Computing power cluster resource consumption evaluation method for computing operation

Abstract

The invention relates to the field of cluster computing, in particular to a computing power cluster resource consumption evaluation method for computing operation, and aims to solve the problem that resource allocation is not matched with actual requirements due to the fact that comprehensive indexes such as dynamically changing computing load cannot be adapted and network bandwidth and storage I / O (input / output) cannot be neglected when a computing power cluster carries out the computing operation. According to the method, the idle computing power resources are allocated according to the job task scale and the idle computing power resources and the timeliness requirement of the job, and the data transmission and throughput time consumption limitation caused by interference resource parameters is fully considered through pre-allocation, one-time allocation and final allocation for multiple times of resource allocation adjustment in the allocation process; according to the method, limitation of storage resources and network resources can be considered during allocation of the computing power resources, the accuracy degree of allocation of the computing power resources is improved, meanwhile, the overall energy consumption of the computing power cluster is evaluated through dynamic comparison between the energy consumption of the computing power resources and the completion degree, and abnormal behaviors of the computing power energy consumption are found in time.

Description

Technical Field

[0001] The present invention relates to the field of cluster computing, and specifically to a method for evaluating the consumption of computing power cluster resources for computing jobs. Background Art

[0002] Computing power refers to the data processing ability, which integrates information computing power, network carrying capacity, and data storage capacity, and mainly provides services to society through computing power infrastructure such as computing power centers. It exists in various intelligent hardware devices such as mobile phones, laptops, and supercomputers. Its original meaning is to represent the computing performance of a certain device or system. With the advent of the intelligent era, the three elements of intelligent computing - computing power, algorithms, and data - have gradually become an important part of the social information infrastructure, and the connotation of "computing power" has been further expanded to the computing performance that users can obtain and is reflected in the actual utility of users;

[0003] Computing power is the new productive force in the digital economy era, and has become the core force driving the development of the digital economy and the solid foundation supporting the development of the digital economy. It plays an important role in promoting scientific and technological progress, promoting the digital transformation of industries, and supporting economic and social development; it is like water conservancy in the agricultural era and electricity in the industrial era, and has become the core productive force in the development of the digital economy;

[0004] With the explosive growth of computing power demand, computing power clusters such as GPU clusters and CPU clusters have become the core infrastructure to support large-scale computing jobs. However, the existing resource consumption evaluation methods have the following problems:

[0005] Traditional methods rely on static resource allocation strategies and cannot adapt to dynamically changing computing loads, resulting in low resource utilization or job performance bottlenecks. At the same time, most methods only focus on CPU / GPU utilization or memory occupancy, ignoring comprehensive indicators such as network bandwidth, storage I / O, and energy consumption. The evaluation results are one-sided, leading to a mismatch between resource allocation and actual needs, making it difficult to balance energy consumption and computing efficiency and unable to meet the requirements of green computing;

[0006] In response to the above technical problems, this application proposes a solution. Summary of the Invention

[0007] In the present invention, based on the scale of the operation task and the idle computing power resources, the idle computing power resources are allocated according to the timeliness requirements of the operation, and during the allocation process, multiple resource allocation adjustments such as pre-allocation, primary allocation, and final allocation are carried out. The data transmission and throughput time-consuming limitations brought by the interference resource parameters are fully considered, so that the storage resources and network resource limitations can be taken into account when allocating computing power resources, improving the accuracy of the computing power resource allocation, and solving the problem that when performing computing operations through a computing power cluster, it is impossible to adapt to dynamic changes in the computing load, ignoring comprehensive indicators such as network bandwidth, storage I / O, and energy consumption, resulting in a mismatch between resource allocation and actual demand, and it is difficult to balance energy consumption and computing efficiency, resulting in low resource utilization. Therefore, a method for evaluating the resource consumption of a computing power cluster for computing operations is proposed.

[0008] The object of the present invention can be achieved by the following technical solutions:

[0009] A method for evaluating the resource consumption of a computing power cluster for computing operations, comprising the following steps:

[0010] Step 1: Obtain the scale of the computing power cluster to obtain the total idle computing power resources;

[0011] Step 2: Obtain the computing operation tasks to obtain the scale of the operation tasks. If there are multiple groups of computing operation tasks, each computing operation task is numbered, the scale of the operation tasks is calculated respectively according to the numbering results, and the operation timeliness is generated according to the completion time requirements of the computing operations;

[0012] Step 3: Compare the total computing power resources with the scale of the operation tasks, perform a primary allocation of computing power resources according to the operation timeliness, and generate an estimated computing time through the primary allocation of computing power resources;

[0013] Step 4: Obtain the scale of the operation tasks and the interference resources, obtain the interference limit time according to the interference resources and the scale of the task operation, compare the estimated computing time with the interference limit time, retain or correct the primary allocation of computing power resources according to the comparison result, and record it as the final computing power resources;

[0014] Step 5: Calculate the energy consumption of the allocated final computing power resources, and at the same time obtain the progress of the completion of the operation tasks;

[0015] Step 6: Compare the energy consumption of the computing power resources with the progress of the completion of the operation to obtain the energy consumption efficiency, compare the energy consumption evaluation with the set standard, and generate an energy consumption evaluation result.

[0016] As a preferred embodiment of the present invention, the operation timeliness generated in Step 2 is the latest time point for completing the computing operation;

[0017] In step 3, when allocating the total computing power resources, first reserve a part of the spare computing power resources from the total computing power resources, record the remaining total computing power resources as available computing power resources, then count all the job task scales to obtain the total job task scale, and evenly distribute the available computing power resources to the total job task scale to complete the pre-allocation of computing power resources;

[0018] Calculate according to the pre-allocation result of computing power resources to obtain the estimated completion time of the job task, compare the estimated completion time with the job timeliness, judge whether the estimated completion time meets the job timeliness, classify the computing job tasks into tasks with insufficient computing power and tasks with redundant computing power, and make mutual adjustments according to the tasks with insufficient computing power and tasks with redundant computing power. Finally, involve the spare computing power resources in the allocation to complete one allocation of computing power resources.

[0019] As a preferred embodiment of the present invention, the interference resources in step 4 include storage resources and network resources. Obtain the data throughput time through the resource throughput speed of the storage resources and the computing task scale, obtain the data transmission time through the resource transmission speed of the network resources and the computing task scale, and record the minimum value of the data throughput time and the data transmission time as the interference limit time-consuming;

[0020] Then compare the interference limit time-consuming with the estimated computing time-consuming. If the estimated computing time-consuming is less than the interference limit time-consuming, correct the primary computing power resources. If the estimated computing time-consuming is greater than or equal to the interference limit time-consuming, retain the primary computing power resources.

[0021] As a preferred embodiment of the present invention, it further includes a computing power cluster resource consumption evaluation system for computing jobs, including a computing power cluster supervision unit, a computing job task interface, a computing power resource allocation unit, an interference resource intervention unit, a computing power resource correction unit, and an energy consumption efficiency evaluation unit;

[0022] The computing power cluster supervision unit is used to connect to the computing power cluster and obtain the total computing power resources of the computing power cluster;

[0023] The computing job task interface is used to receive job tasks and the computing job completion time requirement, analyze the scale of the computing job tasks, and generate job timeliness according to the computing job completion time requirement;

[0024] The computing power resource allocation unit obtains the total computing power resources through the computing power cluster supervision unit, obtains the computing job task scale through the computing job task interface, and comprehensively analyzes the total computing power resources, the computing job task scale, and the job timeliness to generate the estimated computing time-consuming and the primary computing power resource allocation result;

[0025] The interference resource intervention unit can obtain interference resources, including network resources and storage resources, and analyze them according to the interference resources and the scale of the operation tasks to obtain the interference limit time consumption;

[0026] The computing power resource correction unit can compare the interference limit time consumption with the expected computing time consumption, and retain or correct the primary computing power resource allocation according to the comparison result to obtain the final computing power resources;

[0027] The energy consumption efficiency evaluation unit analyzes the operation completion progress and the energy consumption of the computing power resources to obtain the energy consumption efficiency, and generates an energy consumption evaluation result according to the energy consumption efficiency.

[0028] As a preferred embodiment of the present invention, after the computing power resource allocation unit evenly distributes the available computing power resources to the total scale of the operation tasks, it integrates the evenly distributed available computing power resources according to different computing task scales to obtain the pre-allocation of the computing power resources allocated to each computing operation task;

[0029] The computing power resource allocation unit records the tasks with insufficient computing power as tasks with insufficient computing power when the expected computing time consumption does not meet the operation timeliness, records the tasks with redundant computing power as tasks with redundant computing power when the expected computing time consumption meets the operation timeliness, and centralizes the redundant computing power of the tasks with redundant computing power and allocates it to the tasks with insufficient computing power. If there are still tasks with insufficient computing power after the allocation, they are supplemented with standby computing power resources.

[0030] As a preferred embodiment of the present invention, if the computing power resource correction unit obtains that the expected computing time consumption is less than the interference limit time consumption, it reduces the computing power allocated to the computing tasks so that the expected computing time consumption is equal to the interference limit time consumption, and centralizes the reduced computing power into the standby computing power resources.

[0031] As a preferred embodiment of the present invention, the energy consumption efficiency evaluation unit obtains the operation task completion progress, and at the same time obtains the time duration of the computing operation task. It calculates according to the final computing power resources allocated to the computing operation task and the time duration of the computing operation task to obtain the energy consumption of the computing power resources. The energy consumption efficiency evaluation unit calculates the energy consumption of the computing power resources and the operation task completion progress to obtain the total energy consumption when all computing operation tasks are completed, and calculates the ratio of the total energy consumption to the scale of the computing operation task to obtain the unit energy consumption of the computing operation task;

[0032] The energy consumption efficiency evaluation unit makes a standard judgment on the unit energy consumption of the computing operation task to obtain high-energy consumption tasks and qualified-energy consumption tasks.

[0033] As a preferred embodiment of the present invention, the energy consumption efficiency evaluation unit sums up the unit energy consumption of each computing job task to obtain the overall unit energy consumption, and judges the overall unit energy consumption and a set threshold, and generates an energy consumption exceeding standard signal or an energy consumption qualified signal according to the judgment result

[0034] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0035] 1. In the present invention, the scale of the job task and the idle computing power resources are used, and the idle computing power resources are allocated according to the timeliness requirements of the job. During the allocation process, multiple resource allocation adjustments such as pre-allocation, primary allocation, and final allocation are carried out, so that the accuracy of the allocation of computing power resources is improved, and the idle waste caused by redundant resource allocation is avoided.

[0036] 2. In the present invention, by considering the data transmission and throughput time-consuming limitations brought by the interference resource parameters, the storage resources and network resources limitations can be considered when allocating computing power resources, and the allocation of computing power resources is checked and corrected again based on the limitations of storage resources and network resources, so as to avoid excessive allocation of computing power resources, resulting in a large amount of redundant waste of computing power resources due to the short board of storage resources and network resources, and improve the comprehensiveness of consideration when allocating computing power resources.

[0037] 3. In the present invention, the real-time energy consumption of the allocation result of the computing power resources and the job completion degree is calculated. Through the dynamic comparison between the energy consumption and the completion degree, the energy consumption required to complete the computing task is evaluated, and the computing task is classified according to the energy consumption result. At the same time, the overall energy consumption of the computing power cluster is evaluated, so that abnormal energy consumption behaviors of the computing power can be found in time, providing data support for the optimization of subsequent computing power resource scheduling strategies. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] For the convenience of those skilled in the art to understand, the present invention will be further described below with reference to the accompanying drawings.

[0039] Figure 1 is the system block diagram of the present invention;

[0040] Figure 2 is the system flow chart of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0041] The technical solutions of the present invention will be clearly and completely described below in conjunction with the embodiments. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative work shall fall within the protection scope of the present invention.

[0042] Embodiment 1:

[0043] Please refer to Figure 1 - Figure 2 As shown, a method for evaluating the computing power cluster resource consumption of computing jobs includes the following steps:

[0044] Step 1: Obtain the scale of the computing power cluster to get the total computing power resources in the idle state;

[0045] Step 2: Obtain the computing job tasks to get the scale of the job tasks. If there are multiple groups of computing job tasks, number each computing job task, calculate the scale of the job tasks according to the numbering results respectively, and generate job timeliness according to the completion time requirements of the computing jobs. The generated job timeliness is the latest time point to complete the computing jobs;

[0046] Step 3: Compare the total computing power resources with the scale of the job tasks, perform a first allocation of computing power resources according to the job timeliness, and generate an estimated computing time through the first allocation of computing power resources. When allocating the total computing power resources, first reserve a part of the backup computing power resources from the total computing power resources, record the remaining total computing power resources as available computing power resources, then count all the scales of the job tasks to get the total scale of the job tasks, evenly distribute the available computing power resources to the total scale of the job tasks, and then integrate the computing power resources allocated to a computing job task to complete the pre-allocation of computing power resources;

[0047] Calculate according to the pre-allocation result of the computing power resources and the scale of the computing job tasks to get the estimated completion time of the job tasks, and compare the estimated completion time with the job timeliness to determine whether the estimated completion time meets the job timeliness. If the estimated completion time meets the job timeliness, allocate the computing job tasks as computing power redundant tasks. If the estimated completion time does not meet the job timeliness, allocate the computing job tasks as computing power insufficient tasks, and adjust each other according to the computing power insufficient tasks and the computing power redundant tasks, so as to allocate part of the computing power of the computing power redundant tasks to the computing power insufficient tasks, thus avoiding the existence of computing power insufficient tasks. If there are still computing power insufficient tasks after the allocation of the computing power of the computing power redundant tasks, then allocate the backup computing power resources to complete the first allocation of computing power resources. After the first allocation of computing power resources is completed, predict through the allocated computing power resources and the scale of the computing tasks to get the estimated computing time;

[0048] Step 4: Obtain the scale of the job task and interference resources. The interference resources include storage resources and network resources. Calculate the data throughput time based on the resource throughput rate of the storage resources and the scale of the computing task, and calculate the data transmission time based on the resource transmission rate of the network resources and the scale of the computing task. Record the minimum value of the data throughput time and the data transmission time as the interference limit time. Compare the estimated computing time with the interference limit time. If the estimated computing time is less than the interference limit time, correct the primary computing power resources. If the estimated computing time is greater than or equal to the interference limit time, retain the primary computing power resources. When correcting the computing power resources, reduce the computing power allocated to the computing task to make the estimated computing time equal to the interference limit time, and concentrate the reduced computing power into the standby computing power resources. Finally, record the computing power allocation situation after correction or retention as the final computing power resources;

[0049] Step 5: Calculate the energy consumption of the allocated final computing power resources, obtain the duration of the computing job task, and calculate the computing power resource energy consumption based on the final computing power resources allocated for the computing job task and the duration of the computing job task. At the same time, obtain the progress of the job task completion;

[0050] Step 6: Calculate the computing power resource energy consumption and the job task completion progress to obtain the total energy consumption when all computing job tasks are completed. Sum up the total energy consumption of each computing task, and finally calculate the ratio with the total sum of the computing job task scale to obtain the energy consumption efficiency. Compare the energy consumption assessment with the set standard to generate an energy consumption assessment result, including generating an energy consumption exceeding standard signal or an energy consumption qualified signal.

[0051] Embodiment 2:

[0052] Please refer to Figure 1 - Figure 2 As shown in the figure, a computing power cluster resource consumption assessment system for computing jobs includes a computing power cluster supervision unit, a computing job task interface, a computing power resource allocation unit, an interference resource intervention unit, a computing power resource correction unit, and an energy consumption efficiency assessment unit;

[0053] The computing power cluster supervision unit is used to connect to the computing power cluster and obtain the available total computing power resources of the computing power cluster;

[0054] The computing job task interface is used to receive job tasks and the computing job completion time requirements, analyze the scale of the computing job tasks, and generate job timeliness according to the computing job completion time requirements;

[0055] The computing power resource allocation unit obtains the total computing power resources through the computing power cluster supervision unit, obtains the scale of the computing job tasks through the computing job task interface, conducts a comprehensive analysis based on the total computing power resources, the scale of the computing job tasks, and the job timeliness, and evenly distributes the available computing power resources into the total scale of the job tasks to complete the pre-allocation of computing power resources;

[0056] After the computing power resource allocation unit evenly distributes the available computing power resources into the total scale of the job tasks, it integrates the evenly distributed available computing power resources according to different computing task scales to obtain the pre-allocation of the computing power resources allocated to each computing job task;

[0057] The computing power resource allocation unit records the tasks with insufficient computing power for which the estimated computing time does not meet the job timeliness as tasks with insufficient computing power, records the tasks with redundant computing power for which the estimated computing time meets the job timeliness as tasks with redundant computing power, centralizes the redundant computing power of the tasks with redundant computing power, and allocates it to the tasks with insufficient computing power. If there are still tasks with insufficient computing power after the allocation, they are supplemented with backup computing power resources to obtain the first result of computing power resource allocation, and the estimated computing time is generated according to the first result of computing power resource allocation;

[0058] The interference resource intervention unit can obtain interference resources, including network resources and storage resources, and conduct an analysis based on the interference resources and the scale of the job tasks to obtain the interference limit time;

[0059] The computing power resource correction unit can compare the interference limit time with the estimated computing time, and retain or correct the first computing power resource allocation according to the comparison result to obtain the final computing power resources. If the computing power resource correction unit obtains that the estimated computing time is less than the interference limit time, it reduces the computing power allocated to the computing tasks so that the estimated computing time is equal to the interference limit time, and centralizes the reduced computing power into the backup computing power resources;

[0060] The energy consumption efficiency evaluation unit analyzes the job completion progress and the energy consumption of the computing power resources to obtain the energy consumption efficiency, and generates an energy consumption evaluation result according to the energy consumption efficiency;

[0061] The specific evaluation method is as follows:

[0062] The energy consumption efficiency evaluation unit obtains the job task completion progress, and at the same time obtains the duration of the computing job tasks. It calculates according to the final computing power resources allocated to the computing job tasks and the duration of the computing job tasks to obtain the energy consumption of the computing power resources. The energy consumption efficiency evaluation unit calculates the energy consumption of the computing power resources and the job task completion progress to obtain the total energy consumption when all computing job tasks are completed, and conducts a ratio calculation through the total energy consumption and the scale of the computing job tasks to obtain the unit energy consumption of the computing job tasks;

[0063] The energy consumption efficiency evaluation unit makes a standard judgment on the unit energy consumption of the computing job tasks, obtains high-energy-consuming tasks and qualified energy-consuming tasks, and sends the high-energy-consuming tasks and qualified energy-consuming tasks through the network to achieve the visual output of the energy consumption of different computing tasks;

[0064] The energy consumption efficiency evaluation unit calculates the sum according to the unit energy consumption of each computing job task to obtain the overall unit energy consumption, and judges the overall unit energy consumption and the set threshold. If the overall unit energy consumption is greater than the set threshold, an energy consumption exceeding standard signal is generated. If the overall unit energy consumption is not greater than the set threshold, an energy consumption qualified signal is generated.

[0065] The preferred embodiments of the present invention disclosed above are only used to help illustrate the present invention. The preferred embodiments do not describe all the details in detail, nor do they limit the invention to the specific implementation manners. Obviously, many modifications and variations can be made according to the content of this specification. These embodiments are selected and specifically described in this specification to better explain the principle and practical application of the present invention, so that those skilled in the art can understand and utilize the present invention well. The present invention is only limited by the claims and their full scope and equivalents.

Claims

1. A method for evaluating the computing power cluster resource consumption of a computing job, characterized in that, It includes the following steps: Step 1: Obtain the scale of the computing power cluster to get the total computing power resources in the idle state; Step 2: Obtain the computing job tasks to get the scale of the job tasks. If there are multiple groups of computing job tasks, number each computing job task, calculate the scale of the job tasks respectively according to the numbering results, and generate job timeliness according to the computing job completion time requirements; Step 3: Compare the total computing power resources with the scale of the job tasks, perform the first allocation of computing power resources according to the job timeliness, and generate the estimated computing time through the first allocation of computing power resources; Step 4: Obtain the scale of the job tasks and the interference resources, get the interference limit time according to the interference resources and the scale of the task jobs, compare the estimated computing time with the interference limit time, retain or correct the first allocation of computing power resources according to the comparison result, and record it as the final computing power resources; Step 5: Calculate the energy consumption of the allocated final computing power resources, and at the same time obtain the progress of the completion of the job tasks; Step 6: Compare the energy consumption of the computing power resources with the progress of the job completion to obtain the energy consumption efficiency, compare the energy consumption assessment with the set standard, and generate the energy consumption assessment result.

2. The computing power cluster resource consumption evaluation method for computing jobs according to claim 1, wherein The job timeliness generated in Step 2 is the latest time point for completing the computing job; When allocating the total computing power resources in Step 3, first reserve a part of the standby computing power resources from the total computing power resources, record the remaining total computing power resources as available computing power resources, then count all the scales of the job tasks to get the total scale of the job tasks, and evenly distribute the available computing power resources to the total scale of the job tasks to complete the pre-allocation of computing power resources; Calculate according to the result of the pre-allocation of computing power resources to get the estimated completion time of the job tasks, compare the estimated completion time with the job timeliness, judge whether the estimated completion time meets the job timeliness, classify the computing job tasks into tasks with insufficient computing power and tasks with redundant computing power, and adjust them mutually according to the tasks with insufficient computing power and tasks with redundant computing power. Finally, involve the standby computing power resources in the allocation to complete the first allocation of computing power resources.

3. A method for evaluating the computing power cluster resource consumption of a computing job according to claim 1, characterized in that, The interference resources in Step 4 include storage resources and network resources. Obtain the data throughput time through the resource throughput speed of the storage resources and the scale of the computing tasks, obtain the data transmission time through the resource transmission speed of the network resources and the scale of the computing tasks, and record the minimum value of the data throughput time and the data transmission time as the interference limit time; Then compare the interference limit time with the estimated computing time. If the estimated computing time is less than the interference limit time, correct the first allocation of computing power resources. If the estimated computing time is greater than or equal to the interference limit time, retain the first allocation of computing power resources.

4. A method for evaluating the computing power cluster resource consumption of a computing job according to claim 1, characterized in that, It also includes a computing power cluster resource consumption assessment system for computing jobs, including a computing power cluster supervision unit, a computing job task interface, a computing power resource allocation unit, an interference resource intervention unit, a computing power resource correction unit, and an energy consumption efficiency assessment unit; The computing power cluster supervision unit is used to connect to the computing power cluster and obtain the total computing power resources of the computing power cluster; The computing job task interface is used to receive job tasks and computing job completion time requirements, parse the scale of the computing job tasks, and generate job timeliness according to the computing job completion time requirements; The computing power resource allocation unit obtains the total computing power resources through the computing power cluster supervision unit, obtains the scale of the computing job tasks through the computing job task interface, and generates the estimated computing time and the first computing power resource allocation result through comprehensive analysis of the total computing power resources, the scale of the computing job tasks, and the job timeliness; The interference resource intervention unit can obtain interference resources, including network resources and storage resources, and analyze them according to the interference resources and the scale of the job tasks to obtain the interference limit time; The computing power resource correction unit can compare the interference limit time with the estimated computing time, and retain or correct the first computing power resource allocation according to the comparison result to obtain the final computing power resources; The energy consumption efficiency evaluation unit analyzes the job completion progress and the energy consumption of the computing power resources to obtain the energy consumption efficiency, and generates an energy consumption evaluation result according to the energy consumption efficiency; 5. A method for evaluating the computing power cluster resource consumption of a computing job according to claim 4, wherein, After the computing power resource allocation unit evenly distributes the available computing power resources to the total scale of the job tasks, it integrates the evenly distributed available computing power resources according to different computing task scales to obtain the pre-allocation of the computing power resources allocated to each computing job task; The computing power resource allocation unit records the tasks with insufficient computing power for which the estimated computing time does not meet the job timeliness, records the tasks with redundant computing power for which the estimated computing time meets the job timeliness, centralizes the redundant computing power of the tasks with redundant computing power, and allocates it to the tasks with insufficient computing power. If there are still tasks with insufficient computing power after the allocation, they are supplemented with standby computing power resources; 6. A method for evaluating the computing power cluster resource consumption of a computing job according to claim 4, characterized in that If the computing power resource correction unit obtains that the estimated computing time is less than the interference limit time, it reduces the computing power allocated to the computing tasks so that the estimated computing time is equal to the interference limit time, and centralizes the reduced computing power into the standby computing power resources; 7. A method for evaluating the computing power cluster resource consumption of a computing job according to claim 4, characterized in that The energy consumption efficiency evaluation unit obtains the job task completion progress, and at the same time obtains the time duration of the computing job tasks. It calculates according to the final computing power resources allocated to the computing job tasks and the time duration of the computing job tasks to obtain the energy consumption of the computing power resources. The energy consumption efficiency evaluation unit calculates the energy consumption of the computing power resources and the job task completion progress to obtain the total energy consumption when all computing job tasks are completed, and calculates the ratio of the total energy consumption to the scale of the computing job tasks to obtain the unit energy consumption of the computing job tasks; The energy consumption efficiency evaluation unit makes a standard judgment on the unit energy consumption of the computing job tasks to obtain high-energy consumption tasks and qualified-energy consumption tasks; 8. A method for evaluating the computing power cluster resource consumption of a computing job according to claim 7, characterized in that, The energy consumption efficiency evaluation unit calculates the sum of the unit energy consumption of each computing job task to obtain the overall unit energy consumption, and judges the overall unit energy consumption and the set threshold. According to the judgment result, it generates an energy consumption exceeding standard signal or an energy consumption qualified signal.

Images