Computing resource scheduling method and device, medium and electronic equipment

Abstract

The present disclosure provides a computing power resource scheduling method and device, a medium and an electronic device, and relates to the technical field of cloud. The method comprises the following steps: receiving a target application request about computing power resources sent by a resource consumer, wherein the target application request comprises a target type and a target scale of the computing power resources; determining a target name in a name list having a mapping relationship with the address information of the computing power resources according to the target type and the target scale; and sending the address information of the target computing power resources having a mapping relationship with the target name to the resource consumer, so that the resource consumer acquires the target resources according to the address information of the target resources. The computing power resources can be provided in the form of a cloud server. The technical scheme provided by the present disclosure can simplify the operation steps of the resource consumer when using the computing power resources to a certain extent, improve the ease of use of the computing power resources, and optimize the use experience of the computing power resources.

Description

Technical Field

[0001] This disclosure relates to the field of cloud technology, and more specifically, to a computing resource scheduling method, a computing resource scheduling device, and a computer-readable storage medium and electronic device for implementing the above-mentioned computing resource scheduling method. Background Technology

[0002] To provide computing resources, such as GPU (Graphics Processing Unit) resources and CPU (Central Processing Unit) resources, to application scenarios that require such resources in a simple and efficient manner, computing resources can be provided in the form of cloud servers.

[0003] In related technologies, when a party using computing resources (the resource consumer) needs such resources, they generally need to go through steps such as equipment application, environment deployment, and debugging. Furthermore, continuous optimization and iteration are required during the use of computing resources. It is evident that different resource consumers must go through the same or similar R&D environment deployment procedures when applying for and using computing resources, making the process cumbersome and repetitive.

[0004] On the other hand, after successfully obtaining the relevant computer resources, the resource list is usually provided to the resource consumer as an address information list, which the consumer then needs to manage itself. However, this solution is not very user-friendly, resulting in a poor user experience for the resource consumer.

[0005] It should be noted that the information disclosed in the background section above is only used to enhance the understanding of the background of this disclosure, and therefore may include information that does not constitute prior art known to those skilled in the art. Summary of the Invention

[0006] The purpose of this disclosure is to provide a computing resource scheduling method, computing resource scheduling device, electronic device, and computer-readable storage medium, thereby simplifying the operation steps when using computing resources to a certain extent and improving the ease of use of computing resources, thus enhancing the user experience of resource consumers.

[0007] According to one aspect of this disclosure, a method for scheduling computing resources is provided. The method includes: receiving a target request for computing resources sent by a resource consumer, the target request including: target type and target size of the computing resources; determining a target name from a list of names that have a mapping relationship with address information of the computing resources based on the target type and the target size; and sending the address information of the target computing resources that have a mapping relationship with the target name to the resource consumer, so that the resource consumer can obtain the target resources based on the address information of the target resources.

[0008] According to one aspect of this disclosure, a computing resource scheduling device is provided, comprising: a request receiving module, a name determination module, and an address information sending module.

[0009] The request receiving module is configured to receive a target application request for computing resources sent by the resource consumer, the target application request including the target type and target size of the computing resources; the name determination module is configured to determine the target name from a list of names that are mapped to the address information of the computing resources based on the target type and the target size; and the address information sending module is configured to send the address information of the target computing resources that are mapped to the target name to the resource consumer, so that the resource consumer can obtain the target resources based on the address information of the target resources.

[0010] In an exemplary embodiment, based on the foregoing solution, the apparatus further includes a name list determination module.

[0011] The aforementioned name list determination module is configured to: obtain a resource list of N computing devices from the resource provider to obtain the i-th address information of the i-th computing device in the resource list, where i is a positive integer not greater than N and N is a positive integer; based on the interface of the name service, establish a mapping relationship between the i-th address information and the j-th name according to the resource type of the i-th computing device, to obtain the aforementioned name list containing M names, where j is a positive integer not greater than M and M is a positive integer, and the j-th name corresponds to at least one address information.

[0012] In an exemplary embodiment, based on the foregoing solution, the apparatus further includes a name list update module.

[0013] The aforementioned name list update module is configured to: obtain the current resource list of the computing power device of the resource provider and determine the change information between the current resource list and the previously obtained resource list; and update the name list based on the name service interface and the change information.

[0014] In an exemplary embodiment, based on the foregoing solution, the above-mentioned device further includes: a resource detail information determination module.

[0015] The resource detail information determination module is configured as follows: For the xth name in a list of names containing X names, obtain the resource type and resource scale of Y computing devices corresponding to the Y address information that has a mapping relationship with the xth name, where x is a positive integer not greater than X, y is a positive integer not greater than Y, and X and Y are positive integers; use the yth address information in the Y address information as an identifier to cache the resource type and resource scale corresponding to the yth computing device to determine the resource detail information corresponding to the xth name; use the xth name as an identifier to cache the resource detail information corresponding to the xth name to determine the resource detail information corresponding to the xth name, thus obtaining the resource detail information corresponding to the list of names.

[0016] In an exemplary embodiment, based on the foregoing scheme, the name determination module is specifically configured to: determine whether the current resource details contain computing power resources of the target type; if the current resource details contain computing power resources of the target type, determine whether the current scale of the target type contained in the current resource details is not less than the target scale; if the current scale is not less than the target scale, determine the target name based on the target scale.

[0017] In an exemplary embodiment, based on the foregoing solution, the apparatus further includes a suggestion information determination module.

[0018] The aforementioned suggestion information determination module is configured to: determine first suggestion information based on the current scale when it is determined that the current scale is less than the target scale, and send the first suggestion information to the resource consumer; or, determine second suggestion information based on other types of computing resources included in the current resource details when it is determined that the current resource details do not contain the target type of computing resources, and send the second suggestion information to the resource consumer.

[0019] In an exemplary embodiment, based on the foregoing solution, the apparatus further includes a queue management module.

[0020] The queue management module is configured to: after determining the target name from the list of names that are mapped to the address information of computing resources, add the target application request to the resource application request queue with the target name as the head of the queue, so as to manage the multiple computing resource applications received through the resource application request queue.

[0021] In an exemplary embodiment, based on the foregoing scheme, the queue management module is further configured to: when the target application requests to leave the queue, obtain the address information of the target computing resources that have a mapping relationship with the target name.

[0022] In an exemplary embodiment, based on the foregoing scheme, the resource consumer corresponds to a cluster; the address information sending module is specifically configured to broadcast the address information of the target computing power resource that has a mapping relationship with the target name to each device in the cluster, so that each device in the cluster knows the communication address of the target computing power resource.

[0023] In an exemplary embodiment, based on the foregoing scheme, the resource consumer corresponds to a CPU cluster device, and the computing power resource is GPU computing power; the address information sending module is further specifically configured to: broadcast the address information of the target GPU computing power that has a mapping relationship with the target name to each device in the CPU cluster, so that the target GPU computing power is mounted to the CPU cluster device.

[0024] The aforementioned apparatus further includes a task execution module configured to execute a machine learning model training task via the aforementioned CPU cluster device and the mounted target GPU computing power.

[0025] In an exemplary embodiment, based on the foregoing scheme, the task execution module is specifically configured to: perform data cleaning and serialization processing on the CPU cluster device to determine training samples; and train a machine learning model using the training samples based on the target GPU computing power.

[0026] In an exemplary embodiment, based on the foregoing scheme, the apparatus further includes a construction module.

[0027] The aforementioned building module is configured to: construct a GPU computing power cluster at the resource provider end of the aforementioned GPU computing power based on one or more of the following information: network affinity, data center affinity, and GPU card type.

[0028] According to one aspect of this disclosure, a computer-readable storage medium is provided having a computer program stored thereon, which, when executed by a processor, implements the computing resource scheduling method described in any of the above embodiments.

[0029] According to one aspect of this disclosure, an electronic device is provided, comprising: a processor; and a memory for storing executable instructions of the processor; wherein the processor is configured to perform the computing resource scheduling method described in any of the above embodiments by executing the executable instructions.

[0030] According to one aspect of this disclosure, a computer program product or computer program is provided, comprising computer instructions stored in a computer-readable storage medium. A processor of a computer device reads the computer instructions from the computer-readable storage medium and executes the computer instructions, causing the computer device to perform the computing resource scheduling method provided in the various embodiments described above.

[0031] The computing resource scheduling scheme provided in one example embodiment of this disclosure can be applied to an intermediate service node that has data connections with both the resource consumer requesting computing resources and the resource provider offering computing resources. The intermediate service node pre-stores a list of names that are mapped to the address information of computing resources. Specifically, the intermediate service node receives a target request for computing resources from the resource consumer, and then determines the target name from the aforementioned name list based on the type and scale of the computing resources required in the target request. Further, the address information of the target computing resources mapped to the target name is sent to the resource consumer, so that the resource consumer can obtain the target resource based on the address information of the target resource.

[0032] As can be seen, this technical solution names the communication addresses of computing resources, allowing intermediate service nodes to return a list of computing resources to resource consumers using these names. This eliminates the need for individual resource consumers to manually manage the acquired resource lists, thus improving usability and optimizing the user experience. Furthermore, it enables dynamic scheduling of computing resources across different resource consumers, allowing for the attachment and unattaching of resources during and after use, achieving a plug-and-play functionality for computing resources.

[0033] It should be understood that the above general description and the following detailed description are exemplary and explanatory only, and are not intended to limit this disclosure. Attached Figure Description

[0034] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this disclosure and, together with the description, serve to explain the principles of this disclosure. It is obvious that the drawings described below are merely some embodiments of this disclosure, and those skilled in the art can obtain other drawings based on these drawings without any inventive effort.

[0035] Figure 1 A schematic diagram of a computing resource scheduling system that can be applied to embodiments of the present disclosure is shown.

[0036] Figure 2 The illustration shows a flowchart of a computing resource scheduling method according to an embodiment of the present disclosure.

[0037] Figure 3 A schematic diagram of the computing resource scheduling system according to this disclosure is shown.

[0038] Figure 4 The diagram illustrates the information interaction of a computing resource scheduling method according to an embodiment of the present disclosure.

[0039] Figure 5 A flowchart illustrating a method for filtering resource request submissions according to an embodiment of the present disclosure is shown.

[0040] Figure 6 The illustration shows a flowchart of a computing resource scheduling method according to another embodiment of the present disclosure.

[0041] Figure 7 A schematic diagram of a computing resource scheduling device that can be applied to an embodiment of the present disclosure is shown.

[0042] Figure 8 A schematic diagram of the structure of a computer system suitable for implementing the embodiments of the present disclosure is shown. Detailed Implementation

[0043] Example embodiments will now be described more fully with reference to the accompanying drawings. However, example embodiments can be implemented in many forms and should not be construed as limited to the examples set forth herein; rather, these embodiments are provided to make this disclosure more comprehensive and complete, and to fully convey the concept of the example embodiments to those skilled in the art. The described features, structures, or characteristics can be combined in any suitable manner in one or more embodiments. In the following description, numerous specific details are provided to give a full understanding of embodiments of this disclosure. However, those skilled in the art will recognize that the technical solutions of this disclosure can be practiced with one or more of the specific details omitted, or other methods, components, apparatus, steps, etc., can be employed. In other instances, well-known technical solutions are not shown or described in detail to avoid obscuring various aspects of this disclosure.

[0044] Furthermore, the accompanying drawings are merely illustrative of this disclosure and are not necessarily drawn to scale. The same reference numerals in the drawings denote the same or similar parts, and therefore repeated descriptions of them will be omitted. Some block diagrams shown in the drawings are functional entities and do not necessarily correspond to physically or logically independent entities. These functional entities may be implemented in software, in one or more hardware modules or integrated circuits, or in different network and / or processor devices and / or microcontroller devices.

[0045] Cloud technology refers to a managed technology that unifies a series of resources such as hardware, software, and networks within a wide area network or local area network to achieve data computing, storage, processing, and sharing.

[0046] Cloud technology is a collective term for network technologies, information technologies, integration technologies, management platform technologies, and application technologies applied to the cloud computing business model. It can form resource pools, providing flexible and convenient on-demand access. Cloud computing technology will become a crucial support. Backend services of technical network systems require substantial computing and storage resources, such as video websites, image websites, and many portal websites. With the rapid development and application of the internet industry, every item may have its own identification mark in the future, requiring transmission to backend systems for logical processing. Data at different levels will be processed separately, and various industry data will all require robust system support, which can only be achieved through cloud computing.

[0047] Cloud storage is a new concept that extends and develops from the concept of cloud computing. A distributed cloud storage system (hereinafter referred to as a storage system) refers to a storage system that uses cluster applications, grid technology and distributed storage file systems to bring together a large number of storage devices of various types in the network (storage devices are also called storage nodes) to work together through application software or application interfaces to provide data storage and business access functions to the outside world.

[0048] Currently, the storage method in storage systems is as follows: Logical volumes are created, and during creation, physical storage space is allocated to each logical volume. This physical storage space may consist of a single storage device or the disks of several storage devices. Clients store data on a logical volume, which means storing the data on the file system. The file system divides the data into many parts, each part being an object. Each object contains not only the data but also additional information such as a data identifier (ID, ID entity). The file system writes each object to the physical storage space of that logical volume and records the storage location information of each object. Therefore, when a client requests access to data, the file system can allow the client to access the data based on the storage location information of each object.

[0049] The process by which a storage system allocates physical storage space to a logical volume is as follows: the physical storage space is pre-divided into strips according to the capacity estimate of the objects stored in the logical volume (this estimate often has a large margin relative to the actual capacity of the objects to be stored) and the grouping of Redundant Array of Independent Disks (RAID). A logical volume can be understood as a strip, thus allocating physical storage space to the logical volume.

[0050] In this technical solution, the servers in the cluster can be independent physical servers, server clusters or distributed systems composed of multiple physical servers, or cloud servers that provide basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, CDN (Content Delivery Network), and big data and artificial intelligence platforms.

[0051] The business server terminal can be a smartphone, tablet, laptop, desktop computer, smart speaker, smartwatch, etc., but is not limited to these. The terminal and the server can be connected directly or indirectly through wired or wireless communication, which is not limited herein.

[0052] The computing resource scheduling system provided by this technical solution can be used in the following scenarios: With the development of AI (Artificial Intelligence) technology, the application of machine learning models is becoming increasingly widespread. The training process of machine learning models requires considerable computing resources; for example, data processing for neural network models requires heterogeneous computing. Heterogeneous computing refers to utilizing the collaborative computing capabilities of CPUs (Central Processing Units) and GPUs, the collaborative computing capabilities of APUs (Accelerated Processing Units) and other devices (such as FPGAs (Field Programmable Gate Arrays)), etc., to improve system speed by leveraging the computing power of multiple computing devices.

[0053] The computing resource scheduling system provided by this technical solution can also be used in scenarios where video production companies, live streaming platforms, and other businesses experience high concurrency video traffic and require real-time video encoding and decoding. In these cases, the computing resource scheduling system provides access to computing resources in cloud servers, such as GPU (Graphics Processing Unit) computing power. For example, by configuring GPU computing power using the computing resource scheduling method provided by this technical solution, model training can be conveniently and quickly completed using GPU computing power without the resource consumer (i.e., model training end) being aware of the business process. The aforementioned cloud server can be a node in a blockchain network.

[0054] For example, in machine learning model training scenarios, CPU computing resources are more numerous and readily available compared to GPU computing resources. Therefore, multiple CPU resources need to be matched with the same GPU resource to jointly serve the training of different machine learning models. However, to ensure physical isolation of GPU computing power when training different models, the same GPU computing resource cannot be matched with multiple CPU computing resources simultaneously to train multiple machine learning models. The traditional approach to this is to deploy CPU and GPU computing clusters in the same resource pool. That is, when different CPU clusters use the same GPU computing device, the GPU device is deployed to the local CPU computing cluster each time. This has drawbacks such as a long process, high time consumption, the need for manual intervention, and poor operability.

[0055] It is evident that when using CPU and GPU computing power in combination (such as in the training scenario of the AI ​​model mentioned above), given the scarcity of GPU computing power, it is necessary to enable GPU computing power to be dynamically used by multiple CPU computing power clusters. Therefore, how to facilitate the CPU computing power clusters to acquire GPU computing power for use, return it after use, and maintain physical isolation between CPU clusters during use is an urgent technical problem to be solved.

[0056] To address the aforementioned technical problems in related technologies, this technical solution provides a computing resource scheduling method, apparatus, computer-readable storage medium, and electronic device. Specifically, Figure 1 A schematic diagram of the system architecture of an exemplary application environment for a computing resource scheduling system that can be applied to embodiments of the present disclosure is shown.

[0057] like Figure 1As shown, the computing resource scheduling system 100 may include: one or more of terminal devices 101, 102, and 103 as resource consumers, a cloud server 105 as a resource provider, and nodes 106 arranged between the resource consumers and the resource providers. Network 104 serves as a medium for providing a communication link between the resource consumers and nodes 106, and also serves as a medium for providing a communication link between nodes 106 and the cloud server 105. Network 104 may include various connection types, such as wired or wireless communication links or fiber optic cables. Terminal devices 101, 102, and 103 may be various electronic devices with displays, including but not limited to desktop computers, portable computers, smartphones, and tablets.

[0058] It should be understood that Figure 1 The number of terminal devices, networks, and servers shown is merely illustrative. Depending on implementation needs, there can be any number of terminal devices, networks, and servers. For example, node 106 can be set as a virtual node, and cloud server 105 can be a server cluster composed of multiple servers.

[0059] The following details an embodiment of the computing resource scheduling method provided by this technical solution: Figure 2 A flowchart illustrating a computing resource scheduling method according to an embodiment of the present disclosure is shown. Specifically, the resource allocation method provided in this embodiment uses nodes deployed between the resource consumer and the resource provider as the execution entity. (Reference) Figure 2 The embodiment shown in the figure includes: Step S210: Receive a target request for computing resources sent by the resource consumer, wherein the target request includes the target type and target quantity of computing resources. Step S220: Based on the target type and the target size, determine the target name from the list of names that have a mapping relationship with the address information of computing resources; and, Step S230: Send the address information of the target computing power resource that has a mapping relationship with the target name to the resource consumer end, so that the resource consumer end can obtain the target resource according to the address information of the target resource.

[0060] The aforementioned computing resources may include GPU computing power, CPU computing power, Field-Programmable Gate Array (FPGA) computing power, or Application Specific Integrated Circuit (ASIC) chip computing power, etc., and this technical solution does not limit them. The consumer of the aforementioned resources can be any terminal that requires computing resources.

[0061] In an exemplary embodiment, the computing resource scheduling scheme provided by this technical solution can be explained using a scenario where GPU computing power is applied to the training of a machine learning model as an example. Specifically, GPU computing resources are obtained through the GPU computing resource scheduling scheme provided by this technical solution, and then the obtained GPU computing resources are combined with predetermined GPU computing resources to achieve the training of the machine learning model.

[0062] In the context of machine learning model training, the aforementioned resource consumer can refer to a terminal that already has a CPU computing power cluster and requires a GPU computing power cluster.

[0063] It can be seen that, in Figure 2 In the computing resource scheduling scheme provided in the illustrated embodiment, naming the computing resources enables plug-and-play functionality. For example, in machine learning model training scenarios, GPU computing resources can be dynamically embedded into the resource consumer (including CPU computing clusters), thereby enabling the on-demand mounting and offloading of GPU computing resources, effectively maximizing their utilization. This dynamic embedding of GPU computing resources into CPU computing clusters can be termed plug-and-play functionality for GPU computing resources.

[0064] In an exemplary embodiment, Figure 3 A schematic diagram of a computing resource scheduling system according to an embodiment of the present disclosure is shown. (Reference) Figure 3 The computing resource scheduling system includes: a resource consumer (which may include multiple resource consumers) 310, a resource provider 330, and an intermediate service node 320 deployed between the resource consumer 310 and the resource provider 330. For example, Figure 4 A timing diagram of a computing resource scheduling method according to an embodiment of the present disclosure is illustrated. Specifically, the computing resource scheduling method is based on the illustrated computing resource scheduling system, and the following can be combined with... Figure 4 right Figure 2 The specific implementation methods for each step are explained.

[0065] In order to achieve Figure 2 The technical solutions provided in the illustrated embodiments are for reference only. Figure 3 and Figure 4 The intermediate service node 320 determines a list of names that are mapped to the address information of computing resources through the naming service center 322. For example, refer to... Figure 4 Intermediate service node 320 executes step S41: Obtain the address information of each computing device in the resource list. And executes step S42: Determine the name list based on the address information.

[0066] In an exemplary embodiment, reference is made to Figure 5Steps S510 and S520 can be considered as a specific implementation of determining the name list.

[0067] In step S510, a resource list of N computing devices from the resource provider is obtained to acquire the i-th address information of the i-th computing device in the resource list, where i is a positive integer not greater than N and N is a positive integer. In step S520, based on the name-based service interface and according to the resource type of the i-th computing device, a mapping relationship between the i-th address information and the j-th name is established to obtain a name list containing M names, where j is a positive integer not greater than M and M is a positive integer. The j-th name corresponds to at least one address information.

[0068] For example, a GPU computing power cluster provides the aforementioned computing resources as a resource provider. Specifically, during GPU device initialization, a GPU computing power cluster is constructed based on factors such as network affinity, data center affinity, and GPU card type to achieve low latency in network communication between devices within the GPU cluster and to ensure that the devices have the same card type. Including devices of the same type (such as the GPU card type mentioned above) as a factor in constructing the computing power cluster facilitates naming devices within the same cluster and of the same type with the same name. This naming then makes it easier to satisfy different computing resource requests based on the resource details corresponding to the name.

[0069] For example, N devices belonging to the same computing power cluster can be named using M names. For instance, if the computing power cluster contains GPU devices of two different card types, then two names can be determined. Of course, this is not limited to naming computing power devices of the same resource type with the same name; naming can also be based on network affinity or data center affinity according to actual needs. This technical solution does not limit this approach.

[0070] In this technical solution, the aforementioned central service node can provide registration operations for the GPU computing power resource list. Specifically, during the construction of the GPU computing power cluster, the iplist of GPU computing power devices is registered through the naming service interface.

[0071] For example, to ensure the real-time nature of the name list, the aforementioned central service node can repeatedly obtain the current resource list of the computing power devices at the resource provider end, and determine the changes in the current resource list compared to the previously obtained resource list. Furthermore, based on the interface of the aforementioned name service, the name list is updated according to the aforementioned change information. Specifically, the name service can obtain the change information of the GPU computing power resource list through periodic polling, and further update the change information, thereby achieving state synchronization of name categories.

[0072] In this technical solution, the resource details corresponding to the above-mentioned name list are also determined, so as to determine a more suitable name for the received application request for computing power resources based on the resource details information (the specific implementation method will be described in the following embodiments). Therefore, referring to Figure 3 and Figure 4 The intermediate service node 320 determines the resource details corresponding to the aforementioned name list through the naming service center 322. For example, refer to... Figure 4 Intermediate service node 320 executes step S43: obtain the resource type and resource scale corresponding to each computing power resource. And executes step S44: determine the resource details based on the resource type and resource scale corresponding to each computing power resource.

[0073] In an exemplary embodiment, reference is made to Figure 5 Steps S530, S540, and S550 can be considered as a specific implementation method for determining resource details.

[0074] In step S530, for the x-th name in the name list containing X names, the resource types and resource sizes of the Y computing devices corresponding to the Y address information that have a mapping relationship with the x-th name are obtained, where x is a positive integer not greater than X, y is a positive integer not greater than Y, and X and Y are positive integers. In step S540, the resource type and resource size corresponding to the y-th computing device are cached using the y-th address information as an identifier to determine the resource detail information corresponding to the x-th name. And, in step S550, the resource detail information corresponding to the x-th name is cached using the x-th name as an identifier to determine the resource detail information corresponding to the x-th name, thus obtaining the resource detail information corresponding to the name list.

[0075] For example, suppose the name 'name1' in the name list has a mapping relationship with the address information of devices a, b, and c. Then, by obtaining the resource type and resource size of devices a, b, and c respectively, and summing them, the name information of 'name1' can be obtained. For instance, if the resource types of devices a, b, and c all belong to type 's', and the resource sizes of devices a, b, and c are summed to obtain the total size, then the resource details information of 'name1' can be obtained. For example, the resource details information corresponding to this name can be cached using 'name1' as an identifier. Similarly, the information of each name in the name list, such as 'name2', 'name3', etc., can be obtained.

[0076] In an exemplary embodiment, to enable dynamic scheduling of computing resources across different resource consumers and to ensure isolation when computing resources are used, this technical solution employs a queue (denoted as the "resource request queue") to manage resource request requests from resource consumers. Specifically, each received resource request is queued and placed at the end of the queue, undergoing a pending queuing operation. (See reference...) Figure 4 In this embodiment, the resource request queue can be set on the intermediate service node 320. Of course, the resource request queue can also be set on other computing nodes according to actual needs, and this application does not limit this.

[0077] In an exemplary embodiment, reference is made to Figure 4 The resource request queue management module (queue controller-master 321) performs filtering operations on resource request requests and enqueueing operations on resource request requests that meet the requirements. Therefore, in step S45, the aforementioned name list and resource detail information are sent and cached in queue controller-master 321. The specific implementation methods of the above operations are described below: For example, when resource consumer 310 sends a target application request in step S46, the intermediate service node 320 performs the following filtering operation on the received target application request through the queue controller-master 321: Step S47, determine whether the current computing power resources meet the target application request based on the current resource details. Here, "current computing power resources" specifically refers to currently available computing power resources, that is, among the registered computing power resources, resources in an unavailable state (such as those currently in use or under maintenance).

[0078] In an exemplary embodiment, reference is made to Figure 5 Steps S560 and S570 can be used as a specific implementation method to determine whether the current computing power resources meet the above target application request. For example, if the current computing power resources meet the above target application request, then the name of the relevant computing power resources is further assigned to them. If the current computing power resources do not meet the above target application request, then the current resource application request is discarded and relevant suggestion information is returned, thereby realizing the filtering of resource application requests.

[0079] In step S560, it is determined whether the current resource details information contains computing power resources of the target type.

[0080] For example, if it is determined that the current resource details information contains computing power resources of the target type, indicating that the current computing power resources have resources that match the resource type required by the target application request, then it is necessary to further determine whether the scale of the resource type in the current computing power resources meets the requirements of the target application request. For example, if step S570 is executed, it is determined whether the current scale of the target type contained in the current resource details information is not less than the target scale.

[0081] In an exemplary embodiment, if step S47 determines that the current computing power resources satisfy the target application request based on the current resource details, then step S50 is executed to determine the target name corresponding to the target application request.

[0082] For example, if it is determined that the current scale is not less than the target scale, it indicates that the scale of that resource type in the current computing power resources meets the requirements of the target application request. Then, step S580 is executed to determine the target name based on the target scale. For instance, if the target application request includes a target type of computing power resource (type a) and a target scale of "q+w1", and based on the resource details, it is determined that name1 has a resource type of a and a resource scale of q, name2 has a resource type of a and a resource scale of q2, and q2 is greater than w1, then it indicates that the current computing power resources can meet the target application request. It is evident that when a CPU cluster (resource consumer) mounts GPU computing power resources, it is not limited to a single name each time it mounts; multiple names can be mounted and used simultaneously based on the GPU computing power scale required in the application request. This improves scheduling flexibility.

[0083] In an exemplary embodiment, if step S47 determines that the current computing power resources cannot meet the target application request based on the current resource details, then the following steps are executed: step S48, determining suggestion information based on the current resource details and the target application request; and step S49, sending the suggestion information.

[0084] For example, if it is determined that the current scale is less than the target scale, it means that although there are resources in the current computing power resources that match the resource type required by the target application request, the scale of that resource type does not meet the requirements of the target application request. In this case, step S590 is executed to determine the first suggestion information based on the current scale and send the first suggestion information to the resource consumer. For example, the target type in the above target application request is: type a computing power resource, and the target scale is "q+w1". According to the resource details, it is determined that: the resource type of name1 is a and the resource scale is q, the resource type of name2 is a and the resource scale is q3, and q3 is less than w1. Then, the first suggestion information is determined based on the current scale "q+w3" (such as "It is recommended that you change the resource scale of resource type a to 'q+w3'") and the first suggestion information is sent to the resource consumer so that the resource consumer can determine whether to submit an application request again based on the suggestion information.

[0085] In another exemplary embodiment, if it is determined that the current resource details do not contain computing power resources of the target type, it means that the current computing power resources do not have resources matching the resource type required by the target application request. Therefore, it is not necessary to further determine whether the scale of that resource type in the current computing power resources meets the requirements of the target application request. Then, step S5100 is executed, determining second suggestion information based on other types of computing power resources included in the current resource details, and sending the second suggestion information to the resource consumer. For example, the resource types included in the current computing power resources and the scale of each type are sent to the resource consumer as the aforementioned second suggestion information.

[0086] In an exemplary embodiment, after determining the target name corresponding to the target request in step S50, the following is also performed: the resource request is enqueued with the target name as the head of the queue. For example, refer to... Figure 3 The queue head key is the name of the named service, such as "Name1" as the head of the resource consumer "CPU-1-CPU-2...", "Name2" as the head of the resource consumer "CPU-3-CPU-4...", "Target Name" as the head of the resource consumer "CPU-x-CPU-y...", and so on. The queue is managed on a first-come, first-served basis; newly added requests are queued and placed at the end of the queue.

[0087] Further, refer to Figure 4Step S51: The target application request for dequeuing is sent to the naming service center 322. Step S52: The address information of the target computing power resource with a mapping relationship to the target name is obtained. For example, for the resource consumer "CPU-3-CPU-4…", the address information of the target computing power resource with a mapping relationship to "Name2" can be obtained. Step S53: The address information of the target computing power resource is sent to the resource consumer. This allows the resource consumer to execute step S54: Obtain the target resource according to the address information of the target resource.

[0088] For example, the address information (e.g., GPU-iplist) of the target computing resources that have a mapping relationship with the aforementioned target name, such as "Name2", is broadcast to each device in the cluster, such as CPU-3-CPU-4..., so that each device in the cluster knows the communication address of the target computing resources, thereby enabling the GPU computing power to be mounted to the CPU cluster devices, and then performing machine learning model training tasks based on the relevant computing resources. During the execution of the machine learning model training task, data cleaning and serialization processing can be performed on the CPU cluster devices to determine training samples, and the machine learning model can be trained using the training samples based on the GPU computing power.

[0089] In an exemplary embodiment, Figure 6 A flowchart illustrating a computing resource scheduling method according to another embodiment of the present disclosure is shown. (See reference...) Figure 6 ,include: Step S610: The resource consumer sends a target request for GPU computing power resources.

[0090] In step S620, the Master determines whether the application requirements are met. If the application requirements are met, steps S630-S650 are executed. In step S630, the application is added to the pending queue (as described in the resource application request queue above). In step S640, the naming service center is accessed. And in step S650, the GPU-iplist is returned. If the application requirements are not met, step S660 is executed, and suggestion information is returned to the resource consumer.

[0091] The specific implementation methods of each of the above steps are as follows: Figure 4 , Figure 5 The corresponding embodiments have already been described, and will not be repeated here.

[0092] This technical solution is applicable to scenarios where CPU and GPU computing power are used interchangeably during machine learning model training. It enables the dynamic use of GPU computing power by multiple CPU computing power clusters without manual intervention. This solution adds a named service center, which, based on this name-based service, mounts the resource list of the GPU computing power cluster to the CPU computing power cluster, mounting when needed and unmounting when finished. The constructed automated processing flow supports dynamic scheduling of GPU computing power clusters, ensuring physical isolation of GPU computing power while facilitating shared use of GPU computing power and reducing resource scheduling costs.

[0093] Meanwhile, in this technical solution, the resource consumer does not need to directly interact with the resource provider offering computing power. Users on the resource consumer side do not need to concern themselves with the installation, deployment, debugging, and operation of the resources provided by the resource provider, thus simplifying the steps when using computing power resources. Furthermore, resource upgrades and updates by the resource provider are transparent to the resource consumer, meaning the resource consumer does not need to handle issues arising during the resource provider's operation.

[0094] Those skilled in the art will understand that all or part of the steps of the above embodiments are implemented as a computer program executed by a processor (including CPU and GPU). When the computer program is executed by the processor, it performs the functions defined by the methods provided in this disclosure. The program can be stored in a computer-readable storage medium, such as a read-only memory, a magnetic disk, or an optical disk.

[0095] Furthermore, it should be noted that the above figures are merely illustrative representations of the processes included in the methods according to exemplary embodiments of this disclosure, and are not intended to be limiting. It is readily understood that the processes shown in the above figures do not indicate or limit the temporal order of these processes. Additionally, it is readily understood that these processes may be executed synchronously or asynchronously, for example, in multiple modules.

[0096] The following is a description of the computing resource scheduling device provided in this technical solution: This example embodiment provides a computing resource scheduling device. (See reference...) Figure 7 As shown, the computing power resource scheduling device 700 includes: a request receiving module 701, a name determination module 702, and an address information sending module 703.

[0097] The request receiving module 701 is configured to receive a target application request for computing resources sent by the resource consumer, the target application request including the target type and target size of the computing resources; the name determination module 702 is configured to determine the target name from a list of names that are mapped to the address information of the computing resources based on the target type and the target size; and the address information sending module 703 is configured to send the address information of the target computing resources that are mapped to the target name to the resource consumer, so that the resource consumer can obtain the target resources based on the address information of the target resources.

[0098] In an exemplary embodiment, based on the foregoing solution, the above-mentioned device further includes: a name list determination module 704.

[0099] The name list determination module 704 is configured to: obtain a resource list of N computing devices from the resource provider to obtain the i-th address information of the i-th computing device in the resource list, where i is a positive integer not greater than N and N is a positive integer; based on the interface of the name service, establish a mapping relationship between the i-th address information and the j-th name according to the resource type of the i-th computing device to obtain the name list containing M names, where j is a positive integer not greater than M and M is a positive integer, and the j-th name corresponds to at least one address information.

[0100] In an exemplary embodiment, based on the foregoing solution, the above-mentioned device further includes: a name list update module 705.

[0101] The name list update module 705 is configured to: obtain the current resource list of the computing power device of the resource provider and determine the change information between the current resource list and the previously obtained resource list; and update the name list according to the change information based on the interface of the name service.

[0102] In an exemplary embodiment, based on the foregoing solution, the above-mentioned device further includes: a resource detail information determination module 706.

[0103] Specifically, the resource detail information determination module 706 is configured to: for the xth name in a name list containing X names, obtain the resource type and resource scale of Y computing devices corresponding to the Y address information that has a mapping relationship with the xth name, where x is a positive integer not greater than X, y is a positive integer not greater than Y, and X and Y are positive integers; cache the resource type and resource scale corresponding to the yth computing device using the yth address information in the Y address information as an identifier to determine the resource detail information corresponding to the xth name; cache the resource detail information corresponding to the xth name using the xth name as an identifier to determine the resource detail information corresponding to the xth name, thereby obtaining the resource detail information corresponding to the name list.

[0104] In an exemplary embodiment, based on the foregoing scheme, the name determination module 702 is specifically configured to: determine whether the current resource details information contains computing power resources of the target type; if the current resource details information contains computing power resources of the target type, determine whether the current scale of the target type contained in the current resource details information is not less than the target scale; if the current scale is not less than the target scale, determine the target name based on the target scale.

[0105] In an exemplary embodiment, based on the foregoing solution, the above-mentioned apparatus further includes: a suggestion information determination module 707.

[0106] The aforementioned suggestion information determination module 707 is configured to: determine first suggestion information based on the current scale when the current scale is less than the target scale, and send the first suggestion information to the resource consumer; or, determine second suggestion information based on other types of computing resources included in the current resource details when the current resource details do not contain the target type of computing resources, and send the second suggestion information to the resource consumer.

[0107] In an exemplary embodiment, based on the foregoing solution, the above-mentioned device further includes a queue management module 708.

[0108] The queue management module 708 is configured to: after determining the target name from the list of names that have a mapping relationship with the address information of computing resources, add the target application request to the resource application request queue with the target name as the head of the queue, so as to manage the multiple computing resource applications received through the resource application request queue.

[0109] In an exemplary embodiment, based on the foregoing scheme, the queue management module 708 is further configured to: obtain the address information of the target computing resources that have a mapping relationship with the target name when the target application requests to leave the queue.

[0110] In an exemplary embodiment, based on the foregoing scheme, the resource consumer corresponds to a cluster; the address information sending module 703 is specifically configured to broadcast the address information of the target computing power resource that has a mapping relationship with the target name to each device in the cluster, so that each device in the cluster knows the communication address of the target computing power resource.

[0111] In an exemplary embodiment, based on the foregoing scheme, the resource consumer corresponds to a CPU cluster device, and the computing power resource is GPU computing power; the address information sending module 703 is further specifically configured to: broadcast the address information of the target GPU computing power that has a mapping relationship with the target name to each device in the CPU cluster, so that the target GPU computing power is mounted to the CPU cluster device. The aforementioned apparatus further includes a task execution module 709, configured to execute a machine learning model training task via the aforementioned CPU cluster device and the mounted target GPU computing power.

[0112] In an exemplary embodiment, based on the foregoing scheme, the task execution module 709 is specifically configured to: perform data cleaning and serialization processing on the CPU cluster device to determine training samples; and train a machine learning model using the training samples based on the target GPU computing power.

[0113] In an exemplary embodiment, based on the foregoing scheme, the above-mentioned apparatus further includes a construction module 710.

[0114] The aforementioned building module 710 is configured to: build a GPU computing power cluster at the resource provider of the aforementioned GPU computing power based on one or more of the following information: network affinity, data center affinity, and GPU card type.

[0115] The specific details of each module or unit in the aforementioned computing resource scheduling device have been described in detail in the corresponding computing resource scheduling method, so they will not be repeated here.

[0116] Figure 8 A schematic diagram of a computer system suitable for implementing embodiments of the present invention is shown.

[0117] It should be noted that, Figure 8 The computer system 800 of the electronic device shown is merely an example and should not impose any limitation on the functionality and scope of use of the embodiments of the present invention.

[0118] like Figure 8 As shown, the computer system 800 includes a processor 801, which may include a graphics processing unit (GPU) and a central processing unit (CPU). The processor 801 can perform various appropriate actions and processes based on programs stored in read-only memory (ROM) 802 or programs loaded from storage portion 808 into random access memory (RAM) 803. The RAM 803 also stores various programs and data required for system operation. The processor (GPU / CPU) 801, ROM 802, and RAM 803 are interconnected via a bus 804. An input / output (I / O) interface 805 is also connected to the bus 804.

[0119] The following components are connected to I / O interface 805: an input section 806 including a keyboard, mouse, etc.; an output section 807 including a cathode ray tube (CRT), liquid crystal display (LCD), etc., and speakers, etc.; a storage section 808 including a hard disk, etc.; and a communication section 809 including a network interface card such as a LAN (Local Area Network) card, modem, etc. The communication section 809 performs communication processing via a network such as the Internet. A drive 810 is also connected to I / O interface 805 as needed. A removable medium 811, such as a disk, optical disk, magneto-optical disk, semiconductor memory, etc., is installed on drive 810 as needed so that computer programs read from it can be installed into storage section 808 as needed.

[0120] In particular, according to embodiments of this disclosure, the processes described below with reference to the flowcharts can be implemented as computer software programs. For example, embodiments of this disclosure include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via communication section 809, and / or installed from removable medium 811. When the computer program is executed by processor (GPU / CPU) 801, it performs various functions defined in the system of this application. In some embodiments, the computer system 800 may also include an AI (Artificial Intelligence) processor for processing computational operations related to machine learning.

[0121] It should be noted that the computer-readable medium shown in the embodiments of this disclosure can be a computer-readable signal medium or a computer-readable storage medium, or any combination of the two. A computer-readable storage medium can be, for example,—but not limited to—an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of a computer-readable storage medium may include, but are not limited to: an electrical connection having one or more wires, a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM), flash memory, optical fiber, portable compact disc read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination thereof. In this disclosure, a computer-readable storage medium can be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device. In this disclosure, a computer-readable signal medium can include a data signal propagated in baseband or as part of a carrier wave, carrying computer-readable program code. Such transmitted data signals can take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. The computer-readable signal medium can also be any computer-readable medium other than a computer-readable storage medium, which can send, propagate, or transmit a program for use by or in connection with an instruction execution system, apparatus, or device. The program code contained on the computer-readable medium can be transmitted using any suitable medium, including but not limited to wireless, wired, etc., or any suitable combination thereof.

[0122] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of this disclosure. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in a block diagram or flowchart, and combinations of blocks in a block diagram or flowchart, may be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0123] The units described in the embodiments of this disclosure can be implemented in software or hardware, and the described units can also be located in a processor. The names of these units do not necessarily constitute a limitation on the unit itself.

[0124] In another aspect, this application also provides a computer-readable medium, which may be included in the electronic device described in the above embodiments; or it may exist independently and not assembled into the electronic device. The computer-readable medium carries one or more programs, which, when executed by the electronic device, cause the electronic device to perform the methods described in the above embodiments.

[0125] For example, the electronic device described above can achieve the following: Figure 2 As shown: Step S210, receiving a target application request for computing power resources sent by the resource consumer, the target application including: target type and target size of computing power resources; Step S220, determining a target name from a list of names that have a mapping relationship with the address information of computing power resources according to the target type and the target size; and Step S230, sending the address information of the target computing power resources that have a mapping relationship with the target name to the resource consumer, so that the resource consumer can obtain the target resource according to the address information of the target resource.

[0126] It should be noted that although several modules or units for the device used to perform actions have been mentioned in the detailed description above, this division is not mandatory. In fact, according to embodiments of this disclosure, the features and functions of two or more modules or units described above can be embodied in one module or unit. Conversely, the features and functions of one module or unit described above can be further divided and embodied by multiple modules or units.

[0127] From the above description of the embodiments, those skilled in the art will readily understand that the exemplary embodiments described herein can be implemented by software or by combining software with necessary hardware. Therefore, the technical solutions according to the embodiments of this disclosure can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (such as a CD-ROM, USB flash drive, external hard drive, etc.) or on a network, including several instructions to cause a computing device (such as a personal computer, server, touch terminal, or network device, etc.) to execute the methods according to the embodiments of this disclosure.

[0128] Other embodiments of this disclosure will readily occur to those skilled in the art upon consideration of the specification and practice of the invention disclosed herein. This application is intended to cover any variations, uses, or adaptations of this disclosure that follow the general principles of this disclosure and include common knowledge or customary techniques in the art not disclosed herein. The specification and embodiments are to be considered exemplary only, and the true scope and spirit of this disclosure are indicated by the foregoing claims. It should be understood that this disclosure is not limited to the precise structures described above and shown in the drawings, and various modifications and changes can be made without departing from its scope.

Claims

1. A method for scheduling computing resources, characterized in that, The method includes: Receive a target request for computing resources sent by a resource consumer, wherein the target request includes the target type and target quantity of computing resources; If the resource details information corresponding to the name list that has a mapping relationship with the address information of computing power resources contains computing power resources of the target type, and the current scale of the computing power resources of the target type contained in the resource details information is not less than the target scale, then the target name is determined in the name list according to the target scale. The address information of the target computing power resource that has a mapping relationship with the target name is sent to the resource consumer, so that the resource consumer can obtain the target computing power resource according to the address information of the target computing power resource; The resource details information corresponding to the name list is generated in the following way: For the x-th name in the name list, obtain the resource type and resource scale of the Y computing devices corresponding to the Y address information that have a mapping relationship with the x-th name, where x is less than or equal to the number of names contained in the name list, and Y is a positive integer; Using the y-th address information among the Y address information as an identifier, cache the resource type and resource scale corresponding to the y-th computing power device pointed to by the y-th address information to determine the resource details information corresponding to the x-th name, where y is a positive integer not greater than Y; Using the x-th name as an identifier, cache the resource detail information corresponding to the x-th name to generate the resource detail information corresponding to the name list.

2. The computing resource scheduling method according to claim 1, characterized in that, The method further includes: Obtain a resource list of N computing devices from the resource provider to obtain the address information of the i-th computing device in the resource list, where i is a positive integer not greater than N and N is a positive integer; Based on the interface of the name-based service, according to the resource type of the i-th computing power device, a mapping relationship between the i-th address information and the j-th name is established to obtain the name list containing M names, where j is a positive integer not greater than M, and M is a positive integer, and the j-th name corresponds to at least one address information.

3. The computing resource scheduling method according to claim 2, characterized in that, The method further includes: Obtain the current resource list of the computing power device of the resource provider, and determine the change information between the current resource list and the previously obtained resource list; Based on the interface of the naming service, the name list is updated according to the change information.

4. The computing resource scheduling method according to claim 1, characterized in that, The method further includes: If it is determined that the current scale is less than the target scale, first suggestion information is determined based on the current scale, and the first suggestion information is sent to the resource consumer.

5. The computing resource scheduling method according to claim 1, characterized in that, The method further includes: If it is determined that the target type of computing power resources are not included in the resource details information, second suggestion information is determined based on other types of computing power resources included in the current resource details information, and the second suggestion information is sent to the resource consumer.

6. The computing resource scheduling method according to claim 1, characterized in that, After determining the target name from the name list based on the target size, the method further includes: The target application request is added to a resource application request queue with the target name as the head, so as to manage the received multiple computing resource applications through the resource application request queue.

7. The computing resource scheduling method according to claim 6, characterized in that, The method further includes: In the case of the target application requesting dequeueing, obtain the address information of the target computing resource that has a mapping relationship with the target name.

8. The computing resource scheduling method according to claim 1, characterized in that, The resource consumer corresponds to the cluster; sending the address information of the target computing power resource that has a mapping relationship with the target name to the resource consumer includes: The address information of the target computing resource that has a mapping relationship with the target name is broadcast to each device in the cluster, so that each device in the cluster knows the communication address of the target computing resource.

9. The computing resource scheduling method according to claim 1, characterized in that, The resource consumer corresponds to a CPU cluster device, and the computing power resource is GPU computing power; sending the address information of the target computing power resource that has a mapping relationship with the target name to the resource consumer includes: The address information of the target GPU computing power that has a mapping relationship with the target name is broadcast to each device in the CPU cluster device, so that the target GPU computing power is mounted to the CPU cluster device; The method further includes: performing a machine learning model training task through the CPU cluster device and the target GPU computing power mounted thereon.

10. The computing resource scheduling method according to claim 9, characterized in that, Executing machine learning model training tasks using the CPU cluster device and the mounted target GPU computing power includes: Data cleaning and serialization are performed on the CPU cluster device to determine training samples; Based on the target GPU computing power, a machine learning model is trained using the training samples.

11. The computing resource scheduling method according to claim 9, characterized in that, The method further includes: At the resource provider of the GPU computing power, a GPU computing power cluster is constructed based on one or more of the following information: network affinity, data center affinity, and GPU card type.

12. A computing resource scheduling device, characterized in that, The device includes: The receiving module is configured to receive a target request for computing resources sent by the resource consumer, wherein the target request includes the target type and target size of the computing resources. The determination module is configured to: if the resource details information corresponding to the name list that has a mapping relationship with the address information of the computing power resources contains computing power resources of the target type, and the current scale of the computing power resources of the target type contained in the resource details information is not less than the target scale, then determine the target name in the name list according to the target scale; The sending module is configured to send the address information of the target computing power resource that has a mapping relationship with the target name to the resource consumer, so that the resource consumer can obtain the target computing power resource according to the address information of the target computing power resource; The resource details information corresponding to the name list is generated in the following way: For the x-th name in the name list, obtain the resource type and resource scale of the Y computing devices corresponding to the Y address information that have a mapping relationship with the x-th name, where x is less than or equal to the number of names contained in the name list, and Y is a positive integer; Using the y-th address information among the Y address information as an identifier, cache the resource type and resource scale corresponding to the y-th computing power device pointed to by the y-th address information to determine the resource details information corresponding to the x-th name, where y is a positive integer not greater than Y; Using the x-th name as an identifier, cache the resource detail information corresponding to the x-th name to generate the resource detail information corresponding to the name list.

13. A computer-readable storage medium, characterized in that, It contains computer programs; When the computer program is executed by a processor, it implements the method as described in any one of claims 1 to 11.

14. An electronic device, characterized in that, include: processor; as well as Memory for storing the executable instructions of the processor; The processor is configured to perform the method of any one of claims 1 to 11 by executing the executable instructions.

15. A computer program product, characterized in that, The computer program product includes a computer program stored in a computer-readable storage medium, wherein a processor of a computer device reads from the computer-readable storage medium and executes the computer program, causing the computer device to perform the method of any one of claims 1 to 11.

Images