Host machine determination method and device, computer equipment and computer storage medium

Abstract

The invention relates to a host machine determination method and device, computer equipment and a computer storage medium, and relates to the field of cloud computation.The method comprises the steps that multiple target reputation evaluation indexes of candidate host machines are determined according to request information for establishing virtual machine instances; for each candidate host machine, determining a subjective weight of each target reputation evaluation index, and performing correction processing on the subjective weight of each target reputation evaluation index to obtain a comprehensive weight of each target reputation evaluation index; determining a comprehensive evaluation value of each candidate host machine according to each target reputation evaluation index and the comprehensive weight of each target reputation evaluation index; and determining a target host machine from the plurality of candidate host machines according to the comprehensive evaluation value of each candidate host machine. And the target host machine is selected based on the comprehensive evaluation result, so that more reasonable resource allocation can be realized. Therefore, the resources can be distributed to the most suitable host machine, the utilization rate of the resources is improved, and the resource management efficiency of the whole cloud computing system is improved.

Description

Technical Field

[0001] The present application relates to the field of cloud computing technology, and in particular to a host determination method, device, computer equipment, and computer storage medium. Background Art

[0002] With the rapid development of cloud computing and virtualization technology, cloud computing platforms have become an important tool for enterprises and organizations to manage resources and process computing tasks. Among these platforms, the choice of host directly affects the performance of the cloud host and user experience.

[0003] Traditional host selection methods mostly rely on resource utilization, such as CPU utilization, memory utilization, etc., but there are problems such as single resource utilization and unfair resource allocation. Single resource utilization leads to a lack of comprehensiveness and fairness in the selection results, and unfair resource allocation leads to some hosts being overused while other hosts are idle, resulting in low resource utilization. Summary of the invention

[0004] Based on this, it is necessary to provide a host machine determination method, device, computer equipment and computer storage medium that can solve the above-mentioned technical problems.

[0005] In a first aspect, the present application provides a host determination method, the method comprising:

[0006] Determine multiple target reputation evaluation indicators for each candidate host machine based on the request information for establishing the virtual machine instance;

[0007] For each candidate host, determine the subjective weight of each target reputation evaluation index, and modify the subjective weight of each target reputation evaluation index to obtain the comprehensive weight of each target reputation evaluation index;

[0008] Determine the comprehensive evaluation value of each candidate host machine according to each target reputation evaluation index and the comprehensive weight of each target reputation evaluation index;

[0009] A target host is determined from multiple candidate hosts according to the comprehensive evaluation value of each candidate host.

[0010] In one embodiment, the above-mentioned determination of the subjective weight of each target reputation evaluation indicator and the correction of the subjective weight of each target reputation evaluation indicator to obtain the comprehensive weight of each target reputation evaluation indicator include:

[0011] Based on the pre-built hierarchical structure model, the subjective weight calculation process is performed on each target reputation evaluation indicator to obtain the subjective weight of each target reputation evaluation indicator;

[0012] The objective weight calculation process of each target reputation evaluation index is performed using the entropy algorithm to obtain the objective weight of each target reputation evaluation index;

[0013] Each objective weight is used to modify each subjective weight to obtain the comprehensive weight of each target reputation evaluation indicator.

[0014] In one embodiment, the subjective weight calculation process is performed on each target reputation evaluation indicator based on the pre-built hierarchical structure model, and the subjective weight of each target reputation evaluation indicator is obtained, including:

[0015] Input each target reputation evaluation index into the hierarchical structure model for hierarchical decomposition processing to obtain multiple levels; the levels include target level, criterion level and index level;

[0016] For multiple target reputation evaluation indicators in each level, a pairwise comparison matrix is ​​determined; the pairwise comparison matrix is ​​used to represent the relative importance scores of every two target reputation evaluation indicators;

[0017] The eigenvector calculation and consistency check are performed on multiple element values ​​in the pairwise comparison matrix to obtain the subjective weight of each target reputation evaluation index.

[0018] In one embodiment, for the plurality of target reputation evaluation indicators in each level, determining the pairwise comparison matrix includes:

[0019] For each level, historical experience data is used to compare multiple target reputation evaluation indicators in the level in pairs to obtain the indicator score results of the level;

[0020] According to the indicator score results at each level, a pairwise comparison matrix is ​​constructed.

[0021] In one embodiment, the objective weight calculation process of each target reputation evaluation index using the entropy algorithm is performed to obtain the objective weight of each target reputation evaluation index, including:

[0022] Obtain sample data of each target reputation evaluation indicator;

[0023] Performing data standardization processing on sample data of each target reputation evaluation indicator to obtain standardized data of multiple target reputation evaluation indicators;

[0024] Determine the entropy value of each target reputation evaluation indicator according to the standardized data of each target reputation evaluation indicator;

[0025] Determine the credit utility value of each target credit evaluation indicator according to the entropy value of each target credit evaluation indicator;

[0026] Each credit utility value is normalized to obtain the objective weight of each target credit evaluation indicator.

[0027] In one embodiment, the standardized data of the target reputation evaluation index includes sample evaluation indicators of multiple samples; determining the entropy value of each target reputation evaluation indicator according to the standardized data of each target reputation evaluation indicator includes:

[0028] For each sample evaluation indicator, determine the proportion of the sample evaluation indicator in multiple sample evaluation indicators;

[0029] According to the weight of each sample evaluation index and the number of samples obtained in advance, the entropy value of each target reputation evaluation index is determined.

[0030] In one embodiment, determining the target host according to the comprehensive evaluation value of each candidate host includes:

[0031] The comprehensive evaluation values ​​of the candidate host machines are sorted, and the candidate host machine with the highest comprehensive evaluation value is determined as the target host machine.

[0032] In a second aspect, the present application also provides a host determination device. The device includes:

[0033] An indicator determination module, used to determine multiple target reputation evaluation indicators of each candidate host machine according to the request information for establishing a virtual machine instance;

[0034] A comprehensive weight determination module is used to determine the subjective weight of each target reputation evaluation index for each candidate host machine, and to modify the subjective weight of each target reputation evaluation index to obtain the comprehensive weight of each target reputation evaluation index;

[0035] An evaluation value determination module, used to determine the comprehensive evaluation value of each candidate host machine according to each target reputation evaluation index and the comprehensive weight of each target reputation evaluation index;

[0036] The host machine determination module is used to determine a target host machine from a plurality of candidate host machines according to the comprehensive evaluation value of each candidate host machine.

[0037] In a third aspect, the present application further provides a computer device, which includes a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps of the method in the first aspect when executing the computer program.

[0038] In a fourth aspect, the present application further provides a computer-readable storage medium having a computer program stored thereon, which implements the steps of the method in the first aspect when executed by a processor.

[0039] In a fifth aspect, the present application further provides a computer program product, which includes a computer program, and when the computer program is executed by a processor, the steps of the method in the first aspect are implemented.

[0040] The host determination method, device, computer equipment and computer storage medium described above first determine multiple target reputation evaluation indicators for each candidate host according to the request information for establishing a virtual machine instance; for each candidate host, determine the subjective weight of each target reputation evaluation indicator, and modify the subjective weight of each target reputation evaluation indicator to obtain the comprehensive weight of each target reputation evaluation indicator; determine the comprehensive evaluation value of each candidate host according to each target reputation evaluation indicator and the comprehensive weight of each target reputation evaluation indicator; finally, determine the target host from multiple candidate hosts according to the comprehensive evaluation value of each candidate host. Selecting the target host based on the comprehensive evaluation results helps to achieve more reasonable resource allocation. Ensure that resources can be allocated to the most suitable host, improve resource utilization, reduce resource waste, and thus improve the resource management efficiency of the entire cloud computing system. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 A diagram of an application environment of a host machine determination method in one embodiment;

[0042] Figure 2 is a schematic flow chart of a method for determining a host machine in one embodiment;

[0043] Figure 3 A schematic diagram of a process for obtaining the comprehensive weight of each target reputation evaluation index in an embodiment;

[0044] Figure 4 A schematic diagram of a process for obtaining the subjective weight of each target reputation evaluation index in an embodiment;

[0045] Figure 5 A schematic diagram of a process for determining a pairwise comparison matrix in one embodiment;

[0046] Figure 6 A schematic diagram of a process for obtaining objective weights of each target reputation evaluation index in one embodiment;

[0047] Figure 7 A schematic diagram of a process for determining the entropy value of each target reputation evaluation index in one embodiment;

[0048] Figure 8 is a structural block diagram of a host determination device in one embodiment;

[0049] Fig. 9 FIG. 4 is a diagram showing the internal structure of a computer device in one embodiment. DETAILED DESCRIPTION

[0050] 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.

[0051] First of all, before specifically introducing the technical solution of the embodiments of the present application, the technical background based on the embodiments of the present application is introduced.

[0052] With the rapid development of cloud computing and virtualization technology, cloud computing platforms such as OpenStack have become important tools for enterprises and organizations to manage resources and process computing tasks. In these platforms, the choice of host directly affects the performance and user experience of cloud hosts. Traditional host selection methods mostly rely on resource utilization (such as CPU usage, memory usage, etc.), but these methods have the following problems: ① Single resource utilization: Traditional selection methods mainly focus on the resource usage of the host, ignoring factors such as the host's historical performance and user feedback, resulting in a lack of comprehensiveness and fairness in the selection results; ② Lack of reputation evaluation mechanism: The existing methods lack a comprehensive evaluation of the host's reputation, resulting in the possibility of ignoring the reliability and stability of the host during the selection process, thereby affecting the overall system performance and user satisfaction; ③ Unfair resource allocation: Due to the lack of reasonable consideration of comprehensive factors such as the host's historical record and user feedback, the resource allocation process may be unfair, some hosts may be overused, while other hosts are idle, and resource utilization is low.

[0053] In cloud computing platforms, the selection of hosts directly affects the performance and user experience of cloud hosts. Existing host selection methods mostly rely on resource utilization, ignoring the comprehensive evaluation of host reputation, resulting in unfair resource allocation. Resource utilization methods mainly focus on the current resource usage of hosts, such as CPU, memory, disk space, etc., but due to ignoring historical performance and user feedback, they are prone to unstable and unfair selection results. On the contrary, historical reputation evaluation methods can more comprehensively reflect the performance and reliability of hosts by comprehensively analyzing factors such as historical usage records and user evaluations of hosts, but relying solely on historical data may not reflect the current resource status in a timely manner. To overcome these limitations, this study proposes a host selection method that combines subjective and objective evaluation. The practicality of this method lies in that it can comprehensively analyze and quantify the multi-dimensional performance indicators of hosts, such as processing power, memory usage, response time, and historical stability, to ensure that the host that best suits current needs is selected. This precise host selection process not only optimizes resource allocation and reduces resource waste, but also improves the response speed and reliability of cloud services, ultimately providing users with a more stable and efficient cloud computing experience.

[0054] Based on this, the present application provides a host determination method, apparatus, computer device and computer storage medium, aiming to solve the above technical problems.

[0055] The host determination method provided in the embodiment of the present application can be applied to Figure 1 In the application environment shown. Among them, the terminal 102 communicates with the server 104 through the network. The data storage system can store the data that the server 104 needs to process. The data storage system can be integrated on the server 104, or it can be placed on the cloud or other network servers. The server 104 obtains the request information for establishing a virtual machine instance sent by the terminal 102, and determines multiple target reputation evaluation indicators of each candidate host according to the request information; for each candidate host, the subjective weight of each target reputation evaluation indicator is determined, and the subjective weight of each target reputation evaluation indicator is corrected to obtain the comprehensive weight of each target reputation evaluation indicator; according to each target reputation evaluation indicator and the comprehensive weight of each target reputation evaluation indicator, the comprehensive evaluation value of each candidate host is determined; finally, according to the comprehensive evaluation value of each candidate host, the target host is determined from multiple candidate hosts. Among them, the terminal 102 can be, but is not limited to, various personal computers, laptops, smart phones and tablet computers. The server 104 can be implemented with an independent server or a server cluster consisting of multiple servers.

[0056] In an exemplary embodiment, Figure 2 As shown, a host determination method is provided, which is applied to Figure 1 The server in the example is used to illustrate the following steps:

[0057] S201, determining multiple target reputation evaluation indicators of each candidate host machine according to request information for establishing a virtual machine instance.

[0058] Among them, the candidate host refers to a set of potential hosts in the cloud computing environment that can meet the basic conditions for establishing a virtual machine instance (such as hardware configuration, etc.), from which the most suitable host needs to be selected to create a virtual machine instance.

[0059] The target reputation evaluation index is used to evaluate the key indicators of the reputation and performance of the candidate host, including but not limited to the number of downtimes, recovery time, communication time, number of CPUs, memory size, storage space, and network configuration. These indicators comprehensively reflect the various characteristics of the host and are an important basis for selecting the target host.

[0060] In an embodiment of the present application, the server receives a request message for establishing a virtual machine instance sent by a terminal, and the message includes the specification requirements of the virtual machine, such as the required number of CPU cores, memory size, storage space capacity, and network bandwidth. Based on these specification requirements and preset screening rules, candidate hosts that meet the basic conditions are screened out from all available hosts. For each candidate host, multiple target reputation evaluation indicators are determined for evaluating its reputation and performance, such as the number of downtimes, recovery time, communication time, number of CPUs, memory size, storage space, and network configuration. These indicators can be obtained from the system log, performance monitoring data, and related configuration information of the host.

[0061] Another implementation method: In addition to directly obtaining the specification requirements of the virtual machine from the client request information, the overall load of the current cloud computing system and the resource distribution information can also be combined to dynamically adjust the screening conditions of the candidate host machines. For example, if the resource utilization rate of some areas in the system is high, while that of other areas is low, the candidate host machines from the areas with low resource utilization can be selected first.

[0062] When determining the target reputation evaluation indicators, in addition to the common performance and reliability indicators, some specific indicators can be added according to the needs of the actual application scenarios, such as the host machine's security indicators (whether the latest security patches are installed, whether there are security vulnerabilities, etc.), energy-saving indicators (energy consumption efficiency, etc.), etc.

[0063] S202: For each candidate host, determine the subjective weight of each target reputation evaluation indicator, and modify the subjective weight of each target reputation evaluation indicator to obtain the comprehensive weight of each target reputation evaluation indicator.

[0064] Among them, the subjective weight reflects the subjective importance of different indicators in evaluating the host in the decision-making process.

[0065] In an embodiment of the present application, the server can use the analytic hierarchy process (AHP) to construct a hierarchical model of host evaluation indicators, and divide the target reputation evaluation indicators into different levels, such as the target layer (selecting the optimal host), the criterion layer (performance, reliability, etc.) and the indicator layer (specific target reputation evaluation indicators).

[0066] Experts in related fields were invited to compare indicators at the same level in pairs, and scores were given according to the 1-9 scale to construct a pairwise comparison matrix. The eigenvector and maximum eigenvalue of the pairwise comparison matrix were calculated to obtain the subjective weight of each target reputation evaluation indicator. The entropy method was used to process the data of each target reputation evaluation indicator and calculate the objective weight of each indicator.

[0067] Set an adjustment coefficient (such as 0.5), add the subjective weight and the objective weight, and get the comprehensive weight of each target reputation evaluation indicator. That is: comprehensive weight = adjustment coefficient × subjective weight + (1-adjustment coefficient) × objective weight.

[0068] Another implementation method: The server can also use group decision-making to determine the subjective weights, invite multiple experts to form an expert group, each expert compares and scores the indicators in pairs, and then combines the scoring results of multiple experts through a certain method (such as weighted average method) to obtain the final subjective weight. When using the entropy method to calculate the objective weight, the data can be preprocessed to remove outliers or use data smoothing technology to improve the accuracy of the objective weight calculation. In addition to the simple weighted summation method, other more complex combination methods can be tried to determine the comprehensive weight, such as comprehensive evaluation methods based on fuzzy mathematics, neural network methods, etc., to better integrate subjective and objective weights.

[0069] S203: Determine the comprehensive evaluation value of each candidate host machine according to each target reputation evaluation index and the comprehensive weight of each target reputation evaluation index.

[0070] In the embodiment of the present application, for each candidate host, its actual value on each target reputation evaluation index is obtained, and the actual value of each target reputation evaluation index is standardized, and data of different dimensions are converted into dimensionless values ​​for unified comparison and calculation.

[0071] Another implementation method: When standardizing the actual values, nonlinear standardization methods such as logarithmic transformation, exponential transformation, etc. can be used to better reflect the distribution characteristics of the indicator data.

[0072] The time decay factor is introduced to adjust the weight of historical data. For example, the indicator data of the recent time period is given a higher weight, and the weight of the data of the distant time period is gradually reduced to reflect the dynamic changes of the host performance and reputation.

[0073] S204: Determine a target host from a plurality of candidate host machines according to the comprehensive evaluation value of each candidate host machine.

[0074] In the embodiment of the present application, all candidate hosts are sorted from high to low according to the comprehensive evaluation value, and the candidate host with the highest comprehensive evaluation value is selected as the target host for creating a virtual machine instance.

[0075] Another implementation method is to set a threshold of the comprehensive evaluation value, and only when the comprehensive evaluation value of the candidate host is higher than the threshold, it is considered as a selectable host. If the comprehensive evaluation values ​​of all candidate hosts are lower than the threshold, feedback can be given to the user to prompt the user to adjust the specification requirements of the virtual machine instance or wait for the optimization adjustment of system resources.

[0076] In addition to selecting the host with the highest comprehensive evaluation value, you can also consider selecting the top N hosts (N is an integer greater than 1) in the comprehensive evaluation value as the candidate set, and then further screen them based on other factors (such as cost, geographical location, etc.) to finally determine the target host.

[0077] The host determination method described above first determines multiple target reputation evaluation indicators for each candidate host according to the request information for establishing a virtual machine instance; for each candidate host, determines the subjective weight of each target reputation evaluation indicator, and corrects the subjective weight of each target reputation evaluation indicator to obtain the comprehensive weight of each target reputation evaluation indicator; determines the comprehensive evaluation value of each candidate host according to each target reputation evaluation indicator and the comprehensive weight of each target reputation evaluation indicator; and finally determines the target host from multiple candidate hosts according to the comprehensive evaluation value of each candidate host. Selecting the target host based on the comprehensive evaluation results helps to achieve more reasonable resource allocation. It ensures that resources can be allocated to the most suitable host, improves resource utilization, reduces resource waste, and thus improves the resource management efficiency of the entire cloud computing system.

[0078] In an exemplary embodiment, based on the above embodiment, see Figure 3 The embodiment of the present application relates to a process of determining the subjective weight of each target reputation evaluation indicator for each candidate host machine, and correcting the subjective weight of each target reputation evaluation indicator to obtain the comprehensive weight of each target reputation evaluation indicator, including the following steps:

[0079] S301 , performing subjective weight calculation processing on each target reputation evaluation indicator based on a pre-built hierarchical structure model to obtain the subjective weight of each target reputation evaluation indicator.

[0080] Among them, the hierarchical model is a structural model that decomposes complex problems into multiple levels, which is used to analyze and solve problems. In the embodiment of the present application, the host selection problem is divided into the target layer (selecting the optimal host), the criterion layer (such as performance, reliability, etc.) and the indicator layer (specific target reputation evaluation indicators). The relationship between each level and the relative importance of each indicator can be clearly expressed through the hierarchical model.

[0081] In the embodiment of the present application, first, the elements of each level in the hierarchical model and their interrelationships are clarified. For example, performance, reliability, resource utilization and other aspects are determined as the main criteria at the criterion level, and specific target reputation evaluation indicators are determined at the indicator level, such as the number of CPUs, memory size, number of downtimes, recovery time, etc.

[0082] Experts in related fields were invited to compare the elements of the same level in pairs, and a pairwise comparison matrix was constructed based on the comparison results. The comparison was made using a 1-9 scale. If one element is considered slightly more important than another, it is scored as 3; obviously important as 5; strongly important as 7; extremely important as 9; if two elements are equally important, they are scored as 1. Conversely, if the importance of A to B is 3, then the importance of B to A is 1 / 3.

[0083] By calculating the eigenvector and maximum eigenvalue of the pairwise comparison matrix, the subjective weight of each target reputation evaluation index is obtained. The specific calculation method can use mathematical methods such as the eigenvalue method. The calculated eigenvector is the subjective weight of each index after normalization.

[0084] S302, using an entropy algorithm to perform objective weight calculation processing on each target reputation evaluation indicator to obtain an objective weight of each target reputation evaluation indicator.

[0085] Among them, the entropy algorithm is an objective weighting algorithm based on the principle of information entropy, which is used to calculate the objective weight of each indicator. It determines the weight of the indicator by analyzing the discrete degree of the indicator data. The greater the discrete degree of the data, the higher the weight of the indicator, and vice versa.

[0086] In the embodiment of the present application, the actual data of each candidate host on each target reputation evaluation index can be collected. The data is standardized to eliminate the influence of the dimension. The entropy value of each index is calculated using an entropy value algorithm, the information utility value of each index is calculated, and finally the information utility value is normalized to obtain the objective weight of each target reputation evaluation index.

[0087] Another implementation method: In the process of data standardization, the distribution characteristics of the data are considered and an adaptive standardization method is used. For example, for indicators with uneven data distribution, the quantile standardization method is used to map the data to a specific quantile interval to better reflect the actual situation of the data. Time weights are introduced to assign different weights to data at different time points. Recent data has a higher weight, reflecting the current performance and reputation of the host; long-term data has a lower weight, reflecting the reference value of historical data. In this way, the objective weight can be made more adaptable to the dynamic changes in the performance and reputation of the host.

[0088] S303: using each objective weight to modify each subjective weight, to obtain a comprehensive weight of each target reputation evaluation index.

[0089] In the embodiment of the present application, the AHP-entropy combined weighting method is adopted to balance subjective judgment and objective data analysis to ensure the scientificity and comprehensiveness of the host selection process. As a subjective weighting method, the analytic hierarchy process (AHP) allows decision makers to evaluate the importance of different host performance indicators based on experience and knowledge, thereby forming a weight system. However, relying solely on subjective evaluation may lead to deviations in weight distribution. In order to make up for this deficiency, we introduce the entropy method, which is an objective weighting method that determines the weight by analyzing the degree of dispersion of each indicator in the actual data, thereby reducing the influence of subjective bias. In implementation, the AHP method is first used to construct a hierarchical model of host evaluation indicators, and a pairwise comparison matrix is ​​formed through expert scoring, and the subjective weight and consistency index of each indicator are further calculated.

[0090] Subsequently, the entropy method was used to analyze the same data set and calculate the objective weight of each indicator. Finally, through the appropriate adjustment coefficient (usually 0.5), the subjective weight obtained by the AHP method and the objective weight obtained by the entropy method were organically combined to form the final comprehensive weight. This composite strategy that combines subjective and objective weighting methods ensures the rationality and scientificity of the evaluation system, so that the host selection not only relies on the judgment of experts, but also takes into account the data support in actual operations, thereby improving the accuracy and reliability of the host selection process.

[0091] The objective weight calculated by the entropy method is used to adjust the subjective weight of the AHP method. The calculation formula is as described in formula (1):

[0092] .........................(1)

[0093] Among them, β is the adjustment coefficient.

[0094] Another implementation method: a dynamic adjustment coefficient can be used. The coefficient value is dynamically adjusted according to the system operation status and the analysis results of historical data. For example, when the system is initially running or the data is unstable, the proportion of subjective weight is appropriately increased, that is, the value is larger; as the system operation time increases and data accumulates, the proportion of objective weight is gradually increased, that is, the value is gradually reduced. A secondary correction is made to the comprehensive weight. After obtaining the preliminary comprehensive weight, the weights of certain key indicators are fine-tuned according to the needs of the actual application scenario. For example, if the stability of CPU performance is more important for a specific application, the comprehensive weight of CPU-related indicators can be appropriately increased.

[0095] In the embodiment of the present application, by combining the subjective weight determined by the hierarchical model and the objective weight calculated by the entropy algorithm, the importance of each target reputation evaluation indicator in the host selection can be comprehensively and objectively reflected, avoiding the limitations of a single weight determination method.

[0096] In an exemplary embodiment, based on the above embodiment, see Figure 4 The embodiment of the present application relates to a process of calculating and processing the subjective weight of each target reputation evaluation indicator based on a pre-built hierarchical structure model to obtain the subjective weight of each target reputation evaluation indicator, including the following steps:

[0097] S401, input each target reputation evaluation index into the hierarchical structure model for hierarchical decomposition processing to obtain multiple levels, wherein the levels include a target level, a criterion level and an index level.

[0098] In an embodiment of the present application, the target of host selection is clearly defined and used as the target layer. For example, the target is to select a host with the best performance and the highest reliability. The elements of the criterion layer are determined based on the characteristics and influencing factors of the host. Common criterion layer elements may include performance, reliability, resource utilization, security, etc. The elements of the criterion layer are further refined into specific target reputation evaluation indicators to form an indicator layer. For example, performance indicators may include the number of CPUs, memory size, communication time, etc.; reliability indicators may include the number of downtimes, recovery time, etc.

[0099] S402, determining a pairwise comparison matrix for multiple target reputation evaluation indicators in each level.

[0100] Among them, the pairwise comparison matrix is ​​used to characterize the relative importance scores of each two target reputation evaluation indicators.

[0101] In the embodiment of the present application, experts in related fields can be invited to compare the target reputation evaluation indicators of the same level in pairs. For example, for the performance and reliability indicators of the criterion layer, experts judge the importance of performance relative to reliability based on experience and actual conditions, and score them according to the 1-9 scale. If performance is considered slightly more important than reliability, it can be scored as 3; obviously important as 5; very important as 7; extremely important as 9; equally important as 1. Conversely, if the importance of reliability to performance is 3, the importance of performance relative to reliability is 1 / 3.

[0102] The scoring results of all experts are summarized and sorted to construct a pairwise comparison matrix. The rows and columns of the matrix correspond to different target reputation evaluation indicators, and the elements in the matrix are the scores of the pairwise comparison of the corresponding indicators.

[0103] S403, performing feature vector calculation and consistency check processing on multiple element values ​​in the pairwise comparison matrix to obtain the subjective weight of each target reputation evaluation index.

[0104] In the embodiment of the present application, a mathematical method (such as the eigenvalue method) can be used to calculate the eigenvector of the pairwise comparison matrix. After the eigenvector is obtained by calculation, it is normalized so that the sum of each element is 1, and the preliminary subjective weight of each target reputation evaluation index is obtained.

[0105] In order to ensure the logical consistency of the pairwise comparison matrix and the reliability of the expert scoring, a consistency check is required. The consistency ratio CR is calculated as shown in formula (2):

[0106] ..........................................................(2)

[0107] Among them, RI (Random Consistency Index) is a value predefined according to the number of indicators and can be found in the standard table, while CI (Consistency Index) is shown in formula (3):

[0108] ....................................................(3)

[0109] Among them, λmax is the maximum eigenvalue of the pairwise comparison matrix, n is the number of evaluation indicators, and RI (Random Consistency Index) is a value predefined according to the number of indicators.

[0110] If the CR value is less than 0.1, the matrix is ​​considered to have good consistency; if it is greater than 0.1, the pairwise comparison matrix needs to be readjusted until the consistency requirements are met.

[0111] The embodiment of the present application decomposes the complex host selection problem through a hierarchical model, and clearly shows the hierarchical relationship between the target reputation evaluation indicators. The subjective weights are determined by using the paired comparison matrix and eigenvector calculation, which fully considers the experience and knowledge of experts, making the determination of the weights more scientific and reasonable.

[0112] In an exemplary embodiment, based on the above embodiment, see Figure 5 The embodiment of the present application relates to a process of determining a pairwise comparison matrix for multiple target reputation evaluation indicators in each level, including the following steps:

[0113] S501, for each level, use historical experience data to perform pairwise comparison processing on multiple target reputation evaluation indicators in the level to obtain the indicator score results of the level.

[0114] Among them, historical experience data refers to the data accumulated in the past in host selection or related fields. These data contain information about host performance, reputation, etc., and can be used as a reference for determining the pairwise comparison matrix.

[0115] The indicator score result is obtained by comparing multiple target reputation evaluation indicators at the same level in pairs, and is used to reflect the relative importance of each indicator.

[0116] In an embodiment of the present application, first, historical experience data related to host machine selection is collected and organized, including but not limited to performance indicator data of the host machine (such as CPU utilization, memory utilization, etc.), reliability indicator data (such as number of downtimes, recovery time, etc.) and other relevant data.

[0117] For each level, multiple target reputation evaluation indicators are combined in pairs. For example, in the criterion level, performance indicators are compared with reliability indicators, performance indicators are compared with resource utilization indicators, etc.

[0118] Based on historical experience data, analyze the relative importance of each pair of indicators. Statistical analysis methods, such as calculating the correlation coefficient between indicators, comparing the mean and variance of indicators, etc., can be used to determine the relative importance of two indicators. For example, if historical data shows that CPU usage has a greater impact on host performance, while memory usage has a relatively small impact, then when performing a pairwise comparison within the performance indicators, a higher score can be given to the comparison of CPU usage relative to memory usage.

[0119] According to the above analysis results, a score is assigned to each pair of indicators to obtain the hierarchical indicator score results. The value range of the score can be set according to the actual situation.

[0120] S502: construct a pairwise comparison matrix based on the indicator score results at each level.

[0121] In the embodiment of the present application, a pairwise comparison matrix is ​​constructed based on the index score results of each level. The rows and columns of the matrix correspond to each target reputation evaluation index in the same level.

[0122] For each element a in the matrix ij , whose value is the comparison score between index i and index j. Since the pairwise comparison matrix is ​​symmetrical, that is, a ij =1 / a ji, so we only need to determine the element values ​​of the upper or lower triangular part of the matrix.

[0123] In the embodiment of the present application, by using historical experience data to determine the pairwise comparison matrix, past practical experience and data information can be fully utilized to make the construction of the matrix more objective and accurate and reflect the actual situation.

[0124] In an exemplary embodiment, based on the above embodiment, see Figure 6 The embodiment of the present application relates to a process of using an entropy algorithm to perform objective weight calculation processing on each target reputation evaluation indicator to obtain the objective weight of each target reputation evaluation indicator, including the following steps:

[0125] S601, obtaining sample data of each target reputation evaluation index.

[0126] The sample data refers to the actual data collected from each target reputation evaluation indicator within a certain time range or on a certain number of host machines, and these data are used for subsequent weight calculations.

[0127] In the embodiment of the present application, the actual data of each candidate host machine on each target reputation evaluation index can be collected from the monitoring module or related database of the cloud computing system. For example, for the CPU quantity index, the number of CPU cores of each host machine is collected; for the downtime index, the number of downtimes of each host machine within a certain period of time is recorded.

[0128] Ensure the accuracy and completeness of sample data and handle missing or abnormal data. Data problems can be handled by methods such as data filling, data smoothing, or deleting outliers.

[0129] S602, performing data standardization processing on sample data of each target reputation evaluation index to obtain standardized data of multiple target reputation evaluation indicators.

[0130] In the embodiment of the present application, the data standardization methods include range standardization and Z-score standardization, etc. The sample data of each target reputation evaluation index is processed according to the corresponding standardization method to obtain a standardized data set.

[0131] S603: Determine the entropy value of each target reputation evaluation indicator according to the standardized data of each target reputation evaluation indicator by using an entropy value algorithm.

[0132] In the embodiment of the present application, first, the proportion of the i-th sample under each indicator is calculated, and then the entropy value of each indicator is calculated.

[0133] Alternatively, when calculating the entropy value, consider the correlation of the data. For indicators with correlation, methods such as principal component analysis can be used to reduce the dimension, remove redundant information, and then calculate the entropy value to improve the accuracy of entropy calculation. Introduce a dynamic entropy calculation method. According to the operating status of the system and the changes in the data, dynamically adjust the calculation parameters of the entropy value so that the entropy value can reflect the changes in the importance of the indicator more timely.

[0134] S604: Determine the credit utility value of each target reputation evaluation indicator according to the entropy value of each target reputation evaluation indicator.

[0135] Among them, the credit utility value is an indicator calculated based on the entropy value, which is used to indicate the utility of each target credit evaluation indicator in the evaluation system. The larger the credit utility value, the greater the contribution of the indicator to the evaluation result.

[0136] In the embodiment of the present application, the credit utility value of each target reputation evaluation index is calculated as shown in formula (4): .....................................................(4)

[0137] in, is the credit utility value, is the entropy value of the target reputation evaluation index.

[0138] S605, normalizing each credit utility value to obtain an objective weight of each target credit evaluation index.

[0139] In the embodiment of the present application, the information utility value is normalized to obtain the objective weight of each indicator, as shown in formula (5):

[0140] ...............................................(5)

[0141] in, is the objective weight, and m is the number of indicators.

[0142] In the embodiment of the present application, an entropy algorithm is used to calculate the objective weight of each target reputation evaluation index based on sample data, thereby avoiding subjective interference from human factors and making the determination of the weight more objective and accurate.

[0143] In an exemplary embodiment, based on the above embodiment, see Figure 7The standardized data of the target reputation evaluation index of the embodiment of the present application includes sample evaluation indexes of multiple samples; the embodiment of the present application relates to a process of determining the entropy value of each target reputation evaluation index according to the standardized data of each target reputation evaluation index, including the following steps:

[0144] S701, for each sample evaluation indicator, determining the proportion of the sample evaluation indicator in multiple sample evaluation indicators.

[0145] In the embodiment of the present application, the proportion of the i-th sample under each indicator is calculated as shown in formula (6):

[0146] .................................................(6)

[0147] S702: Determine the entropy value of each target reputation evaluation index according to the weight of each sample evaluation index and the number of samples obtained in advance.

[0148] In the embodiment of the present application, the entropy value of each indicator is calculated as shown in formulas (7) and (8):

[0149] .........................................(7)

[0150] ....................................................(8)

[0151] in, is the proportion of the same target reputation evaluation index of the sample, is the sample size.

[0152] The embodiment of the present application can accurately reflect the discrete degree and information amount of the indicator data by calculating the entropy value of each target reputation evaluation indicator, and provide an objective basis for the subsequent determination of the importance of the indicator.

[0153] In an exemplary embodiment, based on the above embodiments, the embodiments of the present application involve a process of determining a target host machine based on a comprehensive evaluation value of each candidate host machine, including: sorting the comprehensive evaluation values ​​of each candidate host machine, and determining the candidate host machine with the highest comprehensive evaluation value as the target host machine.

[0154] The embodiment of the present application can use common sorting algorithms, such as bubble sort, quick sort, etc., to sort the comprehensive evaluation values ​​of each candidate host machine. Taking bubble sort as an example, starting from the first candidate host machine, the comprehensive evaluation values ​​of two adjacent host machines are compared in turn. If the comprehensive evaluation value of the previous host machine is greater than the comprehensive evaluation value of the next host machine, their positions are swapped. Repeat this process until all candidate host machines are arranged in order of comprehensive evaluation values ​​from small to large.

[0155] In the sorted list of candidate hosts, select the first host, that is, the candidate host with the highest comprehensive evaluation value, and determine it as the target host. Record relevant information of the target host, such as the host's ID, IP address, etc., to facilitate the subsequent creation and deployment of virtual machine instances.

[0156] The embodiment of the present application can quickly and accurately find the host machine with the best comprehensive performance by sorting the comprehensive evaluation values, thereby improving the efficiency of host machine selection.

[0157] In an exemplary embodiment, based on the above embodiment, the method of the present application further includes the following steps:

[0158] Step 1: Determine multiple target reputation evaluation indicators for each candidate host machine based on the request information for establishing a virtual machine instance;

[0159] Step 2: For each candidate host, input each target reputation evaluation index into the hierarchical structure model for hierarchical decomposition to obtain multiple levels; for each level, use historical experience data to perform pairwise comparisons on multiple target reputation evaluation indicators in the level to obtain the indicator score results of the level; construct a pairwise comparison matrix based on the indicator score results of each level; perform eigenvector calculation and consistency check on multiple element values ​​in the pairwise comparison matrix to obtain the subjective weight of each target reputation evaluation indicator;

[0160] Step 3: Obtain sample data of each target reputation evaluation indicator; perform data standardization on the sample data of each target reputation evaluation indicator to obtain standardized data of multiple target reputation evaluation indicators; for each sample evaluation indicator, determine the proportion of the sample evaluation indicator in the multiple sample evaluation indicators; determine the entropy value of each target reputation evaluation indicator according to the proportion of each sample evaluation indicator and the number of samples obtained in advance;

[0161] Step 4: Determine the credit utility value of each target credit evaluation indicator according to the entropy value of each target credit evaluation indicator; normalize each credit utility value to obtain the objective weight of each target credit evaluation indicator; use each objective weight to correct each subjective weight to obtain the comprehensive weight of each target credit evaluation indicator;

[0162] Step 5: Determine the comprehensive evaluation value of each candidate host machine according to each target reputation evaluation index and the comprehensive weight of each target reputation evaluation index; sort the comprehensive evaluation values ​​of each candidate host machine, and determine the candidate host machine with the highest comprehensive evaluation value as the target host machine.

[0163] 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.

[0164] Based on the same inventive concept, the embodiment of the present application also provides a host determination device for implementing the host determination method involved above. The implementation scheme for solving the problem provided by the device is similar to the implementation scheme recorded in the above method, so the specific limitations in one or more host determination device embodiments provided below can refer to the limitations of the host determination method above, and will not be repeated here.

[0165] In one embodiment, Figure 8 As shown, a host determination device 800 is provided, including:

[0166] The indicator determination module 801 is used to determine multiple target reputation evaluation indicators of each candidate host machine according to the request information for establishing a virtual machine instance;

[0167] The comprehensive weight determination module 802 is used to determine the subjective weight of each target reputation evaluation index for each candidate host, and to modify the subjective weight of each target reputation evaluation index to obtain the comprehensive weight of each target reputation evaluation index;

[0168] Evaluation value determination module 803, used to determine the comprehensive evaluation value of each candidate host machine according to each target reputation evaluation index and the comprehensive weight of each target reputation evaluation index;

[0169] The host machine determination module 804 is used to determine a target host machine from a plurality of candidate host machines according to the comprehensive evaluation value of each candidate host machine.

[0170] In one embodiment, the comprehensive weight determination module includes:

[0171] A subjective weight determination unit, used to perform subjective weight calculation processing on each target reputation evaluation indicator based on a pre-built hierarchical structure model to obtain the subjective weight of each target reputation evaluation indicator;

[0172] An objective weight determination unit, used to perform objective weight calculation processing on each target reputation evaluation index using an entropy algorithm to obtain an objective weight of each target reputation evaluation index;

[0173] The weight correction unit is used to correct each subjective weight using each objective weight to obtain the comprehensive weight of each target reputation evaluation index.

[0174] In one embodiment, the subjective weight determination unit includes:

[0175] The hierarchical subunit is used to input each target reputation evaluation index into the hierarchical structure model for hierarchical decomposition processing to obtain multiple levels; the levels include the target level, the criterion level and the indicator level;

[0176] The matrix determination subunit is used to determine a pairwise comparison matrix for a plurality of target reputation evaluation indicators in each level; the pairwise comparison matrix is ​​used to represent the relative importance score of each two target reputation evaluation indicators;

[0177] The subjective weight subunit is used to perform eigenvector calculation and consistency check on multiple element values ​​in the pairwise comparison matrix to obtain the subjective weight of each target reputation evaluation indicator.

[0178] In one embodiment, the matrix determination subunit is specifically used for:

[0179] For each level, historical experience data is used to compare multiple target reputation evaluation indicators in the level in pairs to obtain the indicator score results of the level;

[0180] According to the indicator score results at each level, a pairwise comparison matrix is ​​constructed.

[0181] In one embodiment, the objective weight determination unit includes:

[0182] A sample determination subunit is used to obtain sample data of each target reputation evaluation indicator;

[0183] A data processing subunit is used to perform data standardization processing on the sample data of each target reputation evaluation index to obtain standardized data of multiple target reputation evaluation indicators;

[0184] A utility value determination subunit, used to determine the entropy value of each target reputation evaluation indicator according to the standardized data of each target reputation evaluation indicator by using an entropy value algorithm, and determine the credit utility value of each target reputation evaluation indicator according to the entropy value of each target reputation evaluation indicator;

[0185] The objective weight determination subunit is used to normalize each credit utility value to obtain the objective weight of each target credit evaluation indicator.

[0186] In one embodiment, the utility value determination subunit is specifically used for:

[0187] For each sample evaluation indicator, determine the proportion of the sample evaluation indicator in multiple sample evaluation indicators;

[0188] According to the weight of each sample evaluation index and the number of samples obtained in advance, the entropy value of each target reputation evaluation index is determined.

[0189] In one embodiment, the host determination module includes:

[0190] The host machine determination unit is used to sort the comprehensive evaluation values ​​of the candidate host machines and determine the candidate host machine with the highest comprehensive evaluation value as the target host machine.

[0191] Each module in the host determination 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.

[0192] In one embodiment, a computer device is provided. The computer device may be a server, and its internal structure diagram may be as follows: Fig. 9 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. The processor, the memory and the input / output interface are connected through a system bus, and the communication interface is connected to the system bus through the input / output interface. The processor of the computer device is 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 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 host determination method is implemented.

[0193] Those skilled in the art will understand that Fig. 9 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.

[0194] In one embodiment, a computer device is provided, including a memory and a processor, wherein a computer program is stored in the memory, and when the processor executes the computer program, the following steps are implemented:

[0195] Determine multiple target reputation evaluation indicators for each candidate host machine based on the request information for establishing the virtual machine instance;

[0196] For each candidate host, determine the subjective weight of each target reputation evaluation index, and modify the subjective weight of each target reputation evaluation index to obtain the comprehensive weight of each target reputation evaluation index;

[0197] Determine the comprehensive evaluation value of each candidate host machine according to each target reputation evaluation index and the comprehensive weight of each target reputation evaluation index;

[0198] A target host is determined from multiple candidate hosts according to the comprehensive evaluation value of each candidate host.

[0199] In one embodiment, when the processor executes the computer program, the processor further implements the following steps:

[0200] Based on the pre-built hierarchical structure model, the subjective weight calculation process is performed on each target reputation evaluation indicator to obtain the subjective weight of each target reputation evaluation indicator;

[0201] The objective weight calculation process of each target reputation evaluation index is performed using the entropy algorithm to obtain the objective weight of each target reputation evaluation index;

[0202] Each objective weight is used to modify each subjective weight to obtain the comprehensive weight of each target reputation evaluation indicator.

[0203] In one embodiment, when the processor executes the computer program, the processor further implements the following steps:

[0204] Input each target reputation evaluation index into the hierarchical structure model for hierarchical decomposition processing to obtain multiple levels; the levels include target level, criterion level and index level;

[0205] For multiple target reputation evaluation indicators in each level, a pairwise comparison matrix is ​​determined; the pairwise comparison matrix is ​​used to represent the relative importance scores of every two target reputation evaluation indicators;

[0206] The eigenvector calculation and consistency check are performed on multiple element values ​​in the pairwise comparison matrix to obtain the subjective weight of each target reputation evaluation index.

[0207] In one embodiment, when the processor executes the computer program, the processor further implements the following steps:

[0208] For each level, historical experience data is used to compare multiple target reputation evaluation indicators in the level in pairs to obtain the indicator score results of the level;

[0209] According to the indicator score results at each level, a pairwise comparison matrix is ​​constructed.

[0210] In one embodiment, when the processor executes the computer program, the processor further implements the following steps:

[0211] Obtain sample data of each target reputation evaluation indicator;

[0212] Performing data standardization processing on sample data of each target reputation evaluation indicator to obtain standardized data of multiple target reputation evaluation indicators;

[0213] Determine the entropy value of each target reputation evaluation indicator according to the standardized data of each target reputation evaluation indicator by using an entropy value algorithm, and determine the credit utility value of each target reputation evaluation indicator according to the entropy value of each target reputation evaluation indicator;

[0214] Each credit utility value is normalized to obtain the objective weight of each target credit evaluation indicator.

[0215] In one embodiment, when the processor executes the computer program, the processor further implements the following steps:

[0216] For each sample evaluation indicator, determine the proportion of the sample evaluation indicator in multiple sample evaluation indicators;

[0217] According to the weight of each sample evaluation index and the number of samples obtained in advance, the entropy value of each target reputation evaluation index is determined.

[0218] In one embodiment, when the processor executes the computer program, the processor further implements the following steps:

[0219] The comprehensive evaluation values ​​of the candidate host machines are sorted, and the candidate host machine with the highest comprehensive evaluation value is determined as the target host machine.

[0220] According to some embodiments of the present application, a computer program product is also provided, and when the computer program is executed by a processor, the above method can be implemented. The computer program product includes one or more computer instructions. When these computer instructions are loaded and executed on a computer, part or all of the above method can be implemented in whole or in part according to the process or function described in the embodiment of the present application.

[0221] According to some embodiments of the present application, there is also provided a non-transitory computer-readable storage medium including instructions, such as a memory including instructions, and the above instructions can be executed by a processor of an electronic device to complete the above method. For example, the non-transitory computer-readable storage medium can be a ROM, a random access memory (RAM), a CD-ROM, a magnetic tape, a floppy disk, an optical data storage device, etc.

[0222] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, 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 this application can include at least one of non-volatile 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), magnetoresistive 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, etc., but are not limited to this.

[0223] 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 specification.

[0224] 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 method for determining a host machine, characterized in that: The method comprises: Determine multiple target reputation evaluation indicators for each candidate host machine based on the request information for establishing the virtual machine instance; For each of the candidate host machines, determining the subjective weight of each of the target reputation evaluation indicators, and modifying the subjective weight of each of the target reputation evaluation indicators to obtain a comprehensive weight of each of the target reputation evaluation indicators; Determining a comprehensive evaluation value of each of the candidate host machines according to each of the target reputation evaluation indicators and a comprehensive weight of each of the target reputation evaluation indicators; A target host machine is determined from the plurality of candidate host machines according to the comprehensive evaluation value of each of the candidate host machines.

2. The method according to claim 1, characterized in that The determining of the subjective weight of each target reputation evaluation indicator and correcting the subjective weight of each target reputation evaluation indicator to obtain the comprehensive weight of each target reputation evaluation indicator includes: Based on the pre-built hierarchical structure model, a subjective weight calculation process is performed on each of the target reputation evaluation indicators to obtain a subjective weight of each of the target reputation evaluation indicators; Using an entropy algorithm to perform objective weight calculation processing on each of the target reputation evaluation indicators to obtain an objective weight of each of the target reputation evaluation indicators; Each of the subjective weights is corrected using each of the objective weights to obtain a comprehensive weight of each of the target reputation evaluation indicators.

3. The method according to claim 2, characterized in that The subjective weight calculation process is performed on each of the target reputation evaluation indicators based on the pre-built hierarchical structure model to obtain the subjective weight of each of the target reputation evaluation indicators, including: Inputting each of the target reputation evaluation indicators into the hierarchical structure model for hierarchical decomposition processing to obtain multiple levels; the levels include a target level, a criterion level and an indicator level; For the multiple target reputation evaluation indicators in each of the levels, a pairwise comparison matrix is ​​determined; the pairwise comparison matrix is ​​used to represent the relative importance score of each two target reputation evaluation indicators; The characteristic vector calculation and consistency check processing are performed on the multiple element values ​​in the paired comparison matrix to obtain the subjective weight of each target reputation evaluation index.

4. The method according to claim 3, characterized in that Determining a pairwise comparison matrix for the plurality of target reputation evaluation indicators in each of the levels includes: For each of the levels, historical experience data is used to perform pairwise comparison processing on multiple target reputation evaluation indicators in the level to obtain the indicator score results of the level; The pairwise comparison matrix is ​​constructed according to the indicator score results of each level.

5. The method according to claim 2, characterized in that: The method of using an entropy algorithm to perform objective weight calculation processing on each of the target reputation evaluation indicators to obtain the objective weight of each of the target reputation evaluation indicators includes: Obtaining sample data of each of the target reputation evaluation indicators; Performing data standardization processing on the sample data of each target reputation evaluation indicator to obtain standardized data of multiple target reputation evaluation indicators; Determining the entropy value of each target reputation evaluation indicator according to the standardized data of each target reputation evaluation indicator; Determining the credit utility value of each target reputation evaluation indicator according to the entropy value of each target reputation evaluation indicator; Each of the credit utility values ​​is normalized to obtain an objective weight of each of the target credit evaluation indicators.

6. The method according to claim 5, characterized in that The standardized data of the target reputation evaluation index includes sample evaluation indexes of multiple samples; and determining the entropy value of each target reputation evaluation index according to the standardized data of each target reputation evaluation index includes: For each of the sample evaluation indicators, determining the proportion of the sample evaluation indicator in the plurality of sample evaluation indicators; The entropy value of each target reputation evaluation index is determined according to the weight of each sample evaluation index and the number of samples obtained in advance.

7. The method according to claim 6, characterized in that The step of determining the target host machine according to the comprehensive evaluation value of each candidate host machine includes: The comprehensive evaluation values ​​of the candidate host machines are sorted, and the candidate host machine with the highest comprehensive evaluation value is determined as the target host machine.

8. A host determination device, characterized in that: The device comprises: An indicator determination module, used to determine multiple target reputation evaluation indicators of each candidate host machine according to the request information for establishing a virtual machine instance; A comprehensive weight determination module, used to determine the subjective weight of each target reputation evaluation indicator for each candidate host, and to modify the subjective weight of each target reputation evaluation indicator to obtain a comprehensive weight of each target reputation evaluation indicator; An evaluation value determination module, used to determine the comprehensive evaluation value of each candidate host machine according to each target reputation evaluation index and the comprehensive weight of each target reputation evaluation index; The host machine determination module is used to determine a target host machine from a plurality of the candidate host machines according to the comprehensive evaluation value of each of the candidate host machines.

9. 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 7 are implemented.

10. 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 7 are implemented.

Images