Task processing method and device, computer device, readable storage medium and program product

Abstract

The application relates to a task processing method and device, computer equipment, a computer readable storage medium and a computer program product. The method comprises the following steps: obtaining a data processing task from an edge processing node, obtaining task demand information corresponding to each data processing task; obtaining node state information of each processing node in a computing power network; according to the node state information and the task demand information carried by the data processing task, determining a target processing node by taking the maximum resource utilization rate of each processing node in the computing power network as a target; according to the data amount size information and the time delay requirement carried by the data processing task, and the network state and the transmission protocol of the target processing node, determining a target transmission path and a target transmission protocol by taking the maximum transmission bandwidth and the minimum transmission delay as targets; and transmitting the data processing task to the target processing node through the target transmission path and the target transmission protocol. The method can improve the task processing efficiency.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of computers, and in particular to a task processing method and device, computer equipment, a computer readable storage medium, and a computer program product. BACKGROUND

[0002] With the rapid development of cloud computing, edge computing, 5G, AI, and other technologies, the distribution of computing resources is increasingly widespread and diversified, and a computing power network has emerged. The computing power network integrates widely distributed computing nodes (including cloud data centers, edge nodes, terminal devices, and the like) and shares the computing resources of different nodes. However, the efficiency of the current computing power network in task processing is low. SUMMARY

[0003] Therefore, it is necessary to provide a task processing method, device, computer equipment, computer readable storage medium, and computer program product that can improve the efficiency of task processing.

[0004] In a first aspect, the present application provides a task processing method, comprising:

[0005] obtaining data processing tasks from edge processing nodes, and obtaining task requirement information corresponding to each data processing task; the task requirement information includes data size information, computational complexity, and latency requirements;

[0006] obtaining node state information of each processing node in the computing power network, the node state information including CPU state, memory load state, and network delay information; the processing nodes include central processing nodes and edge processing nodes;

[0007] determining target processing nodes corresponding to each data processing task according to the node state information and the task requirement information carried by the data processing tasks, with the goal of maximizing the resource utilization rate of each processing node in the computing power network;

[0008] determining target transmission paths and target transmission protocols of each data processing task to the corresponding target processing nodes according to the data size information and latency requirements carried by the data processing tasks, and the network state and transmission protocol of the target processing nodes, with the goal of maximizing the transmission bandwidth and minimizing the transmission delay;

[0009] transmitting the data processing tasks to the target processing nodes through the target transmission paths and target transmission protocols.

[0010] In one of the embodiments, the target processing node corresponding to each of the data processing tasks is determined according to the node state information and the task requirement information carried by the data processing task, with the maximum resource utilization of each processing node in the computing power network as the target.

[0011] The CPU state, the memory load state, the network delay information, the computing complexity, and the latency requirement are input into a pre-trained task scheduling model, with the maximum resource utilization of each processing node in the computing power network as the target, to obtain the target processing node corresponding to each of the data processing tasks.

[0012] In one of the embodiments, the target transmission path and the target transmission protocol of each of the data processing tasks to the target processing node are determined according to the data size information and the latency requirement carried by the data processing task, and the network state and the transmission protocol of the target processing node, with the maximum transmission bandwidth and the minimum transmission delay as the target.

[0013] The data size information and the latency requirement carried by the data processing task, and the network state and the transmission protocol of the target processing node are input into a pre-trained transmission path and protocol selection model, with the maximum transmission bandwidth and the minimum transmission delay as the target, to obtain the target transmission path and the target transmission protocol of each of the data processing tasks to the target processing node.

[0014] In one of the embodiments, the transmission protocol at least includes one of a QUIC protocol and an RDMA protocol.

[0015] In one of the embodiments, the data corresponding to the data processing task at least includes one of sensor data and a video stream.

[0016] In one of the embodiments, the task requirement information further includes a priority.

[0017] The data processing task is transmitted to the target processing node through the target transmission path and the target transmission protocol.

[0018] The priority corresponding to the data processing task is obtained.

[0019] The data processing task is transmitted to the target processing node through the target transmission path and the target transmission protocol according to the priority corresponding to the data processing task.

[0020] In a second aspect, the application further provides a task processing device, which comprises:

[0021] The task acquisition module is configured to acquire data processing tasks from the edge processing nodes and acquire task requirement information corresponding to each data processing task. The task requirement information includes data size information, computational complexity, and time delay requirement.

[0022] The node state acquisition module is configured to acquire node state information of each processing node in the computing power network. The node state information includes CPU state, memory load state, and network delay information. The processing nodes include central processing nodes and edge processing nodes.

[0023] The processing node determination module is configured to determine target processing nodes corresponding to each data processing task according to the node state information and the task requirement information carried by the data processing tasks, with the maximum resource utilization rate of each processing node in the computing power network as the target.

[0024] The transmission mode determination module is configured to determine target transmission paths and target transmission protocols of each data processing task to the target processing nodes according to the data size information and time delay requirement carried by the data processing tasks, the network state of the target processing nodes, and the transmission protocol, with the maximum transmission bandwidth and the minimum transmission delay as the target.

[0025] The task transmission module is configured to transmit the data processing tasks to the target processing nodes through the target transmission paths and target transmission protocols.

[0026] In a third aspect, the present application further provides a computer device including a memory and a processor. The memory stores a computer program. When the processor executes the computer program, the steps of the method in the first aspect are implemented.

[0027] In a fourth aspect, the present application further provides a computer readable storage medium having a computer program stored thereon. When the computer program is executed by a processor, the steps of the method in the first aspect are implemented.

[0028] In a fifth aspect, the present application further provides a computer program product including a computer program. When the computer program is executed by a processor, the steps of the method in the first aspect are implemented.

[0029] The task processing method, device, computer equipment, computer readable storage medium and computer program product can obtain data processing tasks from an edge processing node by a scheduling server, obtain task demand information corresponding to each data processing task, wherein the task demand information comprises data size information, calculation complexity and time delay requirement; obtain node state information of each processing node in a computing power network, wherein the node state information comprises CPU state, memory load state and network delay information; the processing node comprises a central processing node and an edge processing node; according to the node state information and the task demand information carried by the data processing task, a target processing node corresponding to each data processing task is determined, with the maximum resource utilization rate of each processing node in the computing power network as the target; according to the data size information and time delay requirement carried by the data processing task, and the network state and transmission protocol of the target processing node, a target transmission path and a target transmission protocol of each data processing task to the corresponding target processing node are determined, with the maximum transmission bandwidth and the minimum transmission delay as the target; and the data processing task is transmitted to the target processing node through the target transmission path and the target transmission protocol. In the present application, by real-time acquisition and analysis of node state information (such as CPU state, memory load state and network delay), the most suitable processing node can be effectively selected to process the data processing task, the data processing task can be flexibly distributed, the specific transmission path and protocol can be determined according to the data size, calculation complexity and time delay requirement, the data transmission process is smoother, the processing time is effectively reduced, and the overall task execution efficiency is improved, and the resource utilization rate and response speed of the computing power network are improved. BRIEF DESCRIPTION OF DRAWINGS

[0030] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the related art, the drawings needed to be used in the description of the embodiments of the present application or the related art will be briefly introduced. Obviously, the drawings in the following description are only some embodiments of the present application, and other related drawings can be obtained by those skilled in the art without creative labor.

[0031] Figure 1 A flowchart of a task processing method in an embodiment;

[0032] Figure 2 A flowchart of a task processing method in another embodiment;

[0033] Figure 3 A block diagram of a task processing device in an embodiment;

[0034] Figure 4 An internal structure diagram of a computer equipment in an embodiment. DETAILED DESCRIPTION

[0035] In order to make the purposes, technical solutions and advantages of the present application clearer, further detailed description will be given below in combination with the drawings and examples. It should be understood that the specific examples described herein are only used to explain the present application and not to limit the present application.

[0036] In one exemplary embodiment, as shown in Figure 1 A task processing method is provided, which is applied to a scheduling server in a computing power network. The scheduling server can be any processing node in the computing power network or an additional server. The task processing method includes steps S202 to S210. In the steps:

[0037] Step S202, obtaining data processing tasks from edge processing nodes, and obtaining task demand information corresponding to each data processing task.

[0038] The task demand information includes data size information, computational complexity and latency requirement. The data size information can refer to the size of the input data involved in the task. The computational complexity can refer to the computational resources required for task execution, which is related to the complexity of the algorithm. The latency requirement can refer to the time within which the task must be completed, which can refer to the acceptable delay range.

[0039] The computing power network can refer to a network composed of multiple processing nodes, such as servers, edge devices, etc., which can provide computing and storage capabilities. The computing power network can allow tasks to be allocated and executed between nodes at different locations to improve resource utilization efficiency and system response speed. The scheduling server can be used to manage and allocate data processing tasks. The data processing task can refer to a specific unit of work that needs to be executed, which involves a request to process certain input data. These tasks can include data analysis, computational tasks or other computationally intensive operations. The edge processing node can refer to a type of computing node close to the data source or user, which can process data locally to reduce latency and bandwidth usage.

[0040] Exemplarily, the scheduling server obtains data processing tasks from edge processing nodes, which can be derived from user requests or device monitoring processes. The scheduling server obtains task demand information corresponding to each data processing task.

[0041] Step S204, obtaining node state information of each processing node in the computing power network;

[0042] The node state information includes CPU state, memory load state and network delay information; the processing node includes a central processing node and an edge processing node. The CPU state can refer to the load or usage of the processor, indicating the computing power available to the node. The memory load state can refer to the usage of the current memory resources of the node, including used memory and available memory. The network delay information can refer to the delay of communication between nodes, affecting the speed and performance of data transmission.

[0043] Exemplarily, the scheduling server obtains the state information of all processing nodes (including central nodes and edge nodes) in the computing power network in real time; the CPU state, memory load state and network delay information can be obtained through a monitoring tool or a self-reporting mechanism.

[0044] In step S206, according to the node state information and the task requirement information carried by the data processing task, the target processing node corresponding to each data processing task is determined, with the maximum resource utilization rate of each processing node in the computing power network as the target.

[0045] The resource utilization rate can refer to the usage of resources (such as CPU, memory, bandwidth) of the processing node when executing the task. The target processing node can refer to the processing node most suitable for executing a specific data processing task, which is assigned to the task. The current state of the node and the task requirement are considered when selecting the target processing node to improve processing efficiency.

[0046] Exemplarily, the scheduling server determines the target processing node most suitable for executing each data processing task according to the task requirement information carried by the data processing task and the node state information of each node, and then the scheduling server can use an algorithm (such as a load balancing algorithm) to determine the target processing node most suitable for executing each data processing task, to maximize the resource utilization rate of each processing node in the computing power network.

[0047] In step S208, according to the data size information and the delay requirement carried by the data processing task, and the network state and transmission protocol of the target processing node, the target transmission path and the target transmission protocol of each data processing task to the corresponding target processing node are determined, with the maximum transmission bandwidth and the minimum transmission delay as the target.

[0048] The maximum transmission bandwidth can refer to optimizing the data transmission path to maximize the bandwidth capacity of the network, thereby speeding up the data transmission speed; the minimum transmission delay can refer to optimizing the data transmission process to reduce the time delay between the source node and the target node, thereby improving the system response speed; the target transmission path can refer to the actual network path of data transmission from the source node to the target processing node, i.e. the network devices and nodes through which the data packet passes. The target transmission protocol can refer to the communication protocol used in the data transmission process, such as TCP / IP, UDP, QUIC, RDMA, etc.

[0049] Exemplarily, after determining the target processing node, the scheduling server determines the target transmission path and the target transmission protocol of each data processing task to the corresponding target processing node according to the data amount size information and the time delay requirement carried by the data processing task, and the network state and the transmission protocol of the target processing node, so as to maximize the transmission bandwidth and minimize the transmission delay.

[0050] In step S210, the data processing task is transmitted to the target processing node through the target transmission path and the target transmission protocol.

[0051] Exemplarily, the scheduling server transmits the data processing task and the related data to the target processing node through the target transmission path and the target transmission protocol.

[0052] In the above task processing method, by acquiring and analyzing the node state information (such as CPU state, memory load and network delay) in real time, the method can effectively select the most suitable processing node to process the data processing task, realize flexible allocation of data processing tasks, and determine the specific transmission path and protocol according to the data amount, calculation complexity and time delay requirement, so that the data transmission process is smoother, the processing time is effectively reduced, and the overall task execution efficiency is improved, and the resource utilization rate and response speed of the computing power network are improved.

[0053] The task processing method effectively improves the efficiency and reliability of data processing in the computing power network through intelligent scheduling, dynamic monitoring, load balancing and optimized transmission path, reduces the delay and task completion time, and maximizes the utilization rate of resources.

[0054] In an exemplary embodiment, the above step S206 can specifically include: inputting the CPU state, the memory load state, the network delay information, the calculation complexity and the time delay requirement into a pre-trained task scheduling model, to obtain the target processing node corresponding to each data processing task, with the maximum resource utilization rate of each processing node in the computing power network as the target.

[0055] The pre-trained task scheduling model can be an algorithm model based on machine learning or reinforcement learning, which aims to intelligently allocate data processing tasks to appropriate processing nodes according to the input information, i.e. CPU state, memory load state, network delay information, calculation complexity and time delay requirement. The model can be trained by historical data to achieve the task scheduling goal of maximizing the resource utilization rate of each processing node in the computing power network.

[0056] Exemplarily, the scheduling server inputs the CPU state, the memory load state, the network delay information, the calculation complexity, and the time delay requirement into a pre-trained task scheduling model, to obtain the target processing node corresponding to each data processing task, with the maximum resource utilization of each processing node in the computing power network as the target.

[0057] In this embodiment, by inputting the node state information and the task requirement information into the trained task scheduling model, the target processing node is obtained, the maximum resource utilization in the computing power network is realized, and the execution efficiency of the data processing task and the overall performance of the computing power network are improved.

[0058] In one exemplary embodiment, the above step S208 can specifically include: inputting the data amount size information and the time delay requirement carried by the data processing task, and the network state and the transmission protocol of the target processing node into a pre-trained transmission path and protocol selection model, to obtain the target transmission path and the target transmission protocol of each data processing task to the corresponding target processing node, with the maximum transmission bandwidth and the minimum transmission delay as the target.

[0059] The pre-trained transmission path and protocol selection model can be a model based on machine learning or optimization algorithm, which is used to predict the optimal transmission path and transmission protocol, i.e., to maximize the transmission bandwidth and minimize the transmission delay.

[0060] Exemplarily, the scheduling server inputs the data amount size information and the time delay requirement carried by the data processing task, and the network state and the transmission protocol of the target processing node into a pre-trained transmission path and protocol selection model, to obtain the target transmission path and the target transmission protocol of each data processing task to the corresponding target processing node, with the maximum transmission bandwidth and the minimum transmission delay as the target.

[0061] In this embodiment, by using the pre-trained transmission path and protocol selection model, the optimal target transmission path and target transmission protocol are selected to maximize the transmission bandwidth and minimize the transmission delay, which can optimize the transmission path of data between different nodes, reduce network delay, improve bandwidth utilization, and reduce data transmission bottleneck.

[0062] In an example embodiment, the transmission protocol includes at least one of a QUIC protocol and an RDMA protocol. The QUIC (Quick UDP Internet Connections) protocol can refer to a protocol for reducing latency and increasing transmission speed. The RDMA (Remote Direct Memory Access) protocol can refer to a protocol for allowing direct data transfer from memory to memory, reducing latency. In this embodiment, the improved transmission protocol (such as the QUIC protocol, the RDMA protocol, etc.) can achieve lower latency, higher bandwidth and throughput, enhanced reliability, and better dynamic adaptation.

[0063] In an example embodiment, the data corresponding to the data processing task includes at least one of sensor data and video stream. The sensor data can be information collected from various sensors (such as temperature sensors, humidity sensors, acceleration sensors, gas sensors, etc.). For example, commonly used in the fields of environmental monitoring of Internet of Things (IoT) devices, smart home, industrial automation, agricultural monitoring, etc. The sensor data can include real-time data streams such as temperature, humidity, light, pressure, etc.; it can also be data generated by various sensors on an autonomous vehicle. The video stream can refer to dynamic visual information transmitted by a series of continuous image frames, such as video streams generated in scenarios such as security monitoring, traffic monitoring, video conferencing systems, streaming media services, etc., and for example, video stream data collected by cameras on an autonomous vehicle.

[0064] In an example embodiment, the task requirement information further includes a priority; for example Figure 2 As shown, transmitting the data processing task to the target processing node through the target transmission path and the target transmission protocol can specifically include:

[0065] Step S302, obtaining the priority corresponding to the data processing task.

[0066] Step S304, transmitting the data processing task to the target processing node through the target transmission path and the target transmission protocol according to the priority corresponding to the data processing task.

[0067] The priority can refer to a priority set in advance for the type of data processing task, for example, sensor data is a first priority task, video stream is a second priority task, for example, sensor data is a first priority task, video stream is a second priority task, and the first priority is higher than the second priority.

[0068] Exemplarily, the scheduling server obtains a priority corresponding to the data processing task, and transmits the data processing task to a target processing node through a target transmission path and a target transmission protocol according to the priority corresponding to the data processing task.

[0069] In this embodiment, by setting different data processing task priorities, important tasks can be ensured to be processed first. The scheduling server makes decisions according to the priorities of the tasks when transmitting data, thereby optimizing the use of transmission resources and ensuring that high-priority tasks can reach the target processing node faster.

[0070] The prior art has the following technical problems in the application of the computing power network: data transmission delay and insufficient bandwidth utilization, insufficient flexibility and efficiency of computing task scheduling, lack of cross-level cooperative optimization mechanism, and insufficient cooperation between edge computing nodes and center nodes.

[0071] To solve the technical problems of the prior art in the application of the computing power network, in an exemplary embodiment, a scheduling server is provided, which can be composed of the following four modules: a data generation module responsible for task generation, data input and processing. A transmission optimization module responsible for efficient data transmission, including transmission protocol optimization and data flow control. A computing scheduling module that intelligently allocates computing tasks, balances the load of each computing node, and considers network conditions and node resources. A task execution and feedback module that executes scheduled tasks and returns the calculation results to the source or downstream nodes.

[0072] Based on the above scheduling server, a task processing method can be implemented in the computing power network, wherein the computing power network includes edge nodes and center nodes. The edge nodes are close to the data source and perform simple or real-time tasks with high requirements. The center nodes have strong computing power and are suitable for processing complex computing tasks. The scheduling server is a global control that monitors network status, node load, optimizes task scheduling and transmission path; the scheduling server can be any node in the computing power network or an additional server. The above task processing method can specifically include:

[0073] Step S1, the scheduling server obtains input data and generates data processing tasks. The input data can be generated by a terminal device or an edge node. The data can come from various real-time application scenarios, such as sensor data in autonomous driving, video streams collected by edge devices, etc. Each task is accompanied by data size, delay requirement and other parameters; specific scenarios can include: the camera on the autonomous vehicle continuously generates video stream data as input data and transmits it to the edge node or the center computing node.

[0074] Step S2, computing task scheduling. The scheduling server monitors the network and the computing resource status of each node (node status information) in real time, and assigns tasks to appropriate nodes (edge or center) through intelligent scheduling algorithms. The scheduling decision considers the computing complexity of the task, the delay requirement, the load of each node, and the network condition. Among them, the node load can be predicted through an algorithm based on reinforcement learning or deep learning, and the appropriate node is selected in combination with the task demand and network delay condition. Specific scenarios can include: the scheduling server assigns computationally complex video processing tasks to cloud computing centers, while real-time simple analysis tasks are assigned to edge nodes. The intelligent scheduling algorithm used in this embodiment can be a dynamic task scheduling algorithm based on reinforcement learning, which dynamically adjusts task allocation by continuously learning the load of the node and the task execution effect. The algorithm steps are as follows:

[0075] Input: task demand (computing complexity, delay requirement), node status (CPU, memory load, network delay, etc.).

[0076] Output: optimal node allocation strategy.

[0077] Training process: by simulating the task execution in the environment, the scheduling strategy is continuously adjusted to maximize resource utilization.

[0078] Step S3, transmission path selection. After the data processing task is generated, the transmission optimization module calculates the best transmission path and protocol selection according to the real-time network state and the optimization algorithm of the transmission protocol (such as QUIC, RDMA), maximizes the transmission bandwidth, and minimizes the transmission delay. For example, an optimization algorithm (such as an AI algorithm based on path prediction) can be used to select the transmission path. Specific scenarios can include: a vehicle transmits a video stream to a nearby edge node, and transmits it to a cloud server through an optimized transmission protocol and path to ensure low delay.

[0079] Step S4, transmission task. The data processing task is transmitted between different nodes through the optimized transmission path. Through the RDMA technology, fast data transmission between memories is realized, and the delay of system calls in the transmission process is reduced. For example, QUIC or RDMA protocol is used for sending and receiving data packets, and flow control is performed according to the network state. Specific scenarios can include: the edge node transmits a video stream to a central server through the QUIC protocol for deep learning model calculation.

[0080] After the task is assigned to the appropriate node in step S4, the computing task is executed on the node, generating results. The edge node feeds back the results to the terminal or downstream node after completing the real-time processing task; the results of complex calculations are fed back by the center node. Specific scenarios can include: the edge node returns the analysis results to the vehicle terminal after completing preliminary video processing to help the vehicle make real-time decisions; the results of complex video processing are processed by the center node and fed back to the system.

[0081] The timing of this embodiment can include: data generation phase: the terminal device or edge node collects real-time data and triggers computing tasks. Computing task scheduling phase: the scheduling server monitors network conditions and node resources in real time and dynamically allocates tasks. Transmission path selection phase: the scheduling server selects the optimal transmission path and performs data transmission, and the data starts to be transmitted to the processing node. Task execution and feedback phase: the computing node executes the task, and the edge or center node generates the result and returns it to the requester. For example, terminal device generates data → edge node → scheduling server decides transmission path → data transmission → center node / edge node → executes task → result feedback.

[0082] The above embodiment introduces an end-to-end collaborative optimization mechanism for data transmission and computing scheduling, which breaks through the limitations of independent optimization of transmission and computing in existing technologies by combining improved transmission protocols (such as optimization based on QUIC or RDMA) and intelligent dynamic computing task scheduling algorithms. This scheme can adjust data transmission paths and computing task allocation in real time according to network status, node load and task demand, achieve efficient collaboration in cross-regional and multi-node environments, and significantly improve the resource utilization rate of the computing network and the overall performance of the system. Through the data transmission optimization and computing task scheduling collaborative mechanism, the following technical effects are achieved:

[0083] 1)Improve data transmission efficiency: By introducing improved transmission protocols (such as QUIC, RDMA) and intelligent transmission path selection algorithms, the transmission path of data between different nodes is optimized, network latency is reduced, and bandwidth utilization is improved, especially in cross-regional and complex network environments, significantly reducing data transmission bottlenecks.

[0084] 2)Improve the resource utilization rate of computing tasks: Through intelligent dynamic scheduling algorithms, tasks are allocated to the most suitable computing nodes (edge nodes or center nodes) in real time according to task demand, node load and network status, achieving efficient utilization of computing resources and avoiding resource waste and task execution delays.

[0085] 3) Realize end-to-end collaborative optimization: break the traditional mode of independent optimization of data transmission and computing scheduling, establish an end-to-end collaborative optimization mechanism, which can dynamically adjust between data transmission and computing task allocation, maximize the performance of the entire system, and adapt to the needs of large-scale, heterogeneous computing power network.

[0086] 4) Reduce system response time: through the cooperative working mechanism of edge computing and center computing node, the tasks with high real-time requirements are preferentially allocated to the edge node, while the complex tasks are transmitted to the center node for processing, reducing the overall delay of computing tasks and improving the response speed of the system.

[0087] It should be understood that although each step in the flowchart involved in each embodiment as described above is displayed in sequence according to the arrow, these steps are not necessarily executed in the order indicated by the arrow. Unless otherwise specified herein, the execution of these steps is not strictly limited in sequence, and these steps can be executed in other orders. Moreover, as described above, at least part of the steps in the flowchart involved in each embodiment can include multiple steps or stages, which are not necessarily executed at the same time, but can be executed at different times, and the execution order of these steps or stages is not necessarily sequential, but can be alternately executed with other steps or steps or stages in other steps.

[0088] Based on the same inventive concept, the embodiments of the present application also provide a task processing device for implementing the above-mentioned task processing method. The implementation scheme for solving the problem provided by the device is similar to the implementation scheme described in the above method, so the specific limitations in one or more task processing device embodiments provided below can refer to the limitations of the task processing method described above, which will not be repeated here.

[0089] In one exemplary embodiment, as shown in Figure 3 A task processing device 900 is provided, comprising a task acquisition module 901, a node state acquisition module 902, a processing node determination module 903, a transmission mode determination module 904, and a task transmission module 905, wherein:

[0090] The task acquisition module 901 is configured to acquire data processing tasks from edge processing nodes, and acquire task demand information corresponding to each data processing task; the task demand information includes data size information, computing complexity, and latency requirement.

[0091] The node state acquisition module 902 is configured to acquire node state information of each processing node in the computing power network, the node state information including CPU state, memory load state, and network delay information; and the processing node includes a central processing node and an edge processing node.

[0092] The processing node determination module 903 is configured to determine a target processing node corresponding to each data processing task according to the node state information and task requirement information carried by the data processing task, with maximization of resource utilization of each processing node in the computing power network as a target.

[0093] The transmission mode determination module 904 is configured to determine a target transmission path and a target transmission protocol of each data processing task to the target processing node according to data size information and latency requirement carried by the data processing task, network state and transmission protocol of the target processing node, with maximization of transmission bandwidth and minimization of transmission delay as a target.

[0094] The task transmission module 905 is configured to transmit the data processing task to the target processing node through the target transmission path and the target transmission protocol.

[0095] In an exemplary embodiment, the processing node determination module 903 is further configured to input the CPU state, the memory load state, the network delay information, the computing complexity, and the latency requirement into a pre-trained task scheduling model, to obtain the target processing node corresponding to each data processing task, with maximization of resource utilization of each processing node in the computing power network as a target.

[0096] In an exemplary embodiment, the transmission mode determination module 904 is further configured to input the data size information and the latency requirement carried by the data processing task, and the network state and the transmission protocol of the target processing node into a pre-trained transmission path and protocol selection model, to obtain the target transmission path and the target transmission protocol of each data processing task to the target processing node, with maximization of transmission bandwidth and minimization of transmission delay as a target.

[0097] In an exemplary embodiment, the task requirement information further includes a priority; and the task transmission module 905 is further configured to acquire the priority corresponding to the data processing task, and transmit the data processing task to the target processing node through the target transmission path and the target transmission protocol according to the priority corresponding to the data processing task.

[0098] The various modules in the task processing apparatus can be implemented wholly or partially by software, hardware, and combinations thereof. The various modules can be embedded in or independent of a processor in the computer device in hardware form, or stored in a memory in the computer device in software form, so as to be invoked and executed by the processor to perform operations corresponding to the various modules.

[0099] In an exemplary embodiment, a computer device, which can be a server, has an internal structure diagram as shown in Figure 4 The computer device includes a processor, a memory, an input / output interface, and a communication interface. The processor, the memory, and the input / output interface are connected through a system bus, and the communication interface is connected to the system bus through the input / output interface. The processor of the computer device is configured to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program, and a database. The internal memory provides an environment for running the operating system and the computer program in the non-volatile storage medium. The input / output interface of the computer device is configured to exchange information between the processor and external devices. The communication interface of the computer device is configured to communicate with external terminals through network connection. The computer program is executed by the processor to implement a task processing method.

[0100] Those skilled in the art can understand that Figure 4 The structure shown in the above

[0101] In an embodiment, a computer device is also provided, which includes a memory and a processor. The memory stores a computer program, and the processor executes the computer program to implement the steps in the above method embodiments.

[0102] In an embodiment, a computer readable storage medium is provided, which stores a computer program. The computer program is executed by a processor to implement the steps in the above method embodiments.

[0103] In an embodiment, a computer program product is provided, which includes a computer program. The computer program is executed by a processor to implement the steps in the above method embodiments.

[0104] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a non-volatile computer readable storage medium, and when executed, can include the processes of the above-mentioned embodiment methods. Any reference to memory, database or other medium used in the embodiments provided in the present application can include at least one of non-volatile memory and volatile memory. The non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical storage, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetoresistive random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. The volatile memory can include random access memory (RAM) or external cache memory, etc. As an illustration but not limitation, the RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM), etc. The database involved in the embodiments provided in the present application can include at least one of a relational database and a non-relational database. The non-relational database can include a distributed database based on a block chain, etc., without being limited thereto. The processor involved in the embodiments provided in the present application can be a general-purpose processor, a central processing unit, a graphics processing unit, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, an artificial intelligence (AI) processor, etc., without being limited thereto.

[0105] The technical features of the above embodiments can be combined in any manner. To make the description concise, all possible combinations of the technical features in the above embodiments are not described, but as long as the combinations of the technical features do not exist, they should be considered as the scope of the present application.

[0106] The above-described embodiments are merely illustrative of several embodiments of the present application, and the description is relatively specific and detailed, but should not be understood as a limitation on the scope of the patent. It should be noted that for those skilled in the art, without departing from the concept of the present application, a number of modifications and improvements can be made, which are all within the scope of the present application. Therefore, the scope of protection of the present application should be subject to the appended claims.

Claims

1. A task processing method characterized by, The application discloses a scheduling server applied to a computing power network. Obtaining data processing tasks from edge processing nodes, and obtaining task demand information corresponding to each data processing task; the task demand information comprises data size information, calculation complexity and time delay requirement; Obtaining node state information of each processing node in the computing power network, wherein the node state information comprises CPU state, memory load state and network delay information; the processing node comprises a central processing node and an edge processing node; According to the node state information and the task demand information carried by the data processing task, determining a target processing node corresponding to each data processing task, with the maximum resource utilization rate of each processing node in the computing power network as the target; According to the data size information and the time delay requirement carried by the data processing task, and the network state and the transmission protocol of the target processing node, determining a target transmission path and a target transmission protocol of each data processing task to the corresponding target processing node, with the maximum transmission bandwidth and the minimum transmission delay as the target; Transmitting the data processing task to the target processing node through the target transmission path and the target transmission protocol.

2. The method of claim 1, wherein, The method comprises the following steps: Inputting the CPU state, the memory load state, the network delay information, the calculation complexity and the time delay requirement into a pre-trained task scheduling model, with the maximum resource utilization rate of each processing node in the computing power network as the target, to obtain the target processing node corresponding to each data processing task.

3. The method of claim 1, wherein, The method comprises the following steps: Inputting the data size information and the time delay requirement carried by the data processing task, and the network state and the transmission protocol of the target processing node into a pre-trained transmission path and protocol selection model, with the maximum transmission bandwidth and the minimum transmission delay as the target, to obtain the target transmission path and the target transmission protocol of each data processing task to the corresponding target processing node.

4. The method of claim 1, wherein, The transmission protocol at least comprises one of a QUIC protocol and an RDMA protocol.

5. The method of claim 1, wherein, The data corresponding to the data processing task at least comprises one of sensor data and a video stream.

6. The method of claim 1, wherein, The task demand information further comprises a priority. The method comprises the following steps: Obtaining the priority corresponding to the data processing task; Transmitting the data processing task to the target processing node through the target transmission path and the target transmission protocol according to the priority corresponding to the data processing task.

7. A task processing apparatus characterized by comprising: The device comprises: A task acquisition module is configured to acquire data processing tasks from edge processing nodes and acquire task requirement information corresponding to each data processing task, wherein the task requirement information includes data size information, computational complexity, and time delay requirement; A node state acquisition module is configured to acquire node state information of each processing node in the computing power network, wherein the node state information includes CPU state, memory load state, and network delay information, and the processing node includes a central processing node and an edge processing node; A processing node determination module is configured to determine a target processing node corresponding to each data processing task according to the node state information and the task requirement information carried by the data processing task, and maximize resource utilization of each processing node in the computing power network as a target; A transmission mode determination module is configured to determine a target transmission path and a target transmission protocol of each data processing task to the target processing node according to the data size information and the time delay requirement carried by the data processing task, the network state of the target processing node, and a transmission protocol, and maximize transmission bandwidth and minimize transmission delay as a target; A task transmission module is configured to transmit the data processing task to the target processing node through the target transmission path and the target transmission protocol.

8. A computer device comprising a memory and a processor, the memory storing a computer program, characterized in that, The processor executes the computer program to realize the steps of the method in any one of claims 1 to 6.

9. A computer-readable storage medium having stored thereon a computer program, characterized in that, The computer program is executed by the processor to realize the steps of the method in any one of claims 1 to 6.

10. A computer program product comprising a computer program, characterized in that, The computer program is executed by the processor to realize the steps of the method in any one of claims 1 to 6.

Images