Tidal computing power scheduling method and system, electronic equipment and storage medium

Abstract

The invention provides a tidal computing power scheduling method and system, electronic equipment and a storage medium, and aims to solve the problem of edge computing task abnormal risk caused by tidal computing power change, and the method comprises the following steps: establishing an edge computing power resource pool, obtaining resource information of computing power nodes in each computing power domain, and calculating computing power values of the computing power nodes to obtain an edge computing power resource pool; storing in an edge computing power resource pool; performing time-varying feature division on the computing power nodes in the resource pool to determine tide computing power nodes in the resource pool and corresponding tide periods; receiving service request information of a computing power demand side, calculating the magnitude of the required computing power, and pre-estimating the duration required for completing a computing power task; the computing power value of the computing power node is larger than the computing power needed by the task, and the computing power node does not coincide with the tidal time period within the time range covered by the task duration. The computing power node serves as a target computing power node. According to the method, the resource occupation peak value of the computing power node can be avoided during computing power distribution, and the computing power demand can be matched more reasonably.

Description

Technical Field

[0001] The present disclosure relates to the field of communication technology, and in particular to a tidal computing power scheduling method, a tidal computing power scheduling system, an electronic device, and a computer-readable storage medium. Background Art

[0002] Cloud computing resources are often located far from endpoints, making them inadequate for applications requiring low latency. This results in high computing power, high energy consumption, and high network latency. In contrast, edge computing devices, located close to data sources and integrating computing, storage, networking, and AI capabilities, offer advantages over cloud computing, such as low latency, high bandwidth, local packet processing, and local security.

[0003] Modern industrial production environments often require higher computing power and lower latency, making edge computing the preferred solution. However, existing edge computing scheduling technologies are often limited to edge computing power matching, simply providing sufficient computing power to ensure computing tasks are completed by the demander. This approach ignores the risks associated with potential fluctuations in the available computing power of edge computing nodes. This single-dimensional computing resource scheduling approach is generally suboptimal.

[0004] From the perspective of edge computing scheduling strategies, edge computing users currently primarily develop scheduling systems based on their own business characteristics to meet their needs. These users often use generic, static scheduling strategies provided by third parties. This static scheduling strategy fails to account for the dynamic changes in computing power demanders and providers, and in extreme cases, can prevent edge computing tasks from completing properly. Summary of the Invention

[0005] In order to at least solve the above-mentioned technical problems existing in the prior art, the present disclosure provides a tidal computing power scheduling method, a tidal computing power scheduling system, an electronic device and a computer-readable storage medium, which can avoid the resource occupancy peak of computing power nodes when allocating computing power, so as to more reasonably match the computing power demand, improve the one-time completion rate of edge computing power tasks, avoid the failure of computing tasks due to fluctuations in the available computing power resources of edge computing power nodes, and then need to perform secondary edge computing power resource allocation, which can better improve the efficiency of edge computing.

[0006] In a first aspect, the present disclosure provides a tidal computing power scheduling method, the method comprising:

[0007] Establish an edge computing resource pool, obtain resource information of computing nodes in each computing domain, calculate the computing power value of the computing nodes based on the resource information, and save the resource information and computing power value of each computing node to the edge computing resource pool;

[0008] The computing nodes in the edge computing resource pool are divided by time-varying characteristics to determine the tidal computing nodes in the edge computing resource pool and the tidal periods corresponding to the tidal computing nodes;

[0009] Receive business request information from computing power demanders, calculate the computing power required for the business request, and estimate the time required to complete the computing power task;

[0010] Find a computing node from the computing power resource pool whose computing power value is greater than the computing power required by the task, and whose task duration covers a time range that does not overlap with the tidal period, and use it as the target computing power node for this task.

[0011] Further,

[0012] The resource information includes the device identification, the device CPU (Central Processing Unit) / GPU (Graphics Processing Unit) main frequency, the number of cores, the device memory and the network bandwidth

[0013] Calculating the computing power value of the computing power node according to the resource information includes:

[0014] The computing power value of the computing power node is calculated according to the CPU / GPU parameters to obtain the quantified computing power value.

[0015] Further,

[0016] The time-varying feature division of computing nodes in the edge computing resource pool includes:

[0017] The historical resource occupancy of each computing node in the edge computing resource pool is counted one by one, and a first threshold is set. When the average resource occupancy rate of a computing node in a certain period is greater than the first threshold, it is identified as a tidal computing node in the edge computing resource pool, and the corresponding period is identified as the tidal period of the tidal computing node;

[0018] The method further comprises:

[0019] After determining whether the computing power node is a tidal computing power node and the corresponding tidal period, the characteristic information of all computing power nodes in the edge computing power resource pool is determined. The characteristic information includes device identification, computing power value, whether it is a tidal computing power node and tidal period.

[0020] Furthermore, the method further comprises:

[0021] The average resource utilization of computing nodes is calculated using the following formula:

[0022] Resource utilization rate = α*A+β*B+γ*C

[0023] Among them, α, β, and γ are the weights of CPU, memory, and network bandwidth in the resource usage calculation, respectively. A is the CPU usage of the computing power node, B is the memory usage of the computing power node, and C is the network bandwidth usage of the computing power node.

[0024] Furthermore, the calculation of the computing power required for the business request and the estimation of the time required to complete the computing task include:

[0025] Convert the computing power required for the business request into FLOPS (Floating-point operations per second).

[0026] Categorize computing power based on business request information from computing power demanders into tasks with fixed computing capacity and tasks with unpredictable duration.

[0027] For tasks with fixed computing amount, the task duration is calculated by the computing amount / computing power. For tasks with unpredictable duration, the task duration is set to unpredictable.

[0028] Furthermore, the method further comprises:

[0029] By analyzing the tidal period T associated with the tidal computing power node x [t i , t j ] Set the pre-switching time period Δt and pre-switching time t for the tidal computing power node i -Δt;

[0030] After the pre-switching time is reached, the task mirror of the task in the tidal computing power node is sent to the target computing power node for calculation.

[0031] In a second aspect, the present disclosure provides a tidal computing power scheduling system, the system comprising:

[0032] A resource pool module is configured to establish an edge computing resource pool, obtain resource information of computing nodes within each computing domain, calculate the computing power value of the computing nodes based on the resource information, and save the resource information and computing power value of each computing node to the edge computing resource pool;

[0033] A partitioning module is configured to perform time-varying feature partitioning on the computing nodes in the edge computing resource pool to determine the tidal computing nodes in the edge computing resource pool and the tidal periods corresponding to the tidal computing nodes;

[0034] The demand analysis module is configured to receive business request information from computing power demanders, calculate the amount of computing power required for the business request, and estimate the time required to complete the computing power task;

[0035] The matching module is configured to search the computing power resource pool for a computing power node whose computing power is greater than the computing power required by the task and whose task duration does not overlap with the tidal period, and use it as the target computing power node for this task.

[0036] Further,

[0037] The division module is specifically configured as follows:

[0038] The historical resource occupancy of each computing node in the edge computing resource pool is counted one by one, and a first threshold is set. When the average resource occupancy rate of a computing node in a certain period is greater than the first threshold, it is identified as a tidal computing node in the edge computing resource pool, and the corresponding period is identified as the tidal period of the tidal computing node;

[0039] The resource pool module is further configured to:

[0040] After determining whether the computing power node is a tidal computing power node and the corresponding tidal period, the characteristic information of all computing power nodes in the edge computing power resource pool is determined. The characteristic information includes device identification, computing power value, whether it is a tidal computing power node and tidal period.

[0041] In a third aspect, the present disclosure provides an electronic device comprising a memory and a processor, wherein the memory stores a computer program. When the processor runs the computer program stored in the memory, the processor executes the tidal computing power scheduling method as described in any one of the first aspects.

[0042] In a fourth aspect, the present disclosure provides a computer-readable storage medium having a computer program stored thereon, and when the computer program is executed by a processor, the tidal computing power scheduling method described in any one of the first aspects above is implemented.

[0043] Beneficial effects:

[0044] The tidal computing power scheduling method, tidal computing power scheduling system, electronic device and computer-readable storage medium provided by the present disclosure integrate the computing power resources of heterogeneous computing power nodes, classify them according to the operating rules of edge computing power nodes through statistical analysis of historical data, identify the time-varying rules of the available resources of edge computing power nodes, avoid the resource occupancy peak of computing power nodes when allocating computing power, match computing power requirements more reasonably, improve the one-time completion rate of edge computing power tasks, avoid the problem of computing task failure due to fluctuations in the available computing power resources of edge computing power nodes, and then need to perform secondary edge computing power resource allocation, which can greatly improve edge computing efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] Figure 1 A schematic diagram of a flow chart of a tidal computing power scheduling method provided in the first embodiment of the present disclosure;

[0046] Figure 2 A flowchart of a tidal computing power scheduling method provided in the second embodiment of the present disclosure;

[0047] Figure 3 This is an architecture diagram of a tidal computing power scheduling system provided in Example 3 of the present disclosure;

[0048] Figure 4 This is an architectural diagram of an electronic device provided in Example 4 of the present disclosure. DETAILED DESCRIPTION

[0049] To enable those skilled in the art to better understand the technical solutions of the present disclosure, the present disclosure is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments and drawings described herein are only used to explain the present disclosure, rather than to limit the present disclosure.

[0050] It should be noted that the terms "first", "second", etc. in the specification and claims of the present disclosure and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence; and, in the absence of conflict, the embodiments and features in the embodiments of the present disclosure can be arbitrarily combined with each other.

[0051] The terms used in the embodiments of the present disclosure are for the purpose of describing specific embodiments only and are not intended to limit the present disclosure. The singular forms "a," "an," "the," and "the" used in the embodiments of the present disclosure and the appended claims are also intended to include the plural forms unless the context clearly indicates otherwise.

[0052] In the subsequent description, suffixes such as "module," "component," or "unit" used to represent elements are used only to facilitate the description of the present disclosure and have no specific meaning. Therefore, "module," "component," or "unit" may be used interchangeably.

[0053] When it comes to network edge computing resources, the emphasis is on fully utilizing every device at the edge of the network to provide edge computing services with moderate computing power, moderate energy consumption, and low network latency. With the rapid development of the Internet of Things, edge computing has gradually gained popularity. The emergence of edge computing has changed the traditional relative independence of cloud and network, bringing computing into the network. The efficiency and reliability of edge computing will be deeply coupled with network bandwidth, latency, security, and isolation, achieving efficient services that integrate computing and the network.

[0054] For edge computing users, fully utilizing existing edge nodes requires intelligent scheduling. Different users utilize different scheduling strategies based on their specific business scenarios. However, edge computing users currently primarily develop scheduling systems based on their own business characteristics to meet their needs, often using generic, static scheduling strategies provided by third parties. These static scheduling strategies fail to account for the dynamic dynamics between computing power demanders and providers, and in extreme cases, can prevent edge computing tasks from completing properly.

[0055] The following is a detailed description of the technical solutions of the present invention and how the technical solutions of the present invention solve the above-mentioned technical problems in the prior art with specific embodiments. It will be understood that in the embodiments of the present application, the execution subject may perform some or all of the steps in the embodiments of the present application, and these steps or operations are merely examples. The embodiments of the present application may also perform other operations or variations of various operations. In addition, the various steps may be performed in different orders as presented in the embodiments of the present application, and it may not be necessary to perform all the operations in the embodiments of the present application. Furthermore, the following specific embodiments may be combined with each other, and the same or similar concepts or processes may not be described in detail in certain embodiments.

[0056] Figure 1 This is a flow chart of a tidal computing power scheduling method provided in the first embodiment of the present disclosure, as shown in FIG. Figure 1 As shown, the method includes:

[0057] Step S101: Establish an edge computing resource pool, obtain resource information of computing nodes in each computing domain, calculate the computing power value of the computing nodes based on the resource information, and save the resource information and computing power value of each computing node to the edge computing resource pool;

[0058] Step S102: performing time-varying feature division on the computing nodes in the edge computing resource pool to determine the tidal computing nodes in the edge computing resource pool and the tidal periods corresponding to the tidal computing nodes;

[0059] Step S103: Receive the business request information from the computing power demander, calculate the computing power required for the business request, and estimate the time required to complete the computing power task;

[0060] Step S104: Search the computing power resource pool for a computing power node whose computing power value is greater than the computing power required by the task and whose task duration does not overlap with the tidal period, and use it as the target computing power node for this task.

[0061] The disclosed embodiment establishes an edge computing power resource pool, collects historical data statistics on the computing power resource usage of all edge computing power nodes, and divides the edge computing power nodes into two categories: tidal computing power nodes and non-tidal computing power nodes; the computing power matching module analyzes the computing power demand details, including its business type, required computing power size, expected task duration, etc., and then allocates corresponding tidal computing power nodes or non-tidal computing power nodes to them, thereby effectively avoiding the risk of edge computing task abnormalities caused by tidal computing power changes and the resulting problems of unreliable computing tasks, abnormal exit leading to edge computing task reallocation and corresponding delay increase, while fully meeting the task computing power requirements.

[0062] Further,

[0063] The resource information includes device identification, device CPU / GPU main frequency, number of cores, device memory and network bandwidth

[0064] Calculating the computing power value of the computing power node according to the resource information includes:

[0065] The computing power value of the computing power node is calculated according to the CPU / GPU parameters to obtain the quantified computing power value.

[0066] Obtain the hardware resource information of all nodes (computing power providers) in each computing power domain, calculate the quantified computing power value based on the CPU / GPU parameters, and save the device identification, computing power value, device memory, network bandwidth and other information of each computing power node to the edge computing power resource pool;

[0067] In one embodiment, the computing power value is calculated as follows: computing power of the CPU / GPU chip = main frequency * number of cores * floating-point calculation value in a single cycle. For example, if a processor has 20 cores, a main frequency of 2.10 GHz, and a single-cycle floating-point calculation value of 32, then the computing power of the processor is: 20*2.10*32=1344GFLOPS (FLOPS means the number of floating-point calculations per second).

[0068] Further,

[0069] The time-varying feature division of computing nodes in the edge computing resource pool includes:

[0070] The historical resource occupancy of each computing node in the edge computing resource pool is counted one by one, and a first threshold is set. When the average resource occupancy rate of a computing node in a certain period is greater than the first threshold, it is identified as a tidal computing node in the edge computing resource pool, and the corresponding period is identified as the tidal period of the tidal computing node;

[0071] The method further comprises:

[0072] After determining whether the computing power node is a tidal computing power node and the corresponding tidal period, the characteristic information of all computing power nodes in the edge computing power resource pool is determined. The characteristic information includes device identification, computing power value, whether it is a tidal computing power node and tidal period.

[0073] Considering the fluctuations in resource utilization (including CPU, memory, and available bandwidth) of edge computing devices, some edge computing devices may experience very high resource utilization during certain periods of time. These edge computing device nodes are called tidal computing nodes, while those with low utilization are called non-tidal computing nodes. By adding a tidal computing attribute to edge computing nodes, we can provide an effective reference for avoiding tidal periods.

[0074] In order to prevent the computing task from being completed normally due to conflicts between computing power scheduling and the busy hours of tidal computing power nodes, the embodiment of the present disclosure counts the historical occupancy of edge computing power device resources one by one and sets a first threshold. When the average resource occupancy rate of an edge computing power device in a certain period is greater than the first threshold, it is marked as a tidal computing power node in the edge computing power resource pool, and the corresponding period T x [t i , t j ] is marked as a tidal period. An edge computing device may have multiple tidal periods, which are represented by T = {T1, T2, ...T x The first threshold can be set according to actual conditions, such as when the occupancy exceeds 70%, 80%, etc., and there is no restriction here.

[0075] Furthermore, the method further comprises:

[0076] The average resource utilization of computing nodes is calculated using the following formula:

[0077] Resource utilization rate = α*A+β*B+γ*C

[0078] Among them, α, β, and γ are the weights of CPU, memory, and network bandwidth in the resource usage calculation, respectively. A is the CPU usage of the computing power node, B is the memory usage of the computing power node, and C is the network bandwidth usage of the computing power node.

[0079] By setting the weights α, β, and γ of the device CPU, memory, and network bandwidth in the resource usage calculation, where α + β + γ = 1, the average resource usage of the computing power node can be flexibly calculated to make it more suitable for the corresponding scenario. For example, if the computing task requirements are more focused on the CPU, the weight of the CPU in the resource usage calculation can be increased.

[0080] Furthermore, the calculation of the computing power required for the business request and the estimation of the time required to complete the computing task include:

[0081] Convert the computing power required for the business request into FLOPS;

[0082] Categorize computing power based on business request information from computing power demanders into tasks with fixed computing capacity and tasks with unpredictable duration.

[0083] For tasks with fixed computing amount, the task duration is calculated by the computing amount / computing power. For tasks with unpredictable duration, the task duration is set to unpredictable.

[0084] After receiving the business request information from the computing power demander, obtain its device identification, calculate the amount of computing power required, and convert it into FLOPS; classify it according to the business request information of the computing power demander, and divide it into fixed computing tasks (such as AI training, etc.) and tasks with unpredictable duration (such as video surveillance); then estimate the time required to complete the computing power task: for fixed computing tasks, the task duration can be obtained by dividing the computing amount / computing power; for tasks with unpredictable duration, the task duration is also unpredictable. According to the computing power demand value and task duration obtained in the above steps, find a matching node in the edge computing power resource pool. For computing tasks whose task duration cannot be estimated, avoid using tidal computing nodes to avoid the risk of computing tasks not being completed normally.

[0085] Furthermore, the method further comprises:

[0086] By analyzing the tidal period T associated with the tidal computing power node x [t i , t j ] Set the pre-switching time period Δt and pre-switching time t for the tidal computing power node i -Δt;

[0087] After the pre-switching time is reached, the task mirror of the task in the tidal computing power node is sent to the target computing power node for calculation.

[0088] After obtaining the required computing power and task duration, the computing power value and current system resource utilization rate of each computing power node in the computing power resource pool are obtained one by one. A computing power node (either a tidal computing power node or a non-tidal computing power node) with a computing power value greater than the computing power required for the task and a task duration that does not overlap with the tidal period T is found in the computing power resource pool and is used as the target computing power node for this task.

[0089] For tasks in tidal computing nodes, by pre-setting the pre-switching time period Δt (the duration is set according to the volatility of the start time of the tidal period) and the pre-switching time, for nodes that are about to enter the tidal period, by setting the pre-switching time and target node, pre-switching preparations are made to avoid the risk of tidal node computing redundancy.

[0090] Furthermore, the method further comprises:

[0091] The characteristic information of each computing power node in the edge computing power resource pool is updated regularly to match the computing power demand through the updated edge computing power resource pool.

[0092] By updating the edge computing resource pool, the characteristic information of each computing node can be made more accurate, thereby better realizing the dynamic matching of computing resources.

[0093] The disclosed embodiments integrate the computing power resources of heterogeneous computing power nodes, classify them according to the operating rules of edge computing power nodes through statistical analysis of historical data, identify the time-varying rules of the available resources of edge computing power nodes, avoid the resource occupancy peak of computing power nodes when allocating computing power, and more reasonably match computing power requirements, thereby improving the one-time completion rate of edge computing power tasks, avoiding the problem of computing task failure due to fluctuations in the available computing power resources of edge computing power nodes and the need for secondary edge computing power resource allocation, and can better improve edge computing efficiency.

[0094] The second embodiment of the present disclosure also provides a tidal computing power scheduling method, which establishes an edge computing power resource pool, collects historical data statistics on the computing power resource usage of all edge computing power nodes, and divides the edge computing power nodes into two categories: tidal computing power nodes and non-tidal computing power nodes; the computing power matching module analyzes the computing power demand details, including its business type, required computing power size, expected task duration, etc., and then allocates corresponding tidal computing power nodes or non-tidal computing power nodes to them; Figure 2 As shown, the method specifically includes:

[0095] Step S1: Establish an edge computing resource pool;

[0096] Obtain hardware resource information for all nodes (computing power providers) within each computing power domain, including device identification, device CPU / GPU main frequency, core count, device memory, and network bandwidth; calculate the node computing power value based on the CPU / GPU parameters to obtain a quantified computing power value, and then save the device identification, computing power value, device memory, network bandwidth and other information together to the edge computing power resource pool;

[0097] The calculation method for computing power is: CPU / GPU chip computing power = main frequency * number of cores * floating-point calculation value in a single cycle. For example, if a processor has 20 cores, a main frequency of 2.10GHz, and its single-cycle floating-point calculation value is 32, then the computing power of the processor is: 20 * 2.10 * 32 = 1344 GFLOPS (FLOPS means floating-point calculations per second);

[0098] Step S2: dividing the edge computing resource pool into time-varying features;

[0099] Taking into account the fluctuations in resource utilization (including CPU, memory, available bandwidth, etc.) of edge computing devices, some edge computing devices may have very high resource utilization during a certain period or several periods of time. Such edge computing device nodes are called tidal nodes, and vice versa, they are called non-tidal nodes.

[0100] In order to prevent the computing power scheduling from conflicting with the busy time of the tidal node, which may cause the computing power task to fail to complete normally, the embodiment of the present disclosure counts the historical occupancy of the edge computing power device resources one by one, and sets a first threshold. When the average resource occupancy rate of an edge computing power device in a certain period is greater than the first threshold, it is marked as a tidal computing power node in the edge computing power resource pool, and the corresponding period T x [t i , t j ] is marked as a tidal period. An edge computing device may have multiple tidal periods, which are represented by T = {T1, T2, ...T x}express.

[0101] Calculation method for edge computing device resource utilization: Set three weight values, α, β, and γ, which correspond to the weights of the device CPU, memory, and network bandwidth in the resource utilization calculation. Assume that the CPU utilization rate is A, the memory utilization rate is B, and the network bandwidth utilization rate is C. The corresponding device resource utilization rate is: α*A+β*B+γ*C;

[0102] Through the above method, the characteristic information of all computing nodes in the edge computing resource pool is obtained, including device identification, computing power value, whether it is a tidal node, tidal period T{T1, T2, ...T x}.

[0103] Step S3: Computing power demand analysis;

[0104] 1. Receive business request information from the computing power demander, obtain its device ID, calculate the required computing power, and convert it into FLOPS;

[0105] 2. Classify computing power demanders based on their business request information into tasks with fixed computing load (such as AI training) and tasks with unpredictable duration (such as video surveillance).

[0106] 3. Estimate the time required to complete the computing task: For tasks with fixed computing power, the task duration can be calculated by dividing the computing power by the computing power. For tasks with unpredictable duration, the task duration is also unpredictable.

[0107] 4. Based on the computing power requirements and task duration obtained in the previous steps, search for matching nodes in the edge computing resource pool.

[0108] Step S4: computing power matching;

[0109] 1. Obtain the computing power value and current system resource utilization of each computing power node in the computing power resource pool one by one (the calculation method is the same as step S2);

[0110] 2. Find a computing node (either a tidal node or a non-tidal node) from the computing resource pool whose computing power is greater than the computing power required for the task and whose task duration does not overlap with the tidal period T. This node is used as the target computing node for this task.

[0111] 3. By analyzing the tidal period T associated with the tidal node, set the pre-switching time period Δt (the duration is set according to the volatility of the tidal period start time) and the pre-switching time t for the tidal node. i -Δt;

[0112] 4. After the pre-switching time is reached, the task image is sent to the target computing power node for calculation.

[0113] Through the above method, it is possible to effectively avoid the abnormal risks of edge computing tasks caused by tidal computing power changes, as well as the problems of unreliable computing tasks, abnormal exit leading to edge computing task reallocation and corresponding delay increase, while fully meeting the task computing power requirements.

[0114] The third embodiment of the present disclosure also provides a tidal computing power scheduling system, such as Figure 3 As shown, the system includes:

[0115] The resource pool module 11 is configured to establish an edge computing resource pool, obtain resource information of computing nodes in each computing domain, calculate the computing power value of the computing nodes based on the resource information, and save the resource information and computing power value of each computing node to the edge computing resource pool;

[0116] A division module 12 is configured to divide the computing power nodes in the edge computing power resource pool by time-varying characteristics to determine the tidal computing power nodes in the edge computing power resource pool and the tidal time periods corresponding to the tidal computing power nodes;

[0117] The demand analysis module 13 is configured to receive business request information from computing power demanders, calculate the amount of computing power required for the business request, and estimate the time required to complete the computing power task;

[0118] The matching module 14 is configured to search the computing power resource pool for a computing power node whose computing power value is greater than the computing power required by the task and whose task duration does not overlap with the tidal period, and use it as the target computing power node for this task.

[0119] Further,

[0120] The resource information includes device identification, device CPU / GPU main frequency, number of cores, device memory and network bandwidth

[0121] The resource pool module 11 is specifically configured as follows:

[0122] The computing power value of the computing power node is calculated according to the CPU / GPU parameters to obtain the quantified computing power value.

[0123] Further,

[0124] The division module 12 is specifically configured as follows:

[0125] The historical resource occupancy of each computing node in the edge computing resource pool is counted one by one, and a first threshold is set. When the average resource occupancy rate of a computing node in a certain period is greater than the first threshold, it is identified as a tidal computing node in the edge computing resource pool, and the corresponding period is identified as the tidal period of the tidal computing node;

[0126] The resource pool module 11 is further configured to:

[0127] After determining whether the computing power node is a tidal computing power node and the corresponding tidal period, the characteristic information of all computing power nodes in the edge computing power resource pool is determined. The characteristic information includes device identification, computing power value, whether it is a tidal computing power node and tidal period.

[0128] Furthermore, the division module 12 is further configured to:

[0129] The average resource utilization of computing nodes is calculated using the following formula:

[0130] Resource utilization rate = α*A+β*B+γ*C

[0131] Among them, α, β, and γ are the weights of CPU, memory, and network bandwidth in the resource usage calculation, respectively. A is the CPU usage of the computing power node, B is the memory usage of the computing power node, and C is the network bandwidth usage of the computing power node.

[0132] Furthermore, the demand analysis module 13 is specifically configured as follows:

[0133] Convert the computing power required for the business request into FLOPS;

[0134] Categorize computing power based on business request information from computing power demanders into tasks with fixed computing capacity and tasks with unpredictable duration.

[0135] For tasks with fixed computing amount, the task duration is calculated by the computing amount / computing power. For tasks with unpredictable duration, the task duration is set to unpredictable.

[0136] Furthermore, the system further includes a switching module 15;

[0137] The switching module 15 is configured to analyze the tidal period T associated with the tidal computing node. x [t i , t j ] Set the pre-switching time period Δt and pre-switching time t for the tidal computing power node i -Δt; and

[0138] After the pre-switching time is reached, the task mirror of the task in the tidal computing power node is sent to the target computing power node for calculation.

[0139] The tidal computing power scheduling system of the embodiment of the present disclosure is used to implement the tidal computing power scheduling method in the method embodiment 1 and the second embodiment, so the description is relatively simple. For details, please refer to the relevant description in the previous method embodiment 1 and the second embodiment, which will not be repeated here.

[0140] In addition, if Figure 4 As shown, the fourth embodiment of the present disclosure further provides an electronic device, including a memory 100 and a processor 200, wherein the memory 100 stores a computer program. When the processor 200 runs the computer program stored in the memory 100, the processor 200 executes the above-mentioned various possible methods.

[0141] The memory 100 is connected to the processor 200 . The memory 100 may be a flash memory, a read-only memory, or other memory. The processor 200 may be a central processing unit or a single-chip microcomputer.

[0142] In addition, an embodiment of the present disclosure further provides a computer-readable storage medium, on which a computer program is stored, and the computer program is used by a processor to execute the above-mentioned various possible methods.

[0143] The computer-readable storage medium includes volatile or nonvolatile, removable or non-removable media implemented in any method or technology for storing information (such as computer-readable instructions, data structures, computer program modules or other data). Computer-readable storage media include, but are not limited to, RAM (Random Access Memory), ROM (Read-Only Memory), EEPROM (Electrically Erasable Programmable read only memory), flash memory or other memory technology, CD-ROM (Compact Disc Read-Only Memory), Digital Versatile Disk (DVD) or other optical disk storage, magnetic cassettes, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other medium that can be used to store the desired information and can be accessed by a computer.

[0144] It is understood that the above embodiments are merely exemplary embodiments for illustrating the principles of the present disclosure, and the present disclosure is not limited thereto. Those skilled in the art may make various modifications and improvements without departing from the spirit and substance of the present disclosure, and such modifications and improvements are also considered to be within the scope of protection of the present disclosure.

Claims

1. A tidal computing power scheduling method, characterized in that: The method comprises: Establish an edge computing resource pool, obtain the resource information of computing nodes in each computing domain, calculate the computing power value of the computing node according to the resource information, and save the resource information and computing power value of each computing node to the edge computing resource pool; The computing nodes in the edge computing resource pool are divided by time-varying characteristics to determine the tidal computing nodes in the edge computing resource pool and the tidal time periods corresponding to the tidal computing nodes; Receive business request information from computing power demanders, calculate the computing power required for the business request, and estimate the time required to complete the computing power task; Find a computing node from the computing resource pool whose computing power value is greater than the computing power required by the task and whose task duration does not overlap with the tidal period, and use it as the target computing power node for this task.

2. The method according to claim 1, characterized in that: The resource information includes device identification, main frequency of the device's central processing unit (CPU) / graphics processing unit (GPU), number of cores, device memory, and network bandwidth. The calculating the computing power value of the computing power node according to the resource information includes: The computing power value of the computing power node is calculated according to the CPU / GPU parameters to obtain the quantified computing power value.

3. The method according to claim 1, characterized in that: The time-varying feature division of computing nodes in the edge computing resource pool includes: The historical resource occupancy of each computing power node in the edge computing power resource pool is counted one by one, and a first threshold is set. When the average resource occupancy rate of a computing power node in a certain period is greater than the first threshold, it is identified as a tidal computing power node in the edge computing power resource pool, and the corresponding period is identified as the tidal period of the tidal computing power node; The method further comprises: After determining whether the computing power node is a tidal computing power node and the corresponding tidal period, the characteristic information of all computing power nodes in the edge computing power resource pool is determined, and the characteristic information includes device identification, computing power value, whether it is a tidal computing power node and tidal period.

4. The method according to claim 3, characterized in that The method further comprises: The average resource occupancy rate of computing power nodes is calculated by the following formula: Resource utilization rate = α*A+β*B+γ*C Among them, α, β, and γ are the weights of CPU, memory, and network bandwidth in the resource occupancy calculation, A is the CPU occupancy rate of the computing power node, B is the memory occupancy rate of the computing power node, and C is the network bandwidth occupancy rate of the computing power node.

5. The method according to claim 1, characterized in that The calculation of the computing power required for the business request and the estimation of the time required to complete the computing task include: Convert the computing power required for the business request into floating point operations per second (FLOPS); Classify computing power demanders according to their business request information into fixed computing tasks and unpredictable tasks. For tasks with fixed computing amount, the task duration is calculated by the computing amount / computing power. For tasks with unpredictable duration, the task duration is set to unpredictable.

6. The method according to claim 1 or 3, characterized in that: The method further comprises: By analyzing the tidal period T associated with the tidal computing power node x [t i , t j ] Set the pre-switching time period Δt and pre-switching time t for the tidal computing power node i -Δt; After the pre-switching time is reached, the task image of the task in the tidal computing power node is sent to the target computing power node for calculation.

7. A tidal computing power scheduling system, characterized in that: The system comprises: A resource pool module is configured to establish an edge computing resource pool, obtain resource information of computing nodes in each computing domain, calculate the computing power value of the computing node according to the resource information, and save the resource information and computing power value of each computing node to the edge computing resource pool; A partitioning module is configured to perform time-varying feature partitioning on the computing nodes in the edge computing resource pool to determine the tidal computing nodes in the edge computing resource pool and the tidal time periods corresponding to the tidal computing nodes; A demand analysis module is configured to receive business request information from computing power demanders, calculate the amount of computing power required for the business request, and estimate the time required to complete the computing power task; The matching module is configured to search the computing power resource pool for computing power nodes whose computing power value is greater than the computing power required by the task and whose task duration does not overlap with the tidal period, and use them as the target computing power nodes for this task.

8. The system according to claim 7, characterized in that The division module is specifically configured as follows: The historical resource occupancy of each computing power node in the edge computing power resource pool is counted one by one, and a first threshold is set. When the average resource occupancy rate of a computing power node in a certain period is greater than the first threshold, it is identified as a tidal computing power node in the edge computing power resource pool, and the corresponding period is identified as the tidal period of the tidal computing power node; The resource pool module is also configured to: After determining whether the computing power node is a tidal computing power node and the corresponding tidal period, the characteristic information of all computing power nodes in the edge computing power resource pool is determined, and the characteristic information includes device identification, computing power value, whether it is a tidal computing power node and tidal period.

9. An electronic device, characterized in that: It includes a memory and a processor, wherein the memory stores a computer program, and when the processor runs the computer program stored in the memory, the processor executes the tidal computing power scheduling method as described in any one of claims 1 to 6.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the tidal computing power scheduling method according to any one of claims 1 to 6 is implemented.

Images