Performance evaluation method and device, computer equipment, readable storage medium and program product

Abstract

The invention relates to a performance evaluation method and device, computer equipment, a computer readable storage medium and a computer program product. The method comprises the following steps: acquiring data of calculation efficiency indexes of each to-be-evaluated data center; determining subjective and objective comprehensive weights of the calculation effect indexes based on the data of the calculation effect indexes; obtaining a comprehensive evaluation result of each to-be-evaluated data center through a comprehensive evaluation method according to the data of the calculation effect indexes and the subjective and objective comprehensive weights; based on the comprehensive evaluation result, determining a calculation effect short board index; the calculation effect short board index represents the calculation effect index of which the index performance does not meet the energy consumption requirement. By adopting the method, the calculation efficiency index of the data center of which the performance cannot meet the energy consumption requirement can be determined.

Description

Technical Field

[0001] The present application relates to the field of information technology, and in particular to a performance evaluation method, apparatus, computer equipment, computer-readable storage medium, and computer program product. Background Art

[0002] With the development of information technologies such as artificial intelligence, big data, and the Internet of Things, the demand for computing power has increased significantly, and computing power has become a new type of productivity supporting the development of the digital economy. As a key infrastructure, the computing power and energy consumption of data centers cannot be ignored. At present, the energy consumption status of data centers can be evaluated by the computing efficiency of data centers.

[0003] However, traditional computing efficiency evaluation methods only evaluate computing efficiency through the computing power of the data center and the power consumption of IT (Internet Technology) equipment. This evaluation method cannot fully reflect the real system characteristics of the data center, resulting in the inability of traditional computing efficiency evaluation methods to determine the computing efficiency indicators of data centers whose performance cannot meet the energy consumption requirements. In other words, traditional computing efficiency evaluation methods cannot evaluate the performance shortcomings of data centers. Summary of the invention

[0004] Based on this, it is necessary to provide a performance evaluation method, device, computer equipment, computer-readable storage medium and computer program product that can determine the computing efficiency indicators of data centers whose performance cannot meet energy consumption requirements in response to the above technical problems.

[0005] In a first aspect, in one embodiment, the present application provides a performance evaluation method, the method comprising:

[0006] Obtain the data of computing efficiency indicators of each data center to be evaluated;

[0007] Based on the data of the calculation efficiency indicators, determine the subjective and objective comprehensive weights of the calculation efficiency indicators;

[0008] Based on the data of the computing efficiency indicators and the subjective and objective comprehensive weights, the comprehensive evaluation results of each data center to be evaluated are obtained through a comprehensive evaluation method;

[0009] Based on the comprehensive evaluation results, the computing efficiency shortcoming indicators are determined; the computing efficiency shortcoming indicators refer to the computing efficiency indicators whose indicator performance does not meet the energy consumption requirements.

[0010] In one embodiment, obtaining data on computing efficiency indicators of each data center to be evaluated includes:

[0011] Based on multiple evaluation dimensions of the data center to be evaluated, the data of the computing efficiency indicators of each data center to be evaluated is obtained; the evaluation dimension represents at least one type of evaluation influencing factor of the computing efficiency of the data center.

[0012] In one embodiment, the evaluation dimensions include a first type of evaluation influencing factors, a second type of evaluation influencing factors, a third type of evaluation influencing factors, and a fourth type of evaluation influencing factors;

[0013] Among them, the first type of evaluation influencing factors corresponds to the computing power of the data center, the second type of evaluation influencing factors corresponds to the storage capacity of the data center, the third type of evaluation influencing factors corresponds to the carrying capacity of the data center, and the fourth type of evaluation influencing factors corresponds to the low-carbon power of the data center.

[0014] In one embodiment, based on the data of the calculation efficiency indicator, the subjective and objective comprehensive weights of the calculation efficiency indicator are determined, including:

[0015] Based on the subjective weighting method, the subjective weight corresponding to each calculation efficiency indicator is obtained;

[0016] Based on the objective weighting method, the objective weight corresponding to each calculation efficiency indicator is obtained;

[0017] The subjective weight and objective weight are weighted and integrated to obtain the subjective and objective comprehensive weights corresponding to each calculation efficiency indicator.

[0018] In one embodiment, based on the subjective weighting method, the subjective weight corresponding to each calculation efficiency indicator is obtained, including:

[0019] Compare the calculation efficiency indicators to obtain the indicator comparison results;

[0020] Construct a judgment matrix based on the index comparison results and obtain the maximum eigenvalue of the judgment matrix;

[0021] Based on the maximum eigenvalue of the judgment matrix, when it is determined that the judgment matrix meets the consistency requirement, the matrix eigenvalue vector of the judgment matrix is ​​solved;

[0022] The matrix eigenvector of the judgment matrix is ​​normalized to obtain the subjective weight corresponding to each calculation efficiency index.

[0023] In one embodiment, based on the objective weighting method, the objective weight corresponding to each calculation efficiency indicator is obtained, including:

[0024] Based on the data of the calculation efficiency index, construct the original data matrix;

[0025] The original data matrix is ​​dimensionless through forward transformation and normalization;

[0026] Based on the dimensionless processed original data matrix, obtain the indicator variation ratio of the computing efficiency indicator of each data center to be evaluated;

[0027] According to the indicator variation ratio, obtain the indicator information entropy corresponding to each calculation efficiency indicator;

[0028] Based on the indicator information entropy, the objective weight corresponding to each calculation efficiency indicator is obtained.

[0029] In one embodiment, the subjective weight and the objective weight are weighted and integrated to obtain the subjective and objective comprehensive weights corresponding to each calculation efficiency indicator, including:

[0030] According to the subjective weight and the objective weight, a subjective weight vector and an objective weight vector are constructed;

[0031] Based on the subjective weight vector and the objective weight vector, the subjective and objective comprehensive weights corresponding to each calculation efficiency index are obtained through the linear weighting method.

[0032] In one embodiment, based on the data of the computing efficiency index and the subjective and objective comprehensive weights, a comprehensive evaluation method is used to obtain the comprehensive evaluation results of each data center to be evaluated, including:

[0033] The data of the calculation efficiency index is corrected by adjusting the coefficient to obtain the corrected index data of each calculation efficiency index, and an adjustment matrix is ​​constructed based on the corrected index data;

[0034] The elements of the adjustment matrix are normalized by using a vector normalization method to obtain a standardized matrix;

[0035] Combining the subjective and objective comprehensive weights of each calculation efficiency index, the standardized matrix is ​​weighted to obtain a weighted standardized matrix;

[0036] According to the weighted normalization matrix, positive and negative ideal solutions of the weighted normalization matrix are determined;

[0037] Through the positive and negative ideal solutions, the Euclidean distance corresponding to each data center to be evaluated is obtained;

[0038] According to the Euclidean distance, the relative closeness of each data center to be evaluated is obtained; the relative closeness is used to characterize the comprehensive evaluation result of the data center to be evaluated.

[0039] In one embodiment, based on the comprehensive evaluation results, the short board indicators of computing efficiency are determined, including:

[0040] Based on the Euclidean distance, the dimension evaluation scores of each evaluation dimension of each data center to be evaluated are obtained, and the dimension scoring matrix is ​​constructed according to the evaluation dimension scores;

[0041] Determine the best data center among all the data centers to be evaluated through comprehensive evaluation results;

[0042] According to the dimension scoring matrix, obtain the deviation rate of the evaluation dimension scores of each data center to be evaluated and the optimal data center;

[0043] Based on the deviation rate corresponding to each evaluation dimension, determine the evaluation dimension with the largest deviation contribution;

[0044] Based on the data of the evaluation dimensions and calculation efficiency indicators that contribute the most to the deviation, an indicator deviation analysis matrix is ​​constructed;

[0045] Based on the indicator deviation analysis matrix and combined with the subjective and objective comprehensive weights, the indicator deviation contribution of the calculation efficiency indicator in the evaluation dimension with the largest deviation contribution is obtained;

[0046] The deviation contribution of each indicator is sorted to obtain the short board indicator of calculation efficiency.

[0047] In a second aspect, in one embodiment, the present application provides a performance evaluation device, the device comprising:

[0048] The data acquisition module is used to obtain the data of the computing efficiency indicators of each data center to be evaluated;

[0049] The indicator weight determination module is used to determine the subjective and objective comprehensive weights of the calculation efficiency indicators based on the data of the calculation efficiency indicators;

[0050] The comprehensive evaluation module is used to obtain the comprehensive evaluation results of each data center to be evaluated through a comprehensive evaluation method based on the data of the computing efficiency indicators and the subjective and objective comprehensive weights;

[0051] The shortcoming assessment module is used to determine the computing efficiency shortcoming indicators based on the comprehensive evaluation results; the computing efficiency shortcoming indicators represent computing efficiency indicators whose indicator performance does not meet the energy consumption requirements.

[0052] In a third aspect, in an example, the present application provides a computer device including a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps in each method embodiment of the above-mentioned first aspect when executing the computer program.

[0053] In a fourth aspect, in an example, the present application provides a computer-readable storage medium having a computer program stored thereon, and when the computer program is executed by a processor, the steps in each method embodiment of the above-mentioned first aspect are implemented.

[0054] In a fifth aspect, in an example, the present application provides a computer program product, including a computer program, which implements the steps in the method embodiments of the above-mentioned first aspect when the computer program is executed by a processor.

[0055] The above-mentioned performance evaluation method, device, computer equipment, computer-readable storage medium and computer program product determine the subjective and objective comprehensive weights of the computing efficiency indicators of each data center to be evaluated by obtaining the data of the computing efficiency indicators of each data center to be evaluated, and obtain the comprehensive evaluation results of each data center to be evaluated by using the comprehensive evaluation method based on the data of the computing efficiency indicators and the subjective and objective comprehensive weights, and then determine the computing efficiency short board indicators of the data center based on the comprehensive evaluation results. In the above manner, the present application can more accurately evaluate the computing efficiency of the data center and obtain the current performance short board of the data center. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] In order to more clearly illustrate the technical solutions in the embodiments of the present application or related technologies, the drawings required for use in the embodiments of the present application or related technical descriptions will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without paying creative work.

[0057] Figure 1 is a schematic flow chart of a performance evaluation method in one embodiment;

[0058] Figure 2 A schematic diagram of computing efficiency indicator types involved in computing efficiency evaluation in one embodiment;

[0059] Figure 3 A schematic diagram of a process for obtaining the subjective and objective comprehensive weights of calculation efficiency indicators in an embodiment;

[0060] Figure 4 A schematic diagram of a process for obtaining a subjective weight of a calculation efficiency indicator in one embodiment;

[0061] Figure 5 A schematic diagram of a flow chart for obtaining an objective weight of a calculation efficiency indicator in one embodiment;

[0062] Figure 6 A schematic diagram of a flow chart for obtaining the subjective and objective comprehensive weights of calculation efficiency indicators in another embodiment;

[0063] Figure 7 A schematic diagram of a process for obtaining the relative proximity of a data center to be evaluated in one embodiment;

[0064] Figure 8 A schematic diagram of a process for obtaining a short board indicator of computing efficiency in one embodiment;

[0065] Fig. 9 is a structural block diagram of a performance evaluation device in one embodiment;

[0066] Fig.10 FIG. 4 is a diagram showing the internal structure of a computer device in one embodiment. DETAILED DESCRIPTION

[0067] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0068] With the development of information technologies such as artificial intelligence, big data, and the Internet of Things, the demand for computing power has increased significantly, and computing power has become a new type of productivity supporting the development of the digital economy. Among them, data centers, as key infrastructure, their computing power and energy consumption issues cannot be ignored. It is understandable that the efficient use of computing resources and energy-saving management of data centers can be achieved through the evaluation of the computing efficiency of data centers, so the evaluation of the computing efficiency of data centers is crucial.

[0069] However, traditional computing efficiency evaluation methods generally evaluate the computing efficiency of data centers by calculating the ratio of the computing power of the data center to the power consumption of all IT equipment, that is, "computing power generated per watt of power". However, the actual service provision capabilities of the data center are not taken into account in traditional computing efficiency evaluation methods, and the evaluation method has a single measurement method and lacks comprehensiveness, resulting in the traditional computing efficiency evaluation method actually not being able to fully reflect the system performance characteristics of the data center's actual network, storage and other systems and the green energy-saving characteristics of the data center, resulting in the inability to analyze the weak aspects of the data center's provision capabilities through traditional computing efficiency evaluation methods.

[0070] To facilitate the understanding of the technical solution of the present invention, the technical feature terms that may appear in the embodiments of the present invention are first explained:

[0071] PUE (Power sage effectiveness): is the ratio of the total energy consumption of the data center to the energy consumption of IT (Internet Technology) equipment;

[0072] CUE (Carbon usage effectiveness): is the ratio of the total carbon dioxide emissions of the data center to the energy consumption of IT equipment;

[0073] TOPSIS (Technique for Order Preference by Similarity to an Ideal Solution) is a multi-index evaluation method that can comprehensively evaluate and rank each evaluation object by calculating the distance between each evaluation object and the positive and negative ideal solutions.

[0074] In one embodiment, Figure 1As shown, a performance evaluation method is provided. This embodiment uses the method applied to a server as an example for illustration. It can be understood that the method can also be applied to a terminal, and can also be applied to a system including a terminal and a server, and is implemented through the interaction between the terminal and the server. In this embodiment, the method includes the following steps S102 to S108. Among them:

[0075] Step S102, obtaining the data of the computing efficiency index of each data center to be evaluated.

[0076] The data center to be evaluated may be a data center included in the scope of computing efficiency evaluation, and the data center may provide users with the required computing resources. The computing efficiency index may include various performance indicators that reflect the system performance characteristics of the data center and the green energy-saving characteristics of the data center.

[0077] Exemplarily, the data of the computing efficiency index may include the general computing power, intelligent computing power and super computing power of the data center, which are index data that characterize the system performance characteristics of the data center. The data of the computing efficiency index may also include the PUE index and CUE index of the data center, which are index data that characterize the green energy-saving characteristics of the data center. It can be understood that the types of the above-mentioned computing efficiency indicators are not limited to the implementation methods mentioned in the above-mentioned embodiments, and may also include other index data such as the total storage amount and the storage-computation ratio. The embodiments of this application do not specifically limit the types of the data of the computing efficiency indicators.

[0078] Specifically, the server can obtain data on various computing efficiency indicators from each data center to be evaluated, and perform computing efficiency evaluation of the data center based on the data on these computing efficiency indicators.

[0079] Step S104, based on the data of the calculation efficiency index, determine the subjective and objective comprehensive weights of the calculation efficiency index.

[0080] Among them, the subjective and objective comprehensive weights of the calculation efficiency indicators can be used to characterize the relative importance of various calculation efficiency indicators in the calculation efficiency evaluation process. Exemplarily, based on the data of the calculation efficiency indicators, the subjective and objective comprehensive weights of each calculation efficiency indicator can be obtained through the subjective and objective comprehensive weighting method.

[0081] Specifically, the server can determine the subjective and objective comprehensive weights of various computing efficiency indicators in the computing efficiency evaluation process based on the obtained computing efficiency indicator data, so that the computing efficiency evaluation process is more in line with the actual application scenario and the accuracy of the computing efficiency evaluation results is improved.

[0082] Step S106, obtaining the comprehensive evaluation results of each data center to be evaluated through a comprehensive evaluation method according to the data of the calculation efficiency index and the subjective and objective comprehensive weights.

[0083] Among them, the comprehensive evaluation method can be used to calculate the comprehensive computing efficiency performance score of the data center, and the comprehensive computing efficiency performance score is used to characterize the comprehensive computing efficiency performance of the data center.

[0084] Optionally, the comprehensive evaluation method may include a TOPSIS method using an adjustment coefficient to modify the weights. Exemplarily, based on the data of each computing efficiency indicator and the subjective and objective comprehensive weights, an evaluation model may be established according to the TOPSIS method using an adjustment coefficient to modify the weights; based on the evaluation model, the comprehensive computing efficiency performance score of each data center to be evaluated may be calculated, so that the comprehensive evaluation results of each data center may be determined according to the comprehensive computing efficiency performance score.

[0085] Specifically, the server can obtain the comprehensive evaluation results corresponding to each data center to be evaluated based on the data of the computing efficiency indicators and the subjective and objective comprehensive weights of each computing efficiency indicator using a comprehensive evaluation method, thereby achieving a comprehensive computing efficiency evaluation of the data center.

[0086] Step S108, based on the comprehensive evaluation results, determine the short board indicators of computing efficiency.

[0087] Among them, the short board index of computing efficiency can indicate the computing efficiency index of the data center that the index performance does not meet the energy consumption requirements. It can be understood that the short board index of computing efficiency represents the weak aspects of the business capabilities and / or energy consumption performance of the data center. By determining the short board index of computing efficiency, an objective basis is provided for optimizing the comprehensive computing efficiency of the data center and determining the improvement direction of the data center.

[0088] Specifically, the server can use the comprehensive evaluation results obtained to further determine the computing efficiency shortcomings indicators whose performance does not meet the energy consumption requirements, so that the weak aspects of the indicators of each data center to be evaluated can be obtained through the computing efficiency shortcomings indicator analysis, which will facilitate the optimization of the comprehensive computing efficiency of the data center and determine the future improvement direction of the data center.

[0089] The above performance evaluation method can determine the subjective and objective comprehensive weights of each computing efficiency indicator of the data center by obtaining the data of the computing efficiency indicators of each data center to be evaluated, and obtain the comprehensive evaluation results of each data center to be evaluated through a comprehensive evaluation method based on the data of the computing efficiency indicators and the subjective and objective comprehensive weights, and then determine the computing efficiency short board indicators of the data center based on the comprehensive evaluation results. This application makes the results of computing efficiency evaluation more in line with actual application scenarios by determining the subjective and objective comprehensive weights of each computing efficiency indicator, and also obtains the weak aspects of the indicators of each data center to be evaluated by obtaining the computing efficiency short board indicators for analysis, which is conducive to further optimizing the comprehensive computing efficiency of the data center.

[0090] In one embodiment, obtaining data on computing efficiency indicators of each data center to be evaluated includes:

[0091] Based on multiple evaluation dimensions of the data center to be evaluated, the data of computing efficiency indicators of each data center to be evaluated is obtained.

[0092] The evaluation dimension may represent at least one type of evaluation influencing factor of the data center computing efficiency.

[0093] It can be understood that by obtaining data on computing efficiency indicators of multiple evaluation dimensions of the data center to be evaluated, a multi-dimensional computing efficiency evaluation can be performed on the data center to be evaluated based on the data on computing efficiency indicators of multiple evaluation dimensions, so that the results of the computing efficiency evaluation are more comprehensive and accurate, and the evaluation results can more accurately reflect the actual business provision capabilities of the data center.

[0094] Specifically, the server can obtain the data of computing efficiency indicators of each data center to be evaluated based on multiple evaluation dimensions, thereby realizing multi-dimensional computing efficiency performance analysis of the data center to be evaluated, making the computing efficiency evaluation more objective and comprehensive.

[0095] In one embodiment, the evaluation dimensions include a first type of evaluation influencing factors, a second type of evaluation influencing factors, a third type of evaluation influencing factors, and a fourth type of evaluation influencing factors;

[0096] Among them, the first type of evaluation influencing factors corresponds to the computing power of the data center, the second type of evaluation influencing factors corresponds to the storage capacity of the data center, the third type of evaluation influencing factors corresponds to the carrying capacity of the data center, and the fourth type of evaluation influencing factors corresponds to the low-carbon power of the data center.

[0097] For example, computing power can represent the comprehensive computing capabilities of a data center's servers to process input data and output results. For example, computing power can include general computing power, intelligent computing power, and supercomputing power.

[0098] Optionally, the evaluation dimension corresponding to the first type of evaluation influencing factors may be a computing power evaluation dimension. In some examples, the computing power evaluation dimension includes computing efficiency indicators such as general computing power scale, intelligent computing power scale, intelligent computing power proportion, and shelf rate. The computing efficiency indicators in the computing power evaluation dimension can be used to evaluate the computing power scale, efficiency, and supply and demand status of a data center.

[0099] It can be understood that the computing efficiency indicators under the above-mentioned computing power evaluation dimension are not limited to the implementation methods mentioned in the above embodiments. As long as they can be used to evaluate the computing power scale, efficiency and supply and demand conditions of the data center, this application does not specifically limit the type of computing efficiency indicators under the computing power evaluation dimension.

[0100] Exemplarily, storage capacity may represent and measure the data storage capacity of a data center. For example, the data storage capacity may be the storage capacity of an external storage device of a data center and / or a built-in storage device of a server.

[0101] Optionally, the evaluation dimension corresponding to the second type of evaluation influencing factors may be a storage capacity evaluation dimension. In some examples, the storage capacity evaluation dimension includes computing efficiency indicators such as total storage volume, storage-to-computation ratio, advanced storage ratio, and single-rack storage capacity. Each computing efficiency indicator in the storage capacity evaluation dimension can be used to comprehensively evaluate the storage performance level of a data center.

[0102] It can be understood that the computing efficiency indicators under the above-mentioned storage capacity evaluation dimension are not limited to the implementation methods mentioned in the above embodiments. As long as they can be used to evaluate the storage performance level of the data center, the present application does not specifically limit the type of computing efficiency indicators under the storage capacity evaluation dimension.

[0103] For example, carrying capacity can characterize the performance of a data center in terms of network architecture, bandwidth, transmission latency, and intelligent management and scheduling, such as the data center access capability and data carrying efficiency of the data center.

[0104] Optionally, the evaluation dimension corresponding to the third type of evaluation influencing factors may be a carrying capacity evaluation dimension. In some examples, the carrying capacity evaluation dimension includes computing efficiency indicators such as network exit bandwidth and latency average, and each computing efficiency indicator in the carrying capacity evaluation dimension can be used to evaluate the comprehensive level of the network carrying capacity of the data center.

[0105] It can be understood that the computing efficiency indicators under the above-mentioned carrying capacity evaluation dimension are not limited to the implementation methods mentioned in the above embodiments. As long as they can be used to evaluate the comprehensive level of network carrying capacity of the data center, the present application does not specifically limit the type of computing efficiency indicators under the carrying capacity evaluation dimension.

[0106] For example, low-carbon power is centered on the carbon efficiency level of computing power, and can characterize the performance of data centers in the application of green and low-carbon equipment technology and renewable energy.

[0107] Optionally, the evaluation dimension corresponding to the fourth type of evaluation influencing factors may be a low-carbon power evaluation dimension. In some examples, the low-carbon power evaluation dimension includes computing efficiency indicators such as PUE (Power sage effectiveness), CUE (Carbon usage effectiveness), and green electricity usage ratio. Each computing efficiency indicator in the low-carbon power evaluation dimension can be used to measure the low-carbon performance of each data center to be evaluated.

[0108] It is understandable that the computing efficiency indicators under the above-mentioned low-carbon force evaluation dimension are not limited to the implementation methods mentioned in the above embodiments. As long as they can be used to measure the low-carbon performance of each data center to be evaluated, this application does not specifically limit the type of computing efficiency indicators under the low-carbon force evaluation dimension.

[0109] In one possible implementation, Figure 2 As shown, the server can be based on Figure 2 The multiple evaluation dimensions shown perform multi-dimensional computing efficiency evaluation on the data center. This embodiment can improve the comprehensiveness and accuracy of the computing efficiency evaluation results by obtaining the computing efficiency index data of the data center to be evaluated from multiple evaluation dimensions.

[0110] In one embodiment, if Figure 3 As shown, based on the data of the calculation efficiency index, determining the subjective and objective comprehensive weights of the calculation efficiency index includes the following steps S302 to S306. Among them:

[0111] Step S302: Based on the subjective weighting method, the subjective weight corresponding to each calculation efficiency indicator is obtained.

[0112] Exemplarily, the subjective weighting method may include the analytic hierarchy process, which can be used to obtain the subjective weights corresponding to each efficiency indicator. It can be understood that the subjective weighting method can be applied to situations where the indicator system is complex and data is difficult to collect.

[0113] Specifically, the server can use the subjective weighting method to obtain the subjective weight corresponding to each calculation efficiency indicator. By adopting the subjective weighting method, the present application can effectively improve the versatility of the method.

[0114] Step S304: Based on the objective weighting method, the objective weight corresponding to each calculation efficiency indicator is obtained.

[0115] Exemplarily, the objective weighting method may include an entropy weight method, and the objective weight corresponding to each calculation efficiency indicator may be obtained by using the entropy weight method.

[0116] Specifically, the server can use the objective weighting method to obtain the objective weight corresponding to each computing efficiency indicator. By adopting the objective weighting method, the present application can effectively improve the rigor of the computing efficiency evaluation result.

[0117] Step S306, weighted fusion of the subjective weight and the objective weight is performed to obtain the subjective and objective comprehensive weight corresponding to each calculation efficiency indicator.

[0118] For example, the obtained subjective weight and objective weight can be weighted and integrated through a linear weighting method to obtain a subjective and objective comprehensive weight corresponding to the calculation efficiency index.

[0119] Specifically, the server can perform weighted fusion on the obtained subjective weights and objective weights to obtain the subjective and objective comprehensive weights corresponding to each computing efficiency indicator. This application makes the evaluation results of the data center to be evaluated more in line with the actual situation by obtaining the subjective and objective comprehensive weights corresponding to each computing efficiency indicator, thereby improving the accuracy of the evaluation results.

[0120] In one embodiment, if Figure 4 As shown, based on the subjective weighting method, the subjective weight corresponding to each calculation efficiency indicator is obtained, including steps S402 to S408. Among them:

[0121] Step S402: compare the calculation efficiency indicators to obtain the indicator comparison results.

[0122] The indicator comparison result may include the result of pairwise comparison between the current computing efficiency indicator and other computing efficiency indicators involved in the computing efficiency evaluation.

[0123] Specifically, the server can compare the computing efficiency indicators involved in the computing efficiency evaluation in pairs to obtain the indicator comparison results. For example, the computing efficiency indicator of "general computing power scale" under the computing power evaluation dimension can be compared with the "intelligent computing power scale" under the same evaluation dimension, and can also be compared with the "total storage" computing efficiency indicator under the storage power dimension. Furthermore, based on the knowledge and experience of experts, the subjective weighting method is used to judge the importance of the computing efficiency indicators in pairs to obtain the corresponding indicator comparison results.

[0124] Step S404: construct a judgment matrix according to the index comparison result, and obtain the maximum eigenvalue of the judgment matrix.

[0125] Among them, the indicator comparison results can replace the absolute scale of the calculation efficiency indicator by the relative scale of the comparison of the calculation efficiency indicators pairwise.

[0126] Specifically, the server may construct a judgment matrix based on the indicator comparison results obtained by comparing the calculation efficiency indicators pairwise, and obtain the maximum eigenvalue of the judgment matrix.

[0127] For example, for those participating in the calculation efficiency evaluation Calculation efficiency indicators, each of which is compared with other calculation efficiency indicators The second comparison, The efficiency index and The comparison results of the calculation efficiency indicators can be recorded as , No. The comparison result of the calculation efficiency index with itself is , so based on the above index comparison results, we can construct a In some examples, as shown in the following formula 1, the judgment matrix obtained is It can be expressed as:

[0128] (Formula 1)

[0129] Among them, the judgment matrix The elements in satisfy the following conditions: , , .

[0130] Furthermore, a scale of 1 to 9 can be used to characterize the relative importance of the calculation efficiency indicators. Optionally, the element scale and meaning of the judgment matrix can refer to the AHP (Analytics Hierarchy Process) scale table shown in Table 1 below:

[0131] Table 1

[0132]

[0133] Step S406: based on the maximum eigenvalue of the judgment matrix, when it is determined that the judgment matrix meets the consistency requirement, the matrix eigenvalue vector of the judgment matrix is ​​solved.

[0134] Among them, the consistency requirement characterization judgment matrix belongs to the consistency matrix.

[0135] Specifically, the server may determine whether the judgment matrix meets the consistency requirement based on the obtained maximum eigenvalue of the judgment matrix, and if the judgment matrix meets the consistency requirement, further solve the matrix eigenvalue vector of the judgment matrix.

[0136] Exemplarily, determining whether the judgment matrix meets the consistency requirement includes: using a random consistency ratio coefficient indicator to judge the consistency of the matrix. In some examples, the above consistency ratio coefficient indicator can be calculated according to the following formula 2 and formula 3:

[0137] (Formula 2)

[0138] (Formula 3)

[0139] in, is the consistency ratio index, is the consistency indicator, is the average random consistency index, represents the maximum eigenvalue of the judgment matrix, Represents the order of the judgment matrix.

[0140] Furthermore, the average random consistency index According to the random consistency index shown in Table 2 Look up the corresponding value in the standard value:

[0141] Table 2

[0142]

[0143] Understandably, The smaller the value, the better the consistency of the judgment matrix. During the inspection, if If the value is less than 0.1, the judgment matrix is ​​determined to be a consistency matrix. At this time, the judgment matrix meets the consistency requirements. Otherwise, the judgment matrix needs to be corrected or reconstructed.

[0144] Step S408, normalizing the matrix eigenvector of the judgment matrix to obtain the subjective weight corresponding to each calculation efficiency index.

[0145] Specifically, after the server normalizes the matrix eigenvector of the judgment matrix, the subjective weight corresponding to each calculation efficiency indicator at this time can be obtained.

[0146] In some examples, as shown in the following formula 4, the eigenvalue method can be used to calculate the subjective weight corresponding to each calculation efficiency indicator:

[0147] (Formula 4)

[0148] in, is the corresponding eigenvector, is the judgment matrix, Represents the judgment matrix Further, the eigenvector obtained above is After normalization, the subjective weights corresponding to each efficiency indicator can be obtained. .

[0149] In one embodiment, if Figure 5 As shown, based on the objective weighting method, the objective weight corresponding to each calculation efficiency index is obtained, including the following steps S502 to S510. Among them:

[0150] Step S502: construct an original data matrix based on the data of the calculation efficiency index.

[0151] Specifically, the original data matrix can be constructed based on the collected data of the computing efficiency indicators of each data center. In some examples, according to the following formula 5, the matrix composed of the computing efficiency indicator data can be expressed as:

[0152] (Formula 5)

[0153] in, It is The data center to be evaluated participated in the computing efficiency evaluation The data of the calculation efficiency index, is the number of data centers to be evaluated.

[0154] Step S504, performing dimensionless processing on the original data matrix through forward transformation and normalization.

[0155] Among them, the calculation efficiency indicators corresponding to each element in the original data matrix include positive indicators and negative indicators, and the units and scales of each calculation efficiency indicator are inconsistent.

[0156] Specifically, based on the obtained original data matrix, the server can perform dimensionless processing on the elements of the original data matrix by performing forward and normalization processing on each element of the original data matrix, and solve the problem of inconsistent units and scales of various calculation efficiency indicators.

[0157] Optionally, the range method may be used to normalize and normalize the elements in the original data matrix so as to perform dimensionless processing on the original data matrix.

[0158] Exemplarily, the forward and normalization processing includes:

[0159] If the target element of the current original data matrix corresponds to a positive indicator, the target element can be positively converted and normalized by the following operation process as shown in Formula 6:

[0160] (Formula 6)

[0161] If the target element of the current original data matrix corresponds to a negative indicator, the target element can be positively converted and normalized by the following operation process as shown in Formula 7:

[0162] (Formula 7)

[0163] in, is the element of the original data matrix after forward and normalization, Indicates the number of participants in the calculation efficiency evaluation A data set of calculation efficiency indicators, It is Data center to be reviewed The data of the calculation efficiency index, is the element with the largest value in the original data matrix before forward and normalization processing, is the element with the smallest value in the original data matrix before forward and normalization, that is and Respectively represent the first The minimum and maximum values ​​of the calculation efficiency indicators.

[0164] Step S506, based on the dimensionless processed original data matrix, obtain the indicator variation ratio of the computing efficiency indicator of each data center to be evaluated.

[0165] Among them, the indicator variation ratio can be used to characterize the degree of variation (or variation size) of the data of the calculation efficiency indicator.

[0166] Specifically, the server can obtain the indicator variation ratio of the computing efficiency indicators of each data center to be evaluated based on the original data matrix after dimensionless processing.

[0167] For example, based on the dimensionless processed original data matrix, the indicator variation ratio of the computing efficiency indicator of each data center to be evaluated can be calculated through the operation process shown in the following formula 8:

[0168] (Formula 8)

[0169] in, Indicates The data center to be evaluated The variation ratio of the index of each efficiency index, is the element of the original data matrix after dimensionless processing, is the number of data centers to be evaluated.

[0170] Step S508: Obtain the indicator information entropy corresponding to each calculation efficiency indicator according to the indicator variation ratio.

[0171] Among them, the indicator information entropy can characterize the expected amount of information of the calculation efficiency indicator.

[0172] Specifically, based on the indicator variation proportions of the computing efficiency indicators of the data centers to be evaluated obtained by the above calculations, the indicator information entropy corresponding to each computing efficiency indicator can be further calculated.

[0173] For example, the indicator information entropy corresponding to each calculation efficiency indicator can be calculated through the calculation process shown in the following formula 9:

[0174] (Formula 9)

[0175] in, Indicates The information entropy of the indicator corresponding to the calculation efficiency indicator is: Indicates The data center to be evaluated The variation ratio of the index of the calculation efficiency index is hour, .

[0176] Step S510: Obtain the objective weight corresponding to each calculation efficiency indicator based on the indicator information entropy.

[0177] Optionally, based on the calculation process shown in the following formula 10, the objective weight corresponding to each calculation efficiency indicator can be determined:

[0178] (Formula 10)

[0179] in, Indicates the objective weight corresponding to each calculation efficiency indicator, Indicates The information entropy of the indicator corresponding to the calculation efficiency indicator is: Indicates the number of computing efficiency indicators involved in computing efficiency evaluation.

[0180] Specifically, the server can determine the objective weight corresponding to each computing efficiency indicator based on the indicator information entropy corresponding to each computing efficiency indicator.

[0181] In one embodiment, if Figure 6 As shown, the subjective weight and the objective weight are weighted and integrated to obtain the subjective and objective comprehensive weights corresponding to each calculation efficiency index, including the following steps S602 to S604. Among them:

[0182] Step S602: construct a subjective weight vector and an objective weight vector according to the subjective weight and the objective weight.

[0183] The subjective weight vector may include the subjective weights corresponding to the various calculation efficiency indicators, and the objective weight vector may include the objective weights corresponding to the various calculation efficiency indicators.

[0184] Specifically, the server may construct a subjective weight vector and an objective weight vector according to the subjective weights and objective weights corresponding to the obtained calculation efficiency indicators.

[0185] Step S604: Based on the subjective weight vector and the objective weight vector, the subjective and objective comprehensive weights corresponding to each calculation efficiency index are obtained by linear weighting method.

[0186] Among them, the linear weighted method can be used to calculate the subjective and objective comprehensive weights corresponding to each calculation efficiency indicator in a linear weighted manner.

[0187] Specifically, the server can use a linear weighting method to obtain the subjective and objective comprehensive weights corresponding to each calculation efficiency indicator based on the obtained subjective weight vector and objective weight vector.

[0188] For example, according to the subjective weight vector and the objective weight vector, the subjective and objective comprehensive weights corresponding to each calculation efficiency index can be calculated through the operation process shown in the following formula 11:

[0189] (Formula 11)

[0190] in, Indicates the number of participants in the calculation efficiency evaluation The subjective and objective comprehensive weights corresponding to the calculation efficiency indicators are: is the comprehensive weight, For the The subjective weight corresponding to each effectiveness indicator is For the The objective weight corresponding to each efficiency indicator.

[0191] Furthermore, the comprehensive weighting weight can be determined by the difference coefficient method. For example, the comprehensive weighting weight can be determined by the difference coefficient method according to the calculation process shown in the following formula 12:

[0192] (Formula 12)

[0193] in, is the comprehensive weight, J is the number of computing efficiency indicators involved in computing efficiency evaluation, It is the numerical value after the subjective weight of the calculation efficiency index is sorted from small to large. Determining the comprehensive weight by the difference coefficient method can further improve the objectivity of the calculation result, thereby improving the accuracy of the calculation efficiency evaluation result.

[0194] In one embodiment, if Figure 7 As shown, based on the data of the calculation efficiency index and the subjective and objective comprehensive weights, the comprehensive evaluation results of each data center to be evaluated are obtained through a comprehensive evaluation method, including the following steps S702 to S712.

[0195] Step S702, correcting the data of the calculation efficiency index by adjusting the coefficient to obtain the corrected index data of each calculation efficiency index, and constructing an adjustment matrix based on the corrected index data.

[0196] The adjustment coefficient can be used to characterize the gap between the actual performance of the computing efficiency indicator of the data center to be evaluated and the expected target performance.

[0197] Specifically, the server can use the adjustment coefficient to correct the original computing efficiency index data, obtain the corrected index data corresponding to the computing efficiency index of each data center to be evaluated, and construct a corresponding adjustment matrix based on the corrected index data.

[0198] For example, the evaluation data is corrected according to the ratio between the actual performance and the expected target, that is, the original calculation efficiency index data is multiplied by the corresponding adjustment coefficient to obtain the index value after the achievement. In some examples, such as the calculation process shown in the following formula 13, the adjustment coefficient can be expressed as follows:

[0199] (Formula 13)

[0200] in, Indicates Data center to be reviewed The adjustment coefficient of the efficiency index is For the Data center to be reviewed The data of the calculation efficiency index, Indicates The expected target value of the calculation efficiency index, the value range of the adjustment coefficient is Furthermore, if the ratio of the calculation efficiency index data to the expected target value is greater than 1, the adjustment coefficient is set to 1.

[0201] In some examples, as shown in the following formula 14, the adjustment matrix constructed based on the modified indicator data can be expressed as:

[0202] (Formula 14)

[0203] Among them, B is the adjustment matrix, For the Data center to be reviewed Corrected index data of each efficiency index, For the Data center to be reviewed The data of the calculation efficiency index, is the number of data centers to be evaluated, The number of computing efficiency indicators involved in computing efficiency evaluation.

[0204] Step S704: normalize the elements of the adjustment matrix using a vector normalization method to obtain a standardized matrix.

[0205] Among them, vector normalization can be used to standardize the elements of the matrix.

[0206] Specifically, based on the obtained adjustment matrix, the server may use a vector normalization method to normalize the elements of the adjustment matrix, thereby obtaining a standardized matrix.

[0207] Exemplarily, in the operation process shown in the following formula 15 and formula 16, the elements of the adjustment matrix B are normalized by using the vector normalization method to obtain a standardized matrix:

[0208] (Formula 15)

[0209] (Formula 16)

[0210] in, is the standardized matrix, To correspond to Data center to be reviewed The standardized matrix elements of the efficiency indicators are To adjust the matrix B corresponding to the Data center to be reviewed Corrected index data of each efficiency index, is the number of data centers to be evaluated, The number of computing efficiency indicators involved in computing efficiency evaluation.

[0211] Step S706, weighting the standardized matrix in combination with the subjective and objective comprehensive weights of each calculation efficiency index to obtain a weighted standardized matrix.

[0212] Specifically, the server further performs weighting processing on the above-mentioned standardized matrix based on the subjective and objective comprehensive weights of each calculation efficiency index, thereby obtaining a weighted standardized matrix.

[0213] For example, as shown in the following formula 17, after weighting the standardized matrix by combining the subjective and objective comprehensive weights of each calculation efficiency index, the obtained weighted standardized matrix can be expressed as:

[0214] (Formula 17)

[0215] Where D is the weighted normalization matrix, Indicates the number of participants in the calculation efficiency evaluation The subjective and objective comprehensive weights corresponding to the calculation efficiency indicators are: To correspond to Data center to be reviewed The elements of the standardized matrix C of the calculation efficiency index, is the number of data centers to be evaluated, The number of computing efficiency indicators involved in computing efficiency evaluation.

[0216] Step S708: determining positive and negative ideal solutions of the weighted normalization matrix according to the weighted normalization matrix.

[0217] Specifically, the server may determine the positive and negative ideal solutions corresponding to the weighted normalization matrix based on the elements of the weighted normalization matrix obtained above.

[0218] Exemplarily, the positive and negative ideal solutions corresponding to the weighted normalization matrix may be determined by the calculation process shown in the following formula 18 and formula 19:

[0219] (Formula 18)

[0220] (Formula 19)

[0221] in, is the positive ideal solution among the positive and negative ideal solutions, is the negative ideal solution among the positive and negative ideal solutions, is the weighted normalization matrix corresponding to Data center to be reviewed elements of an efficiency indicator.

[0222] Step S710, obtaining the Euclidean distance corresponding to each data center to be evaluated through positive and negative ideal solutions.

[0223] Specifically, after obtaining the positive and negative ideal solutions corresponding to the weighted normalization matrix, the Euclidean distance corresponding to each data center to be evaluated can be further calculated.

[0224] Exemplarily, according to the calculation process shown in the following formula 20, the Euclidean distance of each data center to be evaluated corresponding to the positive ideal solution can be obtained:

[0225] (Formula 20)

[0226] in, For the The Euclidean distance of the positive ideal solution corresponding to the data center to be evaluated, is a positive ideal solution, is the weighted normalization matrix corresponding to Data center to be reviewed The elements of the calculation efficiency index, The number of computing efficiency indicators involved in computing efficiency evaluation.

[0227] Exemplarily, according to the calculation process shown in the following formula 21, the Euclidean distance of each data center to be evaluated corresponding to the negative ideal solution can be obtained:

[0228] (Formula 21)

[0229] in, For the The Euclidean distance between the negative ideal solution of the data center to be evaluated, is a negative ideal solution, is the weighted normalization matrix corresponding to Data center to be reviewed The elements of the calculation efficiency index, The number of computing efficiency indicators involved in computing efficiency evaluation.

[0230] Step S712: Obtain the relative closeness of each data center to be evaluated based on the Euclidean distance.

[0231] Among them, relative closeness can be used to characterize the comprehensive evaluation results of the data center to be evaluated.

[0232] Specifically, the server can further obtain the relative proximity of each data center to be evaluated based on the Euclidean distance between the positive ideal solution and the negative ideal solution corresponding to each data center to be evaluated, and use the relative proximity to represent the comprehensive score of the data center, so that based on the comprehensive score of the data center, the comprehensive evaluation results of each data center to be evaluated can be obtained.

[0233] Exemplarily, according to the calculation process shown in the following formula 22, the relative closeness of each data center to be evaluated can be calculated:

[0234] (Formula 22)

[0235] in, For the The relative proximity of the data centers to be evaluated, For the The Euclidean distance of the positive ideal solution corresponding to the data center to be evaluated, For the The Euclidean distance between the negative ideal solution and the data center to be evaluated.

[0236] Furthermore, according to the calculation process shown in the following formula 23, the comprehensive computing efficiency of each data center to be evaluated can be calculated:

[0237] (Formula 23)

[0238] in, For the The comprehensive computing efficiency of the data centers to be evaluated. For the The relative proximity of the data centers to be evaluated, For the The energy consumption of the data center to be evaluated.

[0239] In one embodiment, if Figure 8 As shown, based on the comprehensive evaluation results, determining the short board indicator of computing efficiency includes the following steps S802 to S814. Among them:

[0240] Step S802: Based on the Euclidean distance, the dimension evaluation score of each evaluation dimension of each data center to be evaluated is obtained, and a dimension scoring matrix is ​​constructed according to the evaluation dimension scores.

[0241] The Euclidean distance includes the Euclidean distance between the positive ideal solution and the negative ideal solution corresponding to the data center to be evaluated.

[0242] Specifically, the server can further obtain the dimension evaluation scores of each data center to be evaluated under each evaluation dimension according to the Euclidean distance between the positive ideal solution and the negative ideal solution corresponding to the data center to be evaluated, and construct a dimension scoring matrix based on these evaluation dimension scores.

[0243] Exemplarily, according to the calculation process shown in the following formula 24, formula 25 and formula 26, the dimension evaluation score of each data center to be evaluated under each evaluation dimension can be obtained:

[0244] (Formula 24)

[0245] (Formula 25)

[0246] (Formula 26)

[0247] in, For the Data center to be reviewed The dimension evaluation scores of the evaluation dimensions, is the weighted normalization matrix Data center to be reviewed The evaluation dimension The elements of the calculation efficiency index, for Corresponding to Positive ideal solution of evaluation dimension The Euclidean distance of for Corresponding to Negative ideal solution of evaluation dimension The Euclidean distance of For the The number of calculation efficiency indicators under each evaluation dimension.

[0248] Furthermore, as shown in the following formula 24, based on the above evaluation dimension scores, the constructed dimension scoring matrix can be expressed as:

[0249] (Formula 27)

[0250] Among them, V is the dimension rating matrix, For the Data center to be reviewed The dimension evaluation scores of the evaluation dimensions, is the number of data centers to be evaluated, is the number of evaluation dimensions involved in the effectiveness evaluation process.

[0251] Step S804: Determine the best data center among the data centers to be evaluated through comprehensive evaluation results.

[0252] Among them, the comprehensive evaluation result can be obtained according to the relative proximity of each data center.

[0253] Specifically, the server can further determine the optimal data center with the best comprehensive performance among the data centers to be evaluated that participate in the computing efficiency evaluation based on the comprehensive evaluation results obtained above.

[0254] Exemplarily, the optimal data center may be determined according to the calculation process shown in the following formula 28:

[0255] (Formula 28)

[0256] in, For the The relative proximity of the data centers to be evaluated, and when hour, Get the maximum value, that is, the data center to be evaluated (No. data centers to be evaluated) is the best data center; is the number of data centers to be evaluated.

[0257] Step S806, obtaining the deviation rate of the evaluation dimension scores of each data center to be evaluated and the optimal data center according to the dimension scoring matrix.

[0258] Specifically, the server can further obtain the deviation rate of the evaluation dimension score between each data center to be evaluated and the optimal data center based on the dimension scoring matrix constructed above.

[0259] Exemplarily, according to the calculation process shown in the following formula 29, the evaluation dimension scores of each data center to be evaluated and the optimal data center are compared and calculated, and the deviation rate of the evaluation dimension scores between each data center to be evaluated and the optimal data center can be calculated:

[0260] (Formula 29)

[0261] in, For the Data Centers to be Reviewed and Best Data Centers The deviation rate of the evaluation dimension scores between The best data center No. The dimension evaluation scores of the evaluation dimensions, For the Data center to be reviewed The dimension evaluation scores of the evaluation dimensions, is the number of data centers to be evaluated.

[0262] Step S808: Determine the evaluation dimension with the largest deviation contribution based on the deviation rates corresponding to the various evaluation dimensions.

[0263] Specifically, the server may further determine the evaluation dimension with the largest deviation contribution according to the deviation rates corresponding to the various evaluation dimensions obtained above.

[0264] Exemplarily, according to the operation process shown in the following formula 30, the evaluation dimension with the largest deviation contribution is calculated:

[0265] (Formula 30)

[0266] in, For the Data Centers to be Reviewed and Best Data Centers The deviation rate of the evaluation dimension scores between hour, Get the maximum value, that is, the evaluation dimension (No. The evaluation dimension (out of 10 evaluation dimensions) is the evaluation dimension that contributes the most to the deviation; The number of evaluation dimensions involved in the calculation efficiency evaluation.

[0267] Step S810, constructing an indicator deviation analysis matrix based on the evaluation dimension with the largest deviation contribution and the data of the calculation efficiency indicator.

[0268] Specifically, based on the evaluation dimension with the largest deviation contribution obtained above and the data of each calculation efficiency indicator under the evaluation dimension, an indicator deviation analysis matrix can be further constructed.

[0269] For example, as shown in the following formula 31, if there is a maximum deviation contribution evaluation dimension The indicator deviation analysis matrix constructed can be expressed as:

[0270] (Formula 31)

[0271] Where X is the standard deviation analysis matrix, For the Data center to be reviewed The data of the calculation efficiency index, is the number of data centers to be evaluated.

[0272] Step S812, based on the indicator deviation analysis matrix and combined with the subjective and objective comprehensive weights, obtain the indicator deviation contribution of the calculation efficiency indicator in the evaluation dimension with the largest deviation contribution.

[0273] Specifically, the server further calculates the indicator deviation contribution of each computing efficiency indicator in the evaluation dimension with the largest deviation contribution for each data center to be evaluated based on the indicator deviation analysis matrix obtained above and the subjective and objective comprehensive weights corresponding to the corresponding computing efficiency indicators.

[0274] Exemplarily, according to the calculation process shown in the following formula 32, the indicator deviation corresponding to each calculation efficiency indicator in the evaluation dimension with the largest deviation contribution of each data center to be evaluated is calculated:

[0275] (Formula 32)

[0276] in, For the Data center to be reviewed The indicator deviation corresponding to the calculation efficiency indicator is For the Data center to be reviewed The data of the calculation efficiency index, is the number of data centers to be evaluated.

[0277] Furthermore, according to the calculation process shown in the following formula 33, combined with the subjective and objective comprehensive weights corresponding to each calculation efficiency indicator, the indicator deviation contribution corresponding to each calculation efficiency indicator in the evaluation dimension with the largest deviation contribution of each data center to be evaluated is calculated:

[0278] (Formula 33)

[0279] in, For the Data center to be reviewed The indicator deviation contribution corresponding to each calculation efficiency indicator is For the The subjective and objective comprehensive weights corresponding to the calculation efficiency indicators are: For the Data center to be reviewed The indicator deviation corresponding to each calculation efficiency indicator.

[0280] Step S814, sorting the deviation contributions of each indicator to obtain the calculation efficiency short board indicator.

[0281] Among them, the computing efficiency shortcoming index can indicate that the index performance of the data center does not meet the computing efficiency index of the energy consumption requirements.

[0282] Specifically, the server can sort and analyze the indicator deviation contributions obtained above, and can obtain the computing efficiency shortcoming indicators that affect the comprehensive computing efficiency of the data center based on the sorting results.

[0283] The above performance evaluation method can be applied to the evaluation and comparison of computing efficiency of data centers, and can also be further applied to the green construction, operation and management of computing power infrastructure. It can be understood that this application also includes at least the following possible application values: ① This application can provide a basis for the grading evaluation of data centers; ② This application can guide data centers to improve computing efficiency, identify the shortcomings of data center indicators, and provide optimization and improvement directions for data centers; ③ This application can standardize the computing efficiency evaluation system of data centers and promote the standardization process of multi-dimensional computing efficiency evaluation.

[0284] It should be understood that, although the various steps in the flowcharts involved in the above-mentioned embodiments are displayed in sequence according to the indication of the arrows, these steps are not necessarily executed in sequence according to the order indicated by the arrows. Unless there is a clear explanation in this article, the execution of these steps does not have a strict order restriction, and these steps can be executed in other orders. Moreover, at least a part of the steps in the flowcharts involved in the above-mentioned embodiments can include multiple steps or multiple stages, and these steps or stages 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 to be carried out in sequence, but can be executed in turn or alternately with other steps or at least a part of the steps or stages in other steps.

[0285] Based on the same inventive concept, the embodiment of the present application also provides a performance evaluation device for implementing the above-mentioned performance evaluation method. The implementation scheme for solving the problem provided by the device is similar to the implementation scheme recorded in the above-mentioned method, so the specific limitations in one or more performance evaluation device embodiments provided below can refer to the limitations on the performance evaluation method above, and will not be repeated here.

[0286] In an exemplary embodiment, Fig. 9 As shown, a performance evaluation device 900 is provided, comprising:

[0287] The data acquisition module 902 is used to acquire the data of the computing efficiency index of each data center to be evaluated;

[0288] The indicator weight determination module 904 is used to determine the subjective and objective comprehensive weights of the calculation efficiency indicators based on the data of the calculation efficiency indicators;

[0289] Comprehensive evaluation module 906, used to obtain comprehensive evaluation results of each data center to be evaluated through a comprehensive evaluation method according to the data of the calculation efficiency index and the subjective and objective comprehensive weights;

[0290] The shortcoming assessment module 908 is used to determine the computing efficiency shortcoming index based on the comprehensive evaluation results; the computing efficiency shortcoming index represents the computing efficiency index whose index performance does not meet the energy consumption requirements.

[0291] In one embodiment, the data acquisition module 902 is further used to:

[0292] Based on multiple evaluation dimensions of the data center to be evaluated, the data of the computing efficiency indicators of each data center to be evaluated is obtained; the evaluation dimension represents at least one type of evaluation influencing factor of the computing efficiency of the data center.

[0293] In one embodiment, the evaluation dimensions include a first type of evaluation influencing factors, a second type of evaluation influencing factors, a third type of evaluation influencing factors, and a fourth type of evaluation influencing factors;

[0294] Among them, the first type of evaluation influencing factors corresponds to the computing power of the data center, the second type of evaluation influencing factors corresponds to the storage capacity of the data center, the third type of evaluation influencing factors corresponds to the carrying capacity of the data center, and the fourth type of evaluation influencing factors corresponds to the low-carbon power of the data center.

[0295] In one embodiment, the indicator weight determination module 904 is further configured to:

[0296] Based on the subjective weighting method, the subjective weight corresponding to each calculation efficiency indicator is obtained;

[0297] Based on the objective weighting method, the objective weight corresponding to each calculation efficiency indicator is obtained;

[0298] The subjective weight and objective weight are weighted and integrated to obtain the subjective and objective comprehensive weights corresponding to each calculation efficiency indicator.

[0299] In one embodiment, the indicator weight determination module 904 is further configured to:

[0300] Compare the calculation efficiency indicators to obtain the indicator comparison results;

[0301] Construct a judgment matrix based on the index comparison results and obtain the maximum eigenvalue of the judgment matrix;

[0302] Based on the maximum eigenvalue of the judgment matrix, when it is determined that the judgment matrix meets the consistency requirement, the matrix eigenvalue vector of the judgment matrix is ​​solved;

[0303] The matrix eigenvector of the judgment matrix is ​​normalized to obtain the subjective weight corresponding to each calculation efficiency index.

[0304] In one embodiment, the indicator weight determination module 904 is further configured to:

[0305] Based on the data of the calculation efficiency index, construct the original data matrix;

[0306] The original data matrix is ​​dimensionless through forward transformation and normalization;

[0307] Based on the dimensionless processed original data matrix, obtain the indicator variation ratio of the computing efficiency indicator of each data center to be evaluated;

[0308] According to the indicator variation ratio, obtain the indicator information entropy corresponding to each calculation efficiency indicator;

[0309] Based on the indicator information entropy, the objective weight corresponding to each calculation efficiency indicator is obtained.

[0310] In one embodiment, the indicator weight determination module 904 is further configured to:

[0311] According to the subjective weight and the objective weight, a subjective weight vector and an objective weight vector are constructed;

[0312] Based on the subjective weight vector and the objective weight vector, the subjective and objective comprehensive weights corresponding to each calculation efficiency index are obtained through the linear weighting method.

[0313] In one embodiment, the comprehensive evaluation module 906 is further configured to:

[0314] The data of the calculation efficiency index is corrected by adjusting the coefficient to obtain the corrected index data of each calculation efficiency index, and an adjustment matrix is ​​constructed based on the corrected index data;

[0315] The elements of the adjustment matrix are normalized by using a vector normalization method to obtain a standardized matrix;

[0316] Combining the subjective and objective comprehensive weights of each calculation efficiency index, the standardized matrix is ​​weighted to obtain a weighted standardized matrix;

[0317] According to the weighted normalization matrix, positive and negative ideal solutions of the weighted normalization matrix are determined;

[0318] Through the positive and negative ideal solutions, the Euclidean distance corresponding to each data center to be evaluated is obtained;

[0319] According to the Euclidean distance, the relative closeness of each data center to be evaluated is obtained; the relative closeness is used to characterize the comprehensive evaluation result of the data center to be evaluated.

[0320] In one embodiment, the shortcoming assessment module 908 is further configured to:

[0321] Based on the Euclidean distance, the dimension evaluation scores of each evaluation dimension of each data center to be evaluated are obtained, and the dimension scoring matrix is ​​constructed according to the evaluation dimension scores;

[0322] Determine the best data center among all the data centers to be evaluated through comprehensive evaluation results;

[0323] According to the dimension scoring matrix, obtain the deviation rate of the evaluation dimension scores of each data center to be evaluated and the optimal data center;

[0324] Based on the deviation rate corresponding to each evaluation dimension, determine the evaluation dimension with the largest deviation contribution;

[0325] Based on the data of the evaluation dimensions and calculation efficiency indicators that contribute the most to the deviation, an indicator deviation analysis matrix is ​​constructed;

[0326] Based on the indicator deviation analysis matrix and combined with the subjective and objective comprehensive weights, the indicator deviation contribution of the calculation efficiency indicator in the evaluation dimension with the largest deviation contribution is obtained;

[0327] The deviation contribution of each indicator is sorted to obtain the short board indicator of calculation efficiency.

[0328] Each module in the above performance evaluation device can be implemented in whole or in part by software, hardware, or a combination thereof. Each module can be embedded in or independent of a processor in a computer device in the form of hardware, or can be stored in a memory in a computer device in the form of software, so that the processor can call and execute operations corresponding to each module.

[0329] In an exemplary embodiment, a computer device is provided. The computer device may be a server, and its internal structure diagram may be as shown in FIG. Fig.10 As shown. The computer device includes a processor, a memory, an input / output interface (Input / Output, referred to as I / O) and a communication interface. Among them, 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. Among them, the processor of the computer device is used 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 the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store data on the computing efficiency indicators of each data center. The input / output interface of the computer device is used to exchange information between the processor and an external device. The communication interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, a performance evaluation method is implemented.

[0330] Those skilled in the art will understand that Fig.10 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0331] In one embodiment, a computer device is further provided, including a memory and a processor, wherein a computer program is stored in the memory, and the processor implements the steps in the above method embodiments when executing the computer program.

[0332] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the steps in the above-mentioned method embodiments are implemented.

[0333] In one embodiment, a computer program product is provided, including a computer program, which implements the steps in the above method embodiments when executed by a processor.

[0334] A person of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiment method can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to the 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. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. As an illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM). The database involved in each embodiment provided in this application may include at least one of a relational database and a non-relational database. Non-relational databases may include distributed databases based on blockchains, etc., but are not limited to this. The processor involved in each embodiment provided in this application may be a general-purpose processor, a central processing unit, a graphics processor, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, an artificial intelligence (AI) processor, etc., but are not limited to this.

[0335] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this application.

[0336] The above-described embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the present application. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the attached claims.

Claims

1. A performance evaluation method, characterized in that: The method comprises: Obtain the data of computing efficiency indicators of each data center to be evaluated; Based on the data of the calculation efficiency indicator, determining the subjective and objective comprehensive weights of the calculation efficiency indicator; Obtaining a comprehensive evaluation result of each of the data centers to be evaluated through a comprehensive evaluation method according to the data of the calculation efficiency index and the subjective and objective comprehensive weights; Based on the comprehensive evaluation result, a computing efficiency shortcoming indicator is determined; the computing efficiency shortcoming indicator represents the computing efficiency indicator whose indicator performance does not meet the energy consumption requirements.

2. The method according to claim 1, characterized in that: The data of computing efficiency indicators of each data center to be evaluated is obtained, including: Based on multiple evaluation dimensions of the data center to be evaluated, data of the computing efficiency index of each of the data centers to be evaluated is obtained; the evaluation dimension represents at least one type of evaluation influencing factor of the computing efficiency of the data center.

3. The method according to claim 2, characterized in that The evaluation dimensions include first-category evaluation influencing factors, second-category evaluation influencing factors, third-category evaluation influencing factors, and fourth-category evaluation influencing factors; Among them, the first category of evaluation influencing factors corresponds to the computing power of the data center, the second category of evaluation influencing factors corresponds to the storage capacity of the data center, the third category of evaluation influencing factors corresponds to the carrying capacity of the data center, and the fourth category of evaluation influencing factors corresponds to the low-carbon power of the data center.

4. The method according to claim 2, characterized in that: Determining the subjective and objective comprehensive weights of the calculation efficiency indicators based on the data of the calculation efficiency indicators includes: Based on the subjective weighting method, the subjective weight corresponding to each of the calculation efficiency indicators is obtained; Based on the objective weighting method, the objective weight corresponding to each of the calculation efficiency indicators is obtained; The subjective weight and the objective weight are weighted and integrated to obtain the subjective and objective comprehensive weights corresponding to each of the calculation efficiency indicators.

5. The method according to claim 4, characterized in that The subjective weighting method is based on which the subjective weights corresponding to the various calculation efficiency indicators are obtained, including: Comparing the calculation efficiency indicators to obtain an indicator comparison result; Constructing a judgment matrix according to the index comparison result, and obtaining the maximum characteristic root of the judgment matrix; Based on the maximum eigenvalue of the judgment matrix, when it is determined that the judgment matrix meets the consistency requirement, the matrix eigenvalue vector of the judgment matrix is ​​solved; The matrix eigenvector of the judgment matrix is ​​normalized to obtain the subjective weight corresponding to each of the calculation efficiency indicators.

6. The method according to claim 4, characterized in that The objective weighting method is based on which the objective weights corresponding to the various calculation efficiency indicators are obtained, including: Based on the data of the calculation efficiency index, construct an original data matrix; The original data matrix is ​​dimensionally processed by forward transformation and normalization; Based on the dimensionless processed original data matrix, obtaining the indicator variation ratio of the computing efficiency indicator of each of the data centers to be evaluated; According to the indicator variation ratio, obtaining the indicator information entropy corresponding to each of the calculation efficiency indicators; Based on the indicator information entropy, the objective weight corresponding to each of the calculation efficiency indicators is obtained.

7. The method according to claim 4, characterized in that The weighted fusion of the subjective weight and the objective weight to obtain the subjective and objective comprehensive weight corresponding to each of the calculation efficiency indicators includes: Constructing a subjective weight vector and an objective weight vector according to the subjective weight and the objective weight; Based on the subjective weight vector and the objective weight vector, the subjective and objective comprehensive weights corresponding to each of the calculation effectiveness indicators are obtained through a linear weighting method.

8. The method according to claim 2, characterized in that: The comprehensive evaluation results of each of the data centers to be evaluated are obtained by a comprehensive evaluation method based on the data of the calculation efficiency index and the subjective and objective comprehensive weights, including: Correcting the data of the calculation efficiency index by adjusting the coefficient to obtain the corrected index data of each calculation efficiency index, and constructing an adjustment matrix based on the corrected index data; Using a vector normalization method to normalize the elements of the adjustment matrix to obtain a standardized matrix; Combining the subjective and objective comprehensive weights of each of the calculation efficiency indicators, weighting the standardized matrix to obtain a weighted standardized matrix; Determining positive and negative ideal solutions of the weighted normalization matrix according to the weighted normalization matrix; Obtaining the Euclidean distance corresponding to each of the to-be-evaluated data centers through the positive and negative ideal solutions; According to the Euclidean distance, the relative closeness of each of the to-be-evaluated data centers is obtained; the relative closeness is used to characterize the comprehensive evaluation result of the to-be-evaluated data center.

9. The method according to claim 8, characterized in that Determining the short board indicators of computing efficiency based on the comprehensive evaluation results includes: Based on the Euclidean distance, obtaining dimension evaluation scores of each evaluation dimension of each of the to-be-evaluated data centers, and constructing a dimension scoring matrix according to the evaluation dimension scores; Determining the best data center among the data centers to be evaluated based on the comprehensive evaluation results; According to the dimension scoring matrix, obtaining the deviation rate of the scores of each of the to-be-evaluated data centers and the optimal data center with respect to the evaluation dimension; Based on the deviation rates corresponding to the evaluation dimensions, determine the evaluation dimension with the largest deviation contribution; Constructing an indicator deviation analysis matrix based on the evaluation dimension with the largest deviation contribution and the data of the calculation efficiency indicator; Based on the indicator deviation analysis matrix and in combination with the subjective and objective comprehensive weights, the indicator deviation contribution of the calculation efficiency indicator in the evaluation dimension with the largest deviation contribution is obtained; The deviation contributions of each of the indicators are sorted to obtain the calculation efficiency short board indicator.

10. A performance evaluation device, characterized in that: The device comprises: The data acquisition module is used to obtain the data of the computing efficiency indicators of each data center to be evaluated; An indicator weight determination module, used to determine the subjective and objective comprehensive weight of the calculation efficiency indicator based on the data of the calculation efficiency indicator; A comprehensive evaluation module, used to obtain a comprehensive evaluation result of each of the data centers to be evaluated through a comprehensive evaluation method according to the data of the calculation efficiency index and the subjective and objective comprehensive weights; The shortcoming assessment module is used to determine the computing efficiency shortcoming index based on the comprehensive evaluation result; the computing efficiency shortcoming index represents the computing efficiency index whose index performance does not meet the energy consumption requirements.

11. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 9 are implemented.

12. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 9 are implemented.

13. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 9 are implemented.

Images