodes to provide computing power services for the three devices is device 3, device 2, and device 1.

[0125] Table 3

[0126] The scheduling method provided in this application embodiment involves a first gateway receiving first requests from multiple first devices. Each first request requests services to a first device. The first request includes first information and second information. The first information identifies the computing power service requested by the first device, and the second information characterizes the computing power service requirements of the first device. The first gateway aggregates the multiple first requests according to service type and based on the first information to obtain a first aggregation result. If the first aggregation result indicates that the computing power service types requested by the multiple first devices are the same, a computing power node is selected for each of the multiple first devices to provide computing power services based on the second information. The solution provided in this application embodiment aggregates the received first requests from multiple first devices according to service type. When the aggregation result indicates that the computing power service types corresponding to the multiple first requests are the same, the first gateway selects a computing power node to provide computing power services for each of the multiple first devices, comprehensively considering the computing power service requirements corresponding to the multiple first requests. This achieves optimal scheduling.

[0127] The following section provides a more detailed description of this application with reference to application examples.

[0128] This application example provides a concurrent computing power routing system, including multiple UEs, a computing network-connected router, multiple computing power nodes (which can also be understood as servers providing computing power services), and intermediate nodes (specifically, intermediate routers) between the computing power router and the computing power nodes. The computing network-connected router can specifically include an ingress gateway (i.e., the first gateway mentioned above) and an egress gateway (also known as the second gateway). Typically, the ingress gateway can receive service requests (i.e., the first requests mentioned above) sent by multiple UEs and schedule multiple service requests to appropriate computing power nodes through the corresponding intermediate nodes. That is, each UE corresponding to a service request selects a computing power node that provides computing power services.

[0129] Based on the above system, this application example provides a method for concurrent computing power routing, as shown in Figure 4, which includes the following operations:

[0130] Operation 401: Multiple UEs (i.e., the first device mentioned above) send service requests to the ingress gateway;

[0131] Here, the service request includes service identification information (i.e., the first information mentioned above) and service demand information (i.e., the second information mentioned above).

[0132] Operation 402: The ingress gateway receives service requests from multiple UEs and aggregates the received service requests according to service type to obtain the aggregation result;

[0133] Operation 403: For each service type, the ingress gateway determines the computing power requirement information for each request corresponding to the service type, as well as the total computing power requirement information for all requests of the service type (which may include the total computing power resource amount corresponding to the service type).

[0134] Operation 404: For each service type, the ingress gateway combines different selection strategies and comprehensively considers computing power demand information and computing power node status information to select computing power nodes that provide computing power services for multiple UEs;

[0135] In practical applications, the selection strategy may include strategies such as serving as many users as possible, optimizing global energy consumption, and ensuring optimal fairness.

[0136] Specifically, when the selection strategy includes serving as many users as possible, the ingress gateway may consider the user priority carried in concurrent service requests and / or the priority of service types pre-configured in the ingress gateway when selecting computing nodes to provide computing services for multiple UEs.

[0137] When the selection strategy includes global energy optimization, the ingress gateway may consider centrally scheduling service requests to several heavily loaded computing nodes, provided that the service needs can be met, so that the lightly loaded computing nodes can enter a dormant state, thereby reducing energy consumption.

[0138] When the selection strategy includes optimal fairness, the ingress gateway can count the number of times each computing power node provides services and determine the fairness value of the computing power node. This allows the fairness value of each computing power node to be considered when selecting computing power nodes to provide computing power services for multiple UEs, thereby making more effective use of the resources of the entire computing power network system and improving resource utilization efficiency.

[0139] Operation 405: The ingress gateway generates a scheduling policy based on the selected computing power node and schedules multiple service requests to the selected computing power node.

[0140] The solution provided in this application example, when selecting computing nodes to provide computing power services for multiple concurrent service requests, simultaneously considers the computing power requirements of multiple service requests, the status information of computing nodes, the priority of the UE, the energy consumption information of computing nodes, and the fairness information of computing nodes. In this way, by aggregating and comprehensively considering multiple concurrent requests, optimal scheduling can be achieved; at the same time, by batch scheduling multiple concurrent service requests, the computing power service requirements of multiple service requests on the UE side can be fully considered, enabling more efficient service request scheduling; furthermore, considering multiple factors such as priority, energy consumption, and fairness when scheduling service requests can improve the scheduling efficiency of service requests.

[0141] To implement the method of this application embodiment, this application embodiment also provides a scheduling device, which is installed on a gateway, as shown in FIG5. The device includes:

[0142] The receiving unit 501 is configured to receive a first request sent by a plurality of first devices. The first request is used to request services to be provided to the first devices. The first request includes first information and second information. The first information is used to identify the computing power service requested by the first devices, and the second information represents the computing power service needs of the first devices.

[0143] The aggregation unit 502 is used to aggregate multiple first requests according to service type and based on multiple first pieces of information to obtain a first aggregation result;

[0144] The scheduling unit 503 is used to select a computing power node to provide computing power services for each of the multiple first devices based on multiple second information, when the first aggregation result indicates that the types of computing power services requested by the multiple first devices are the same.

[0145] In one embodiment, the scheduling unit 503 is specifically used for:

[0146] Using third information and multiple second information, a computing power node is selected for each of the multiple first devices to provide computing power services. The third information includes the status information of multiple computing power nodes.

[0147] In one embodiment, the scheduling unit 503 is specifically used for:

[0148] Using the plurality of second information, a fourth piece of information is determined, wherein the fourth piece of information represents the priority of the plurality of first devices;

[0149] Using the fourth and third information, a computing power node is selected to provide computing power services for each of the multiple first devices.

[0150] In one embodiment, the status information of the computing node includes the load information of the computing node, and the scheduling unit 503 is specifically used for:

[0151] Using the load information and multiple second information, one or more computing power nodes with load information greater than a first threshold are selected to provide computing power services for each of the multiple first devices.

[0152] In one embodiment, the status information of the computing node includes the number of times the computing node provides services, and the scheduling unit 503 is specifically used for:

[0153] Using the number of service provision times and multiple pieces of second information, one or more computing power nodes with a number of service provision times less than a second threshold are selected to provide computing power services for each of the multiple first devices.

[0154] or,

[0155] The priority of the computing power nodes is determined by the number of times the service is provided; the priority of the computing power nodes is used to select a computing power node to provide computing power services for each of the multiple first devices.

[0156] In one embodiment, the scheduling unit 503 is further configured to:

[0157] Obtain the third information.

[0158] In one embodiment, the scheduling unit 503 is specifically used for:

[0159] Receive status information sent by multiple computing power nodes;

[0160] The latest received status information is used as the third information.

[0161] In one embodiment, the receiving unit 501 is further configured to:

[0162] The system receives a second request from one or more second devices. The second request is used to request services to the second devices. The second request includes fifth information and sixth information. The fifth information is used to identify the computing power service requested by the second devices, and the sixth information represents the computing power service requirements of the second devices. The service type corresponding to the second request is different from the service type corresponding to the first request.

[0163] The aggregation unit 502 is further configured to aggregate multiple first requests and one or more second requests according to service type and based on multiple first pieces of information and one or more fifth pieces of information to obtain a second aggregation result; the second aggregation result includes at least two service types;

[0164] The scheduling unit 503 is further configured to, for each of the at least two service types of devices, select a computing power node to provide computing power services for the device based on the device's computing power service requirements; and when selecting a computing power node to provide computing power services for the device, combine the seventh information to select a computing power node to provide computing power services for the device, wherein the seventh information represents the correspondence between service type and service priority.

[0165] In practical applications, the receiving unit 501 can be implemented by the communication interface in the scheduling device, the aggregation unit 502 can be implemented by the processor in the scheduling device, and the scheduling unit 503 can be implemented by the processor in the scheduling device in combination with the communication interface.

[0166] It should be noted that the scheduling device provided in the above embodiments is only illustrated by the division of the above-described program units. In practical applications, the above processing can be assigned to different program units as needed, that is, the internal structure of the device can be divided into different program units to complete all or part of the processing described above. In addition, the scheduling device and scheduling method embodiments provided in the above embodiments belong to the same concept, and their specific implementation process can be found in the method embodiments, which will not be repeated here.

[0167] Based on the hardware implementation of the above program modules, and in order to implement the method of this application embodiment, this application embodiment also provides a gateway, as shown in FIG6, the gateway 600 includes:

[0168] The communication interface 601 enables information exchange with other devices;

[0169] The processor 602 is connected to the communication interface 601 to enable information interaction with other devices and to execute the methods provided by one or more of the above-mentioned technical solutions when running a computer program;

[0170] The computer program is stored in memory 603.

[0171] Specifically, the communication interface 601 is used for:

[0172] Receive a first request sent by a plurality of first devices, the first request being used to request services to the first devices, the first request including first information and second information, the first information being used to identify the computing power service requested by the first devices, and the second information representing the computing power service needs of the first devices;

[0173] The processor 602 is used for:

[0174] According to the service type and based on multiple first information, multiple first requests are aggregated to obtain a first aggregation result; and when the first aggregation result indicates that the computing power service types requested by the multiple first devices are the same, a computing power node that provides computing power services is selected for each of the multiple first devices based on multiple second information.

[0175] In one embodiment, the processor 602 is specifically used for:

[0176] Using third information and multiple second information, a computing power node is selected for each of the multiple first devices to provide computing power services. The third information includes the status information of multiple computing power nodes.

[0177] In one embodiment, the processor 602 is specifically used for:

[0178] Using third information and multiple second information, a computing power node is selected for each of the multiple first devices to provide computing power services. The third information includes the status information of multiple computing power nodes.

[0179] In one embodiment, the processor 602 is specifically used for:

[0180] Using the plurality of second information, a fourth piece of information is determined, wherein the fourth piece of information represents the priority of the plurality of first devices;

[0181] Using the fourth and third information, a computing power node is selected to provide computing power services for each of the multiple first devices.

[0182] In one embodiment, the status information of the computing node includes the load information of the computing node, and the processor 602 is specifically used for:

[0183] Using the load information and multiple second information, one or more computing power nodes with load information greater than a first threshold are selected to provide computing power services for each of the multiple first devices.

[0184] In one embodiment, the status information of the computing node includes the number of times the computing node provides services, and the processor 602 is specifically used for:

[0185] Using the number of service provision times and multiple pieces of second information, one or more computing power nodes with a number of service provision times less than a second threshold are selected to provide computing power services for each of the multiple first devices.

[0186] or,

[0187] The priority of the computing power nodes is determined by the number of times the service is provided; the priority of the computing power nodes is used to select a computing power node to provide computing power services for each of the multiple first devices.

[0188] In one embodiment, the processor 602 is further configured to:

[0189] The third information is obtained by combining the communication interface 601.

[0190] In one embodiment, the processor 602 is specifically used for:

[0191] Combined with the communication interface 601, status information sent by multiple computing nodes is received;

[0192] The latest received status information is used as the third information.

[0193] In one embodiment, the communication interface 601 is used for:

[0194] The system receives a second request from one or more second devices. The second request is used to request services to the second devices. The second request includes fifth information and sixth information. The fifth information is used to identify the computing power service requested by the second devices, and the sixth information represents the computing power service requirements of the second devices. The service type corresponding to the second request is different from the service type corresponding to the first request.

[0195] The processor 602 is used for:

[0196] According to the service type, and based on multiple first pieces of information and one or more fifth pieces of information, multiple first requests and one or more second requests are aggregated to obtain a second aggregation result; the second aggregation result includes at least two service types; for each service type of device, based on the device's computing power service requirements, a computing power node is selected to provide computing power services for the device; and when selecting a computing power node to provide computing power services for the device, a seventh piece of information is combined to select a computing power node to provide computing power services for the device, wherein the seventh piece of information represents the correspondence between service type and service priority.

[0197] It should be noted that the specific processing procedures of the processor 602 and the communication interface 601 can be understood with reference to the above method.

[0198] Of course, in practical applications, the various components in gateway 600 are coupled together through bus system 604. It can be understood that bus system 604 is used to implement communication between these components. In addition to the data bus, bus system 604 also includes a power bus, a control bus, and a status signal bus. However, for clarity, all buses are labeled as bus system 604 in Figure 6.

[0199] The memory 603 in this embodiment is used to store various types of data to support the operation of the gateway 600. Examples of such data include any computer program used to operate on the gateway 600.

[0200] The methods disclosed in the embodiments of this application can be applied to the processor 602, or implemented by the processor 602. The processor 602 may be an integrated circuit chip with signal processing capabilities. In the implementation process, each operation of the above methods can be completed by the integrated logic circuit of the hardware in the processor 602 or by instructions in the form of software. The processor 602 may be a general-purpose processor, a digital signal processor (DSP), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The processor 602 can implement or execute the methods, operations, and logic block diagrams disclosed in the embodiments of this application. The general-purpose processor may be a microprocessor or any conventional processor, etc. The operation of the methods disclosed in the embodiments of this application can be directly reflected as being executed by a hardware decoding processor, or being executed by a combination of hardware and software modules in the decoding processor. The software modules may be located in a storage medium, which is located in the memory 603. The processor 602 reads the information in the memory 603 and combines its hardware to complete the operation of the aforementioned methods.

[0201] In an exemplary embodiment, the gateway 600 may be implemented by one or more application-specific integrated circuits (ASICs), DSPs, programmable logic devices (PLDs), complex programmable logic devices (CPLDs), field-programmable gate arrays (FPGAs), general-purpose processors, controllers, microcontrollers (MCUs), microprocessors, or other electronic components to perform the aforementioned method.

[0202] It is understood that the memory (memory 603) in this embodiment of the application can be volatile memory or non-volatile memory, or both. Non-volatile memory can be read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), ferromagnetic random access memory (FRAM), flash memory, magnetic surface memory, optical disc, or compact disc read-only memory (CD-ROM); magnetic surface memory can be disk storage or magnetic tape storage. Volatile memory can be random access memory (RAM), which is used as an external cache. By way of example, but not limitation, many forms of RAM are available, such as Static Random Access Memory (SRAM), Synchronous Static Random Access Memory (SSRAM), Dynamic Random Access Memory (DRAM), Synchronous Dynamic Random Access Memory (SDRAM), Double Data Rate Synchronous Dynamic Random Access Memory (DDRSDRAM), Enhanced Synchronous Dynamic Random Access Memory (ESDRAM), SyncLink Dynamic Random Access Memory (SLDRAM), and Direct Rambus Random Access Memory (DRRAM).The memories described in the embodiments of this application are intended to include, but are not limited to, these and any other suitable types of memories.

[0203] In an exemplary embodiment, this application also provides a storage medium, namely a computer storage medium, specifically a computer-readable storage medium, such as a memory 603 storing a computer program, which can be executed by the processor 602 of the gateway 600 to perform the operations described in the aforementioned method. The computer-readable storage medium may be a memory such as FRAM, ROM, PROM, EPROM, EEPROM, Flash Memory, magnetic surface memory, optical disc, or CD-ROM.

[0204] In an exemplary embodiment, this application also provides a computer program product, including a computer program that can be executed by the processor 602 of the gateway 600 to perform the operations described in the aforementioned method.

[0205] It should be noted that terms such as "first" and "second" are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence.

[0206] Furthermore, the technical solutions described in the embodiments of this application can be combined arbitrarily without conflict.

[0207] The above description is only a part of the embodiments of this application and is not intended to limit the scope of protection of this application.

Claims

1. A scheduling method, wherein, The application is applied to a first gateway, comprising: receiving a plurality of first devices sending a first request, the first request is used to request to provide service for the first device, the first request contains first information and second information, the first information is used to identify the computing power service requested by the first device, and the second information represents the computing power service demand of the first device; According to the service type, and based on a plurality of the first information, a plurality of the first request is converged to obtain a first convergence result; In the case where the first convergence result represents that the computing power service types requested by the plurality of first devices are the same, based on a plurality of the second information, the computing power node providing the computing power service is selected for each of the plurality of first devices.

2. The method of claim 1, wherein, The computing power node providing the computing power service for each of the plurality of first devices includes: Using third information and a plurality of the second information, the computing power node providing the computing power service for each of the plurality of first devices is selected, and the third information contains state information of a plurality of computing power nodes.

3. The method of claim 2, wherein, The computing power node providing the computing power service for each of the plurality of first devices includes: Using the plurality of second information, fourth information is determined, and the fourth information represents the priority of the plurality of first devices; Using the fourth information and the third information, the computing power node providing the computing power service for each of the plurality of first devices is selected.

4. The method of claim 3, wherein, The fourth information is determined by using the plurality of second information, comprising: For each of the plurality of first devices, the other demand information of each of the plurality of first devices is determined by using the second information; the other demand information includes: needing to be served as soon as possible and acceptable delay service; According to the other demand information of each of the plurality of first devices, the priority of each of the plurality of first devices is sorted, and the fourth information is determined.

5. The method of claim 4, wherein, After the fourth information is determined, the scheduling method comprises: According to the priority of each of the plurality of first devices from high to low, the computing power node providing the computing power service for each of the plurality of first devices is selected.

6. The method of claim 4, wherein, The scheduling method comprises: When the computing power node providing the computing power service for each of the plurality of first devices is selected, the computing power node is selected for the plurality of first devices.

7. The method of claim 2, wherein, The state information of the computing power node contains the load information of the computing power node, and the computing power node providing the computing power service for each of the plurality of first devices includes: Using the load information and a plurality of the second information, one or more computing power nodes with load information greater than a first threshold value are selected to provide service for the computing power service of each of the plurality of first devices.

8. The method of claim 7, wherein, The load information includes the used computing power resource amount and / or load occupancy rate of the computing power node.

9. The method of claim 2, wherein, The state information of the computing power node includes the service providing times of the computing power node, and the computing power node providing the computing power service for each of the plurality of first devices comprises: Using the service providing times and a plurality of the second information, one or more computing power nodes with service providing times less than a second threshold are selected to provide services for the computing power service of each of the plurality of first devices; Or, Using the service providing times, the priority of the computing power node is determined, and the computing power node providing the computing power service for each of the plurality of first devices is selected using the priority of the computing power node.

10. The method of claim 2, wherein, The method further comprises: Obtaining the third information.

11. The method of claim 10, wherein, The obtaining of the third information comprises: Receiving the state information sent by the plurality of computing power nodes; Taking the plurality of latest received state information as the third information.

12. The method according to any one of claims 1 to 11, wherein, The method further comprises: Receiving a second request sent by one or more second devices, the second request being used to request to provide services for the second device, the second request comprising fifth information and sixth information, the fifth information being used to identify the computing power service requested by the second device, and the sixth information representing the computing power service demand of the second device; the service type corresponding to the second request is different from the service type corresponding to the first request; According to the service type and based on a plurality of the first information and one or more of the fifth information, a plurality of the first request and one or more of the second request are aggregated to obtain a second aggregation result; the second aggregation result comprises at least two service types; For each device of each service type in the at least two service types, a computing power node providing a computing power service for the device is selected based on the computing power service demand of the device; wherein When the computing power node providing the computing power service for the device is selected, the method further comprises: The computing power node providing the computing power service for the device is selected in combination with seventh information representing the correspondence between the service type and the service priority.

13. A scheduling apparatus, wherein, The gateway comprises: A receiving unit configured to receive a first request sent by a plurality of first devices, the first request being used to request to provide services for the first device, the first request comprising first information and second information, the first information being used to identify the computing power service requested by the first device, and the second information representing the computing power service demand of the first device; An aggregation unit configured to aggregate a plurality of the first request according to the service type and based on a plurality of the first information to obtain a first aggregation result; A scheduling unit configured to, in a case where the first aggregation result represents that the computing power service types requested by the plurality of first devices are the same, select a computing power node providing the computing power service for each of the plurality of first devices based on a plurality of the second information.

14. A gateway, wherein, Comprise: A processor and a communication interface; wherein The communication interface is configured to receive a plurality of first requests sent by a plurality of first devices, the first requests being used to request to provide the first devices with services, the first requests comprising first information and second information, the first information being used to identify computing power service types requested by the first devices, and the second information representing computing power service requirements of the first devices. The processor is configured to: aggregate the plurality of first requests according to service types and based on the plurality of first information to obtain a first aggregation result; and in a case where the first aggregation result represents that the computing power service types requested by the plurality of first devices are the same, select, based on the plurality of second information, a computing power node providing the computing power service for each of the plurality of first devices.

15. The gateway of claim 14, wherein, The processor is configured to: select, based on the third information and the plurality of second information, the computing power node providing the computing power service for each of the plurality of first devices, the third information comprising state information of a plurality of computing power nodes.

16. The gateway of claim 15, wherein, The processor is configured to: determine fourth information representing priorities of the plurality of first devices based on the plurality of second information; and select, based on the fourth information and the third information, the computing power node providing the computing power service for each of the plurality of first devices.

17. The gateway of claim 15, wherein, The state information of the computing power node comprises load information of the computing power node, and the processor is configured to: select, based on the load information and the plurality of second information, one or more computing power nodes with load information greater than a first threshold value to provide the computing power service for each of the plurality of first devices.

18. A gateway, wherein, comprise: a processor and a memory for storing a computer program capable of running on the processor, wherein the processor is configured to execute the computer program to perform the operations of the method in any one of claims 1 to 12.

19. A storage medium having stored thereon a computer program, wherein, The computer program is executed by the processor to perform the operations of the method in any one of claims 1 to 12.

20. A computer program product comprising a computer program, wherein, The computer program is executed by the processor to perform the operations of the method in any one of claims 1 to 12. The computer program is executed by the processor to perform the operations of the method in any one of claims 1 to 12.

Images