Computing task execution method, computing task scheduling device, and storage medium

Abstract

The present application discloses a computing power task execution method, computing power scheduling equipment, and storage medium, relating to the field of data processing technology. The method comprises: obtaining task attribute information of a target computing power task, obtaining matching historical computing power tasks and a set of historical computing power nodes that execute the historical computing power tasks based on the task attribute information; determining the correlation information between the historical computing power tasks and the target computing power tasks; determining the target computing power node set corresponding to the target computing power task based on the correlation information and the set of historical computing power nodes; and scheduling computing power nodes in the target computing power node set to execute the target computing power task. The present application aims to ensure that the target computing power task requirements are met while improving the efficiency of computing power scheduling to meet the requirements for efficient computing power task execution.

Description

Technical Field

[0001] The present application relates to the field of data processing technology, and in particular to a computing task execution method, a computing task scheduling device, and a storage medium. Background Art

[0002] In the computing power leasing process, it is generally necessary to first build a computing power node cluster, and then select computing power nodes that meet the computing task requirements (such as virtual machines or cloud containers, etc.) to schedule and execute the corresponding computing power tasks.

[0003] In related technologies, when selecting computing power nodes, it is generally necessary to traverse each computing power node in the calculation to filter out a set of nodes that meet the requirements of the computing task. However, this screening method is inefficient, affects the efficiency of computing power scheduling, and is difficult to meet the requirements of efficient computing power task execution. Summary of the Invention

[0004] The main purpose of this application is to provide a computing task execution method, computing task scheduling device and storage medium, aiming to improve the efficiency of computing task scheduling to meet the needs of efficient computing task execution.

[0005] To achieve the above objectives, this application proposes a computing task execution method, which includes:

[0006] Obtaining task attribute information of the target computing task, and obtaining matching historical computing tasks and a set of historical computing nodes that execute the historical computing tasks based on the task attribute information;

[0007] Determine the correlation information between the historical computing power task and the target computing power task;

[0008] Determine the target computing power node set corresponding to the target computing power task according to the correlation information and the historical computing power node set;

[0009] Schedule the computing nodes in the target computing node set to execute the target computing task.

[0010] In one embodiment, the step of determining the correlation information between the historical computing power task and the target computing power task includes:

[0011] Determining first state information indicating a difference in data source location based on location data of data sources corresponding to the historical computing power task and the target computing power task respectively;

[0012] Determining second status information indicating task differences based on execution requirements of a plurality of subtasks respectively included in the historical computing power task and the target computing power task;

[0013] Determining third status information indicating a difference in computing power types based on the computing power types corresponding to the historical computing power task and the target computing power task respectively;

[0014] The association information includes the first state information, the second state information and the third state information.

[0015] In one embodiment, the step of determining the target computing power node set corresponding to the target computing power task based on the correlation information and the historical computing power node set includes:

[0016] When the third state information includes that the computing power type of the historical computing power task is consistent with the computing power type of the target computing power task, determining the similarity between the historical computing power task and the target computing power task according to the first state information and the second state information;

[0017] When the similarity is greater than a first preset threshold, determining the historical computing power node set as the target computing power node set;

[0018] When the similarity is greater than a second preset threshold and less than or equal to the first preset threshold, adjusting the historical computing power node set according to the first state information and the second state information to obtain the target computing power node set;

[0019] When the similarity is less than or equal to the second preset threshold, traverse all computing nodes in the computing node cluster and take a set of nodes that meet the preset computing scheduling strategy as the target computing node set;

[0020] The second preset threshold is smaller than the first preset threshold.

[0021] In one embodiment, the first state information includes a first deviation value indicating a difference in data source location, and the second state information includes a second deviation value indicating a difference in computing power processing capability required by the task. The step of determining the similarity between the historical computing power task and the target computing power task based on the first state information and the second state information includes:

[0022] Priority information between the data source location and computing power processing capability corresponding to the target computing power task;

[0023] Determine, according to the priority information, a first weight corresponding to the first deviation value and a second weight corresponding to the second deviation value;

[0024] The similarity is calculated according to the first deviation value and the corresponding first weight, and the second deviation value and the corresponding second weight.

[0025] In one embodiment, the step of determining the target computing power node set corresponding to the target computing power task based on the correlation information and the historical computing power node set includes:

[0026] When the third state information includes that the computing power type of the historical computing power task is inconsistent with the computing power type of the target computing power task, determining the similarity between the historical computing power task and the target computing power task according to the first state information and the second state information;

[0027] When the similarity is greater than a first preset threshold, adjusting the historical computing power node set according to the first state information, the second state information, and the third state information to obtain the target computing power node set;

[0028] When the similarity is less than the first preset threshold, all computing nodes in the computing node cluster are traversed, and a set of nodes that meet the preset computing power scheduling strategy is used as the target computing node set.

[0029] In one embodiment, the step of determining first status information indicating a difference in data source location based on the location data of the data sources corresponding to the historical computing power task and the target computing power task respectively includes:

[0030] Determine a first center position of all data sources corresponding to the historical computing power task, and determine a second center position of all data sources corresponding to the target computing power task;

[0031] A first deviation value indicating a difference in data source locations is determined according to a distance between the first center location and the second center location, and the first state information includes the first deviation value.

[0032] In one embodiment, the execution requirement includes the total number of subtasks and the proportion of the number of subtasks allowed to be executed in parallel to the total number of tasks. The step of determining the second status information representing the task difference based on the execution requirements of the multiple subtasks respectively included in the historical computing power task and the target computing power task includes:

[0033] Determine a first difference between the total number of tasks corresponding to the historical computing power tasks and the total number of tasks corresponding to the target computing power tasks, and determine a second difference between the proportion corresponding to the historical computing power tasks and the proportion corresponding to the target computing power tasks;

[0034] A second deviation value representing a difference in computing power processing capability required by the task is determined based on the first difference value and the second difference value, and the second state information includes the second deviation value.

[0035] In one embodiment, the task attribute information includes a task application scenario and a billing method, and the step of obtaining a matching historical computing task based on the task attribute information includes:

[0036] Determine the computing power tasks that meet the preset conditions and are executed before the current moment as the historical computing power tasks;

[0037] The preset conditions include that the corresponding task application scenario and billing method are consistent with the task application scenario and billing method of the target computing power task.

[0038] In addition, to achieve the above-mentioned purpose, the present application also proposes a computing power scheduling device, which includes: a memory, a processor, and a computer program stored on the memory and runnable on the processor, and the computer program is configured to implement the steps of the computing power task execution method described above.

[0039] In addition, to achieve the above-mentioned purpose, the present application also proposes a storage medium, which is a computer-readable storage medium. A computer program is stored on the storage medium, and when the computer program is executed by the processor, the steps of the computing task execution method described above are implemented.

[0040] One or more technical solutions proposed in this application have at least the following technical effects: the solution first obtains the set of historical computing power nodes used when executing the matching historical computing power task based on the task attribute information of the target computing power task, and then determines the target computing power node set that executes the target computing power task based on the correlation information between the historical computing power task and the target computing power task and the set of historical computing power nodes. Based on this method, the analysis process of the target computing power node set that is currently executing the target computing power task can effectively reduce the process of traversing and screening all nodes in the computing power node cluster one by one, which can ensure that the target computing power task requirements are met while improving the efficiency of computing power scheduling to meet the requirements of efficient computing power task execution. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present application and, together with the description, serve to explain the principles of the present application.

[0042] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, for ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0043] Figure 1 A flowchart of the first embodiment of the computing task execution method of this application is provided;

[0044] Figure 2 This is a schematic diagram of the device structure of the hardware operating environment involved in the computing task execution method in the embodiment of the present application.

[0045] The purpose, features and advantages of this application will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0046] It should be understood that the specific embodiments described herein are merely used to explain the technical solutions of the present application and are not intended to limit the present application.

[0047] In order to better understand the technical solution of the present application, a detailed description will be given below in conjunction with the accompanying drawings and specific implementation methods.

[0048] The main solution of the embodiment of the present application is: obtaining task attribute information of the target computing power task, obtaining matching historical computing power tasks and a set of historical computing power nodes that execute the historical computing power tasks based on the task attribute information; determining the correlation information between the historical computing power tasks and the target computing power tasks; determining the target computing power node set corresponding to the target computing power task based on the correlation information and the historical computing power node set; and scheduling the computing power nodes in the target computing power node set to execute the target computing power task.

[0049] In this embodiment, for ease of description, the following description is made with the computing power scheduling device as the execution entity.

[0050] In the existing technology, when selecting computing power nodes, it is generally necessary to traverse each computing power node in the calculation to filter out a set of nodes that meet the requirements of the computing task. However, this screening method is inefficient, affects the efficiency of computing power scheduling, and is difficult to meet the requirements of efficient computing power task execution.

[0051] The present application provides the above-mentioned solution, which first obtains the set of historical computing power nodes used when executing the matching historical computing power task based on the task attribute information of the target computing power task, and then determines the target computing power node set for executing the target computing power task based on the correlation information between the historical computing power task and the target computing power task and the set of historical computing power nodes. Based on this method, the analysis process of the target computing power node set currently executing the target computing power task can effectively reduce the process of traversing and screening all nodes in the computing power node cluster one by one, which can ensure that the target computing power task requirements are met while improving the efficiency of computing power scheduling to meet the requirements of efficient computing power task execution.

[0052] It should be noted that the execution subject of this embodiment can be a computing service device with data processing, network communication, and program execution functions, such as a tablet computer, personal computer, mobile phone, etc., or an electronic device or computing power scheduling device capable of implementing the above functions. The following uses the computing power scheduling device as an example to illustrate this embodiment and the following embodiments.

[0053] Based on this, the embodiment of the present application provides a method for executing a computing task, referring to Figure 1 , Figure 1 This is a flow chart of the first embodiment of the computing task execution method of this application.

[0054] In this embodiment, the computing task execution method includes steps S10 to S40:

[0055] Step S10: Obtain task attribute information of the target computing task, and obtain matching historical computing tasks and a set of historical computing nodes that execute the historical computing tasks based on the task attribute information;

[0056] The target computing task is the computing task that is currently executed by the required computing node. Task attribute information may include at least one of the following: application scenario, computing power requirement information (such as computing power type, required computing power performance, performance mode, etc.), data source type, etc.

[0057] The number of historical computing tasks matched here can be one or more than one.

[0058] The historical computing power node set may include at least one of a virtual machine, a cloud container, etc. In this embodiment, the computing power unit used by the nodes in the historical computing power node set is a GPU unit.

[0059] Compare the historical task attribute information corresponding to all computing power tasks executed before the current moment with the task attribute information of the target computing power task, and determine the computing power task corresponding to the historical task attribute information that matches the task attribute information of the target computing power task as the historical computing power task. In this embodiment, the task attribute information includes the task application scenario and the billing method, and the computing power task that meets the preset conditions and is executed before the current moment is determined to be the historical computing power task; the preset conditions include the corresponding task application scenario and the billing method being consistent with the task application scenario and the billing method of the target computing power task. Among them, the task application scenario may include one of the following: artificial intelligence (AI) large model training task, image generation task, video generation task, three-dimensional content creation task, etc.; the billing method may include hourly calculation or processing volume billing, etc.

[0060] Step S20: determining the correlation information between the historical computing power task and the target computing power task;

[0061] The correlation information indicates the similarity between the computing power usage status when the historical computing power task is executed and the target computing power task is executed.

[0062] The correlation information can be obtained by analyzing the relevant status parameters during the computing power usage process corresponding to the historical computing power tasks and the target computing power tasks (such as at least one of the location data of the data source, the execution requirements of the subtask, and the computing power type), or by analyzing the task description tag corresponding to the target computing power task and the task description tag corresponding to the historical computing power task (the task description tag here can be selected by the computing power demander from multiple preset tags).

[0063] When there are more than one historical computing power tasks, the correlation information between each historical computing power task and the target computing power task is determined separately.

[0064] Step S30: determining a target computing power node set corresponding to the target computing power task based on the correlation information and the historical computing power node set;

[0065] In one implementation, if the correlation information satisfies a preset similarity condition, the historical computing power node set is determined as the target computing power node set. If the correlation information does not satisfy the preset similarity condition, the target computing power node set can be obtained by adjusting the historical computing power node set based on the correlation information.

[0066] In another implementation method, when there are more than one historical computing power tasks, the historical computing power task with the highest similarity to the target computing power task can be determined as the target task among more than one historical computing power tasks based on more than one piece of correlation information, and the historical computing power node set corresponding to the target task can be used as the target computing power node set.

[0067] In another implementation method, when there is more than one historical computing power task, the evaluation scores corresponding to each historical computing power task in different service indicators can be obtained, and the indicator with the highest priority among the target computing power task in different service indicators is obtained as the target indicator. The historical computing power task with an evaluation score corresponding to the target indicator greater than or equal to the preset score is used as the reference task, and the target computing power node set is determined based on the correlation information corresponding to the reference task and the historical computing power node set. When there is more than one reference task, the reference task with the highest similarity to the target computing power task can be determined as the target task based on the correlation information, and the historical computing power node set corresponding to the target task is used as the target computing power node set.

[0068] Step S40: Schedule the computing nodes in the target computing node set to execute the target computing task.

[0069] In the process of executing the target computing power task, all computing power nodes in the target computing power node set can be called at the same time, or the corresponding computing power nodes can be called to execute the target computing power task in time periods or to adapt to the needs of different subtasks in the target computing power task.

[0070] This embodiment provides a computing power task execution method, which first obtains the historical computing power node set used when executing the matching historical computing power task based on the task attribute information of the target computing power task, and then determines the target computing power node set for executing the target computing power task based on the correlation information between the historical computing power task and the target computing power task and the historical computing power node set. Based on this method, the analysis process of the target computing power node set currently executing the target computing power task can effectively reduce the process of traversing and screening all nodes in the computing power node cluster one by one, which can ensure that the target computing power task requirements are met while improving the efficiency of computing power scheduling to meet the requirements of efficient computing power task execution.

[0071] In a feasible implementation, the step of determining the correlation information between the historical computing power task and the target computing power task includes: determining first status information representing the difference in data source location based on the location data of the data sources corresponding to the historical computing power task and the target computing power task respectively; determining second status information representing the difference in task based on the execution requirements of multiple subtasks respectively contained in the historical computing power task and the target computing power task; determining third status information representing the difference in computing power type based on the computing power types corresponding to the historical computing power task and the target computing power task respectively; wherein the correlation information includes the first status information, the second status information and the third status information.

[0072] The greater the difference in the comprehensive representation of the first state information, the second state information, and the third state information, the smaller the similarity between the corresponding historical computing power task and the target computing power task.

[0073] The first status information may include whether the data source locations are consistent or the location deviations between data source locations. The second status information may include whether the execution requirements are consistent or the requirement deviations between execution requirements. The third status information may include whether the computing power types are consistent or the type deviations between computing power types.

[0074] Among them, the data source is the source of data required during the execution of computing tasks. The data source corresponding to a computing task can be one or more, and multiple data sources can be in the same location or different locations.

[0075] In this embodiment, a first center position of all data sources corresponding to the historical computing task is determined, and a second center position of all data sources corresponding to the target computing task is determined. A first deviation value representing the difference in data source locations is determined based on the distance between the first center position and the second center position, and the first status information includes the first deviation value. If there are multiple data sources corresponding to the historical computing task or the target computing task, the locations of the multiple data sources can be connected by lines. Among the multiple closed areas enclosed by the multiple lines, the largest closed area is determined as the data source area, and the location of the center of gravity of the data source area can be used as the corresponding center position. The distance between the first center position and the second center position can be used as the first deviation value. Here, the first deviation value can accurately reflect the difference in data transmission time corresponding to the execution of the historical computing task and the target computing task. In other implementations, a third center position of the hardware resources corresponding to the set of historical computing nodes can also be determined, a first distance between the first center position and the third center position and a second distance between the second center position and the third center position can be determined, and the difference between the first and second distances can be used as the first deviation value.

[0076] In this embodiment, the execution requirement includes the total number of subtasks and the proportion of the number of subtasks allowed to be executed in parallel in the total number of tasks. The step of determining the second status information representing the task difference based on the execution requirements of the multiple subtasks respectively contained in the historical computing power task and the target computing power task includes: determining the first difference between the total number of tasks corresponding to the historical computing power task and the total number of tasks corresponding to the target computing power task, determining the second difference between the proportion corresponding to the historical computing power task and the proportion corresponding to the target computing power task; determining the second deviation value representing the difference in computing power processing capabilities required for the task based on the first difference and the second difference, and the second status information includes the second deviation value. Among the multiple subtasks contained in each computing power task, subtasks that do not need to be processed in the set order are allowed to be executed in parallel, and subtasks that need to be processed in the set order are not allowed to be executed in parallel. The total number of tasks can reflect the total number of threads required to be used by the computing power node scheduled when executing the corresponding computing power task. The first difference accurately reflects the difference in the total size of subtasks required to be processed by the historical computing power task and the target computing power task, while the second difference accurately reflects the difference in the allocation rules of multiple subtasks in the computing power nodes between the historical computing power task and the target computing power task. Based on this, the second deviation value determined by combining the first and second differences can accurately reflect the difference in computing power processing capabilities required by the task.

[0077] In this embodiment, when both the historical computing power task and the target computing power task correspond to the same computing power type (e.g., both are intelligent computing power or both are general computing power), and the computing power type corresponding to the historical computing power task is the same as the computing power type corresponding to the target computing power task, then the computing power type corresponding to the historical computing power task is determined to be consistent with the computing power type corresponding to the target computing power task. When both the historical computing power task and the target computing power task correspond to more than one computing power type (e.g., including intelligent computing power and general computing power), the number of subtasks corresponding to each computing power type may be counted. If the historical computing power task and the target computing power task correspond to the same computing power type, the difference in the number of subtasks of the same computing power type in the historical computing power task and the target computing power task is determined. If the data difference corresponding to each computing power type is less than a preset threshold, then the computing power type corresponding to the historical computing power task is determined to be consistent with the computing power type corresponding to the target computing power task.

[0078] In this embodiment, the first state information, the second state information and the third state information can accurately reflect the similarity between the target computing power task and the historical computing power task. The target computing power node set determined based on this ensures that the computing power provided by the target computing power node set can accurately meet the needs of the target computing power task while ensuring the improvement of scheduling efficiency.

[0079] In a feasible embodiment, the step of determining the target computing power node set corresponding to the target computing power task based on the correlation information and the historical computing power node set includes: when the third state information includes that the computing power type of the historical computing power task is consistent with the computing power type of the target computing power task, determining the similarity between the historical computing power task and the target computing power task based on the first state information and the second state information; when the similarity is greater than a first preset threshold, determining the historical computing power node set as the target computing power node set; when the similarity is greater than a second preset threshold and less than or equal to the first preset threshold, adjusting the historical computing power node set according to the first state information and the second state information to obtain the target computing power node set; when the similarity is less than or equal to the second preset threshold, traversing all computing power nodes in the computing power node cluster, and taking the node set that meets the preset computing power scheduling strategy as the target computing power node set; wherein, the second preset threshold is less than the first preset threshold.

[0080] Here, after the above-mentioned reference task is screened out from more than one historical computing power task, the similarity between the reference task and the target computing power task can be determined based on the first state information and the second state information corresponding to the reference task. When the similarity is greater than the first preset threshold, the historical computing power node set corresponding to the reference task is determined as the target computing power node set; when the similarity is greater than the second preset threshold and less than or equal to the first preset threshold, the historical computing power node set corresponding to the reference task is adjusted according to the first state information and the second state information corresponding to the reference task to obtain the target computing power node set; when the similarity is less than or equal to the second preset threshold, all computing power nodes in the computing power node cluster are traversed, and the node set that meets the preset computing power scheduling strategy is used as the target computing power node set.

[0081] In the process of adjusting the historical computing power node set according to the first state information and the second state information, computing power nodes can be added, reduced, or replaced in the historical computing power node set. For example, when the first deviation value is greater than the first preset threshold value, the first number of computing power nodes in the historical computing power node set that are farthest away from the location of the corresponding hardware resources and the second center location can be replaced with other first computing power nodes of the same computing power type, the capacity deviation between the computing power processing capacity corresponding to the first computing power node and the computing power processing capacity of the replaced computing power node is less than the preset deviation, and the distance between the location of the hardware resources corresponding to the first computing power node and the second center location is less than the preset distance, wherein the first number can be determined based on the deviation between the first deviation value and the first preset threshold value; the second deviation value is greater than the second preset threshold value, and the computing power processing capacity of the target computing power task requirement corresponding to the second deviation value is greater than the historical computing power task requirement When the computing power processing capability is insufficient, a second number of computing power nodes with the largest computing power processing capability in the node set whose corresponding hardware resource locations in the historical computing power node set are located at a distance greater than the preset distance from the second center location can be deleted, and the second number can be determined based on the deviation between the second deviation value and the second preset threshold value; when the second deviation value is greater than the second preset threshold value, and the computing power processing capability required by the target computing power task corresponding to the second deviation value is less than the computing power processing capability required by the historical computing power task, a third number of second computing power nodes can be added to the historical computing power node set, and the distance between the location of the hardware resource corresponding to the second computing power node and the second center location is less than the preset distance, and the computing power type is the same as the original computing power type in the historical computing power node set.

[0082] The preset computing power scheduling strategy here can be a pre-set fixed strategy, or a strategy determined by user needs.

[0083] In this embodiment, the above method is used to improve the computing power scheduling efficiency, ensure efficient task execution, and ensure that the target computing power task requirements can be accurately met.

[0084] In a feasible implementation, the first state information includes a first deviation value indicating a difference in data source location, and the second state information includes a second deviation value indicating a difference in computing power processing capability required by the task. The first deviation value and the second deviation value here are determined in the manner mentioned in the above embodiment and will not be described in detail here. The step of determining the similarity between the historical computing power task and the target computing power task based on the first state information and the second state information includes: determining the priority information between the data source location and computing power processing capability corresponding to the target computing power task based on the priority information; determining the first weight corresponding to the first deviation value and the second weight corresponding to the second deviation value based on the priority information; and calculating the similarity based on the first deviation value and the corresponding first weight, and the second deviation value and the corresponding second weight.

[0085] The priority information here can be set by the user based on their needs, or determined based on the relationship between the number of subtasks corresponding to different computing power types within the target computing power task requirements. For example, when the proportion of subtasks corresponding to general computing power is greater than the proportion of subtasks corresponding to intelligent computing power, the priority information will be that the data source location takes precedence over the computing power processing power priority; when the proportion of subtasks corresponding to general computing power is less than the proportion of subtasks corresponding to intelligent computing power, the priority information will be that the data source location takes precedence over the computing power processing power priority.

[0086] When the priority information indicates that the priority of the data source location is higher than the priority of the computing power processing capability, the first weight is greater than the second weight; when the priority information indicates that the priority of the data source location is lower than the priority of the computing power processing capability, the first weight is less than the second weight.

[0087] The similarity may be obtained by performing weighted average calculation on the first deviation value and the second deviation value based on the first weight and the second weight.

[0088] In this embodiment, it is helpful to further improve the accuracy of the target computing power node set determined subsequently, thereby further ensuring the effective balance between the efficiency of computing power task execution and the accuracy of demand matching.

[0089] In a feasible implementation manner, the step of determining the target computing power node set corresponding to the target computing power task according to the correlation information and the historical computing power node set includes:

[0090] When the third state information includes that the computing power type of the historical computing power task is inconsistent with the computing power type of the target computing power task, the similarity between the historical computing power task and the target computing power task is determined based on the first state information and the second state information; when the similarity is greater than a first preset threshold, the historical computing power node set is adjusted according to the first state information, the second state information and the third state information to obtain the target computing power node set; when the similarity is less than the first preset threshold, all computing power nodes in the computing power node cluster are traversed, and the node set that meets the preset computing power scheduling strategy is used as the target computing power node set.

[0091] The same expressions as those in the above embodiments refer to the same concepts, and the same parts in the corresponding solutions can be implemented in the manner mentioned in the above embodiments, and will not be repeated here.

[0092] Among them, in the process of adjusting the historical computing power node set according to the first state information, the second state information and the third state information to obtain the target computing power node set, analogously to the process of adjusting the historical computing power node set according to the first state information and the second state information, it is necessary to ensure that the computing power type corresponding to the adjusted historical computing power node set is consistent with the computing power type required by the target computing power task.

[0093] In this embodiment, the above-mentioned method can further ensure the effective balance between the efficiency of computing task execution and the accuracy of demand matching.

[0094] It should be noted that the above examples are only used to understand this application and do not constitute a limitation on the computing task execution method of this application. More simple transformations based on this technical concept are all within the scope of protection of this application.

[0095] Reference Figure 2 The present application provides a computing power scheduling device 1, which may include: at least one processor 1001; and a memory 1002, an input / output (I / O) interface 1003, etc., which are communicatively connected to the at least one processor 1001; wherein the memory 1002 stores instructions that can be executed by the at least one processor 1001, and the instructions are executed by the at least one processor 1001 so that the at least one processor 1001 can execute the computing power task execution method in the above-mentioned embodiment 1.

[0096] Reference below Figure 2 , which shows a schematic diagram of the structure of a computing power scheduling device 1 suitable for implementing an embodiment of the present application. The computing power scheduling device 1 in the embodiment of the present application may include but is not limited to mobile terminals such as mobile phones, laptops, etc. and fixed terminals such as digital TVs, desktop computers, etc. Figure 2The computing power scheduling device 1 shown is merely an example and should not impose any limitations on the functions and scope of use of the embodiments of the present application.

[0097] like Figure 2 As shown, the computing power scheduling device 1 may include a processor 1001 (e.g., a central processing unit (CPU) or graphics processing unit (GPU)), which can perform various appropriate actions and processes based on programs stored in memory 1002. The programs in memory 1002 can be programs in read-only memory (ROM) or programs loaded from a storage device into random access memory (RAM). RAM also stores various programs and data required for the operation of the computing power scheduling device 1. The processor 1001, memory 1002 (ROM and RAM), and input / output (I / O) interface 1003 are interconnected via a communication bus 1007. Typically, the following systems can be connected to the I / O interface 1003: input devices 1004, such as a touchscreen, touchpad, keyboard, mouse, image sensor, microphone, accelerometer, and gyroscope; output devices 1005, such as a liquid crystal display (LCD), speaker, and vibrator; and communication devices 1006. The communication device 1006 can allow the computing power scheduling device 1 to communicate with other devices wirelessly or wired to exchange data. Although the figure shows the computing power scheduling device 1 with various devices, it should be understood that it is not required to implement or have all of the devices shown. More or fewer devices may be implemented or have alternatively.

[0098] In particular, according to the embodiments disclosed in this application, the method flow described in the above embodiments can be implemented as a computer software program. For example, the embodiments disclosed in this application include a computer program product, which includes a computer program carried on a computer-readable medium, and the computer program contains program code for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from the network through a communication device, or installed from the memory 1002. When the computer program is executed by the processor 1001, the above-mentioned functions defined in the computing power task execution method of the embodiment disclosed in this application are executed.

[0099] The computing power scheduling device provided in this application, using the computing power task execution method in the above-mentioned embodiment, can solve the technical problem of how to ensure that the target computing power task requirements are met while improving the efficiency of computing power scheduling to meet the requirements of efficient computing power task execution. Compared with the existing technology, the beneficial effects of the computing power scheduling device provided in this application are the same as the beneficial effects of the computing power task execution method provided in the above-mentioned embodiment, and the other technical features of the computing power scheduling device are the same as those disclosed in the method of the previous embodiment, and will not be repeated here.

[0100] The present application provides a computer-readable storage medium having computer-readable program instructions (i.e., a computer program) stored thereon, and the computer-readable program instructions are used to execute the computing power task execution method in the above-mentioned embodiment.

[0101] The computer-readable storage medium provided herein may be, for example, a USB flash drive, but is not limited to electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems, systems, or devices, or any combination thereof. More specific examples of computer-readable storage media 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 or flash memory), optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination thereof. In this embodiment, the computer-readable storage medium may be any tangible medium that contains or stores a program that can be used by or in conjunction with an instruction execution system, system, or device. The program code contained on the computer-readable storage medium may be transmitted using any suitable medium, including, but not limited to, wires, optical cables, RF (Radio Frequency), etc., or any suitable combination thereof.

[0102] The above-mentioned computer-readable storage medium may be included in the computing power scheduling device; or it may exist independently without being assembled into the computing power scheduling device.

[0103] The above-mentioned computer-readable storage medium carries one or more programs. When the above-mentioned one or more programs are executed by the computing power scheduling device, the computing power scheduling device executes the following process: obtaining task attribute information of the target computing power task, obtaining matching historical computing power tasks and a set of historical computing power nodes that execute the historical computing power tasks based on the task attribute information; determining the correlation information between the historical computing power tasks and the target computing power task; determining the target computing power node set corresponding to the target computing power task based on the correlation information and the historical computing power node set; and scheduling the computing power nodes in the target computing power node set to execute the target computing power task.

[0104] Computer program code for performing the operations of the present application may be written in one or more programming languages, or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, C++, and conventional procedural programming languages ​​such as "C" or similar programming languages. The program code may be executed entirely on the user's computer, partially on the user's computer, as a stand-alone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer via any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0105] The computer-readable storage medium provided in this application is a computer-readable storage medium that stores computer-readable program instructions (i.e., a computer program) for executing the aforementioned computing task execution method. This computer-readable storage medium can address the technical problem of ensuring that target computing task requirements are met while improving the efficiency of computing task scheduling to meet the requirements for efficient computing task execution. Compared to the prior art, the beneficial effects of the computer-readable storage medium provided in this application are the same as those of the computing task execution method provided in the aforementioned embodiments, and are not further elaborated here.

[0106] The flow charts and block diagrams in the accompanying drawings illustrate the possible architecture, functions and operations of the systems, methods and computer program products according to various embodiments of the present application. In this regard, each box in the flow chart or block diagram can represent a module, program segment or a part of code, and the module, program segment or a part of code includes one or more executable instructions for realizing the prescribed logical function. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in a sequence different from that marked in the accompanying drawings. For example, two boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram and / or flow chart, and the combination of the boxes in the block diagram and / or flow chart can be implemented using a dedicated hardware-based system that performs the prescribed function or operation, or can be implemented using a combination of dedicated hardware and computer instructions.

[0107] It should be understood that the various parts disclosed in this application can be implemented using hardware, software, firmware, or a combination thereof. The modules described in the embodiments of this application can be implemented using software or hardware. The name of a module does not, in some cases, constitute a limitation on the unit itself. In the description of the above embodiments, specific features, structures, materials, or characteristics can be combined in any appropriate manner in any one or more embodiments or examples.

[0108] The above descriptions are merely some embodiments of the present application and are not intended to limit the scope of the present application. Any equivalent structural transformations made within the technical concept of the present application using the contents of the present specification and drawings, or any direct or indirect application in other related technical fields, are included within the scope of the present application. Therefore, the scope of protection of the present application shall be based on the scope of protection of the claims.

Claims

1. A computing task execution method, characterized in that: The method includes: Obtaining task attribute information of the target computing power task, and obtaining, based on the task attribute information, matching historical computing power tasks and a set of historical computing power nodes that execute the historical computing power tasks, wherein the computing power task corresponding to the historical task attribute information that matches the task attribute information of the target computing power task is determined as the historical computing power task; Determine first state information indicating a difference in data source location based on location data of the data sources corresponding to the historical computing power task and the target computing power task, determine second state information indicating a difference in task based on execution requirements of the multiple subtasks respectively included in the historical computing power task and the target computing power task, determine third state information indicating a difference in computing power type based on the computing power types corresponding to the historical computing power task and the target computing power task, the correlation information between the historical computing power task and the target computing power task includes the first state information, the second state information, and the third state information, the correlation information indicates a degree of similarity between computing power usage states when the historical computing power task and the target computing power task are executed, and the execution requirements include the total number of subtasks and the proportion of the number of subtasks allowed to be executed in parallel in the total number of tasks; Determine the target computing power node set corresponding to the target computing power task according to the correlation information and the historical computing power node set; Scheduling the computing nodes in the target computing node set to execute the target computing task; The step of determining the target computing power node set corresponding to the target computing power task according to the correlation information and the historical computing power node set includes: When the third state information includes that the computing power type of the historical computing power task is consistent with the computing power type of the target computing power task, determining the similarity between the historical computing power task and the target computing power task according to the first state information and the second state information; When the similarity is greater than a first preset threshold, determining the historical computing power node set as the target computing power node set; When the similarity is greater than the second preset threshold and less than or equal to the first preset threshold, the historical computing power node set is adjusted according to the first state information and the second state information to obtain the target computing power node set, and the second preset threshold is less than the first preset threshold.

2. The computing task execution method according to claim 1, wherein: After the step of determining the similarity between the historical computing power task and the target computing power task according to the first state information and the second state information, the method further includes: When the similarity is less than or equal to the second preset threshold, all computing nodes in the computing node cluster are traversed, and the node set that meets the preset computing scheduling strategy is used as the target computing node set.

3. The computing task execution method according to claim 1, wherein: The first state information includes a first deviation value indicating a difference in data source locations, and the second state information includes a second deviation value indicating a difference in computing power processing capabilities required by the task. The step of determining the similarity between the historical computing power task and the target computing power task based on the first state information and the second state information includes: Priority information between the data source location and computing power processing capability corresponding to the target computing power task; Determine, according to the priority information, a first weight corresponding to the first deviation value and a second weight corresponding to the second deviation value; The similarity is calculated according to the first deviation value and the corresponding first weight, and the second deviation value and the corresponding second weight.

4. The computing task execution method according to claim 1, wherein: The step of determining the target computing power node set corresponding to the target computing power task according to the correlation information and the historical computing power node set includes: When the third state information includes that the computing power type of the historical computing power task is inconsistent with the computing power type of the target computing power task, determining the similarity between the historical computing power task and the target computing power task according to the first state information and the second state information; When the similarity is greater than a first preset threshold, adjusting the historical computing power node set according to the first state information, the second state information, and the third state information to obtain the target computing power node set; When the similarity is less than the first preset threshold, all computing nodes in the computing node cluster are traversed, and a set of nodes that meet the preset computing power scheduling strategy is used as the target computing node set.

5. The computing task execution method according to claim 1, wherein: The step of determining first status information indicating a difference in data source location based on the location data of the data sources corresponding to the historical computing power task and the target computing power task respectively comprises: Determine a first center position of all data sources corresponding to the historical computing power task, and determine a second center position of all data sources corresponding to the target computing power task; A first deviation value indicating a difference in data source locations is determined according to a distance between the first center location and the second center location, and the first state information includes the first deviation value.

6. The computing task execution method according to claim 1, wherein: The step of determining the second status information representing the task difference according to the execution requirements of the multiple subtasks respectively included in the historical computing power task and the target computing power task includes: Determine a first difference between the total number of tasks corresponding to the historical computing power tasks and the total number of tasks corresponding to the target computing power tasks, and determine a second difference between the proportion corresponding to the historical computing power tasks and the proportion corresponding to the target computing power tasks; A second deviation value representing a difference in computing power processing capability required by the task is determined based on the first difference value and the second difference value, and the second state information includes the second deviation value.

7. The computing task execution method according to any one of claims 1 to 6, characterized in that: The task attribute information includes a task application scenario and a billing method. The step of obtaining a matching historical computing task based on the task attribute information includes: Determine the computing power tasks that meet the preset conditions and are executed before the current moment as the historical computing power tasks; The preset conditions include that the corresponding task application scenario and billing method are consistent with the task application scenario and billing method of the target computing power task.

8. A computing power scheduling device, characterized in that: The computing power scheduling device includes: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the computer program is configured to implement the steps of the computing power task execution method according to any one of claims 1 to 7.

9. A storage medium, characterized in that: The storage medium is a computer-readable storage medium, and a computer program is stored on the storage medium. When the computer program is executed by the processor, the steps of the computing task execution method according to any one of claims 1 to 7 are implemented.

Images