Initialized configuration method and device, computer equipment and storage medium

Abstract

The invention relates to an initialization configuration method and device, computer equipment and a storage medium. The method comprises the steps of obtaining historical instance data according to historical cloud instances; acquiring a user input parameter, and performing intelligent recommendation in combination with the historical instance data to obtain a first configuration parameter; obtaining a performance prediction model based on logistic regression and hyper-parameter tuning according to the historical instance data; according to the performance prediction model, correcting the first configuration parameter to obtain a second configuration parameter; and realizing initial configuration of the cloud instance according to the second configuration parameter through a cloud instance customization tool. By adopting the method, different requirements can be flexibly met, the configuration parameters can be intelligently recommended, and automatic configuration of the cloud instance is realized, so that the flexibility and accuracy of configuration are improved, and the resource utilization rate is further improved.

Description

Technical Field

[0001] The present application relates to the technical field of initialization configuration, and in particular to an initialization configuration method, apparatus, computer equipment and storage medium. Background Art

[0002] With the development of cloud computing technology, it has become the core of modern information technology architecture. Enterprises and individual users are increasingly relying on cloud platforms to meet computing, storage and network needs. In a cloud computing environment, a cloud instance is a virtualized computing resource unit, similar to a physical server, and is usually composed of a combination of multiple hardware and software resources, including core resources such as CPU, memory, storage and network bandwidth. Users can configure cloud instances according to their needs to meet computing and performance requirements in different scenarios. Among them, the reasonable initialization configuration of cloud instance resources is crucial to system performance, resource utilization efficiency and cost control.

[0003] However, there are many deficiencies in the initialization configuration of cloud instances in the existing technology. Traditional cloud instance initialization usually relies on predefined templates or manual configuration, such as Cloud-Init, Ansible, etc. These tools automatically perform complex configuration tasks through scripts or configuration files, reducing manual intervention. However, the fixed template-based approach is difficult to flexibly respond to the personalized needs of different users, which can easily lead to unreasonable resource allocation, resource waste, performance bottlenecks, or failure to meet user-specific needs, and lacks intelligent and targeted configuration optimization capabilities. Summary of the invention

[0004] Based on this, it is necessary to provide an initialization configuration method, device, computer equipment and storage medium that can flexibly respond to different needs, intelligently recommend configuration parameters, and realize automatic configuration of cloud instances, thereby improving the flexibility and accuracy of configuration, and further improving resource utilization.

[0005] On the one hand, an initialization configuration method is provided, the method comprising:

[0006] According to the historical cloud instance, historical instance data is obtained;

[0007] Obtaining user input parameters, combining the historical instance data, performing intelligent recommendation, and obtaining first configuration parameters;

[0008] According to the historical instance data, a performance prediction model is obtained based on logistic regression and hyperparameter tuning;

[0009] According to the performance prediction model, the first configuration parameter is modified to obtain a second configuration parameter;

[0010] Through the cloud instance customization tool, the cloud instance initialization configuration is implemented according to the second configuration parameters.

[0011] In one embodiment, historical instance data is obtained based on historical cloud instances, including:

[0012] Based on the system log, extract the log records of the historical cloud instance;

[0013] According to the log records, the original data of historical instances are obtained;

[0014] Performing data preprocessing on the original data of the historical instance to obtain the historical instance data, wherein the data preprocessing includes at least one of the following:

[0015] Filling missing values ​​in the original data of the historical instance;

[0016] Removing outliers from the original data of the historical instance;

[0017] The duplicate data in the original data of the historical embodiment is deleted.

[0018] In one embodiment, obtaining the user input parameter, combining the historical instance data, and performing intelligent recommendation to obtain the first configuration parameter includes:

[0019] Acquiring the user input parameters, wherein the user input parameters include: the service required by the user and the performance expected by the user;

[0020] According to the historical instance data, historical installation services and historical performances are obtained;

[0021] According to the service required by the user, combined with the historical installed services, a service similarity calculation is performed to obtain a service similarity;

[0022] According to the user expected performance, combined with the historical performance, a performance similarity calculation is performed to obtain a performance similarity;

[0023] Combining the service similarity and the performance similarity to obtain scene similarity;

[0024] Sorting based on the scene similarity, selecting in turn a number of the historical cloud instances with the highest scene similarity, and obtaining a recommended reference instance group;

[0025] Acquire the historical instance data of the recommended reference instance group, extract the corresponding historical configuration parameters, and obtain the recommended reference configuration parameters;

[0026] A comprehensive calculation is performed based on the recommended reference configuration parameters to obtain the first configuration parameters.

[0027] In one embodiment, a performance prediction model is obtained based on the historical instance data, logistic regression and hyperparameter tuning, including:

[0028] Acquiring historical instance data that has undergone the data preprocessing;

[0029] Performing feature engineering on the historical instance data to obtain a historical instance standard data set;

[0030] Dividing the historical instance standard data set into a training set, a validation set, and a test set according to a preset ratio;

[0031] Constructing a logistic regression model based on the training set and the validation set;

[0032] Based on the training set and combined with hyperparameter tuning, a random forest model is constructed;

[0033] According to the test set, model performance evaluation is performed on the logistic regression model and the random forest model respectively to obtain a logistic regression model performance evaluation result and a random forest model performance evaluation result;

[0034] In response to the performance evaluation result of the logistic regression model being better than the performance evaluation result of the random forest model, the logistic regression model is selected as the performance prediction model; otherwise, the random forest model is selected as the performance prediction model.

[0035] In one embodiment, constructing a logistic regression model based on the training set and the validation set includes:

[0036] Constructing a plurality of basic models, and using the training set to train the plurality of basic models to obtain a plurality of basic model parameters;

[0037] According to the several basic model training parameters, the verification set is predicted by the several basic models to obtain a verification prediction result;

[0038] Constructing an original logistic regression model, and training the original logistic regression model according to the verification prediction result to obtain the logistic regression model;

[0039] Based on the training set, combined with hyperparameter tuning, a random forest model is constructed, including:

[0040] Build the original random forest model;

[0041] Selecting several key hyperparameters, and optimizing the several key hyperparameters based on grid search to obtain several preferred hyperparameters;

[0042] According to the several preferred hyperparameters and in combination with the training set, the original random forest model is trained to obtain the random forest model.

[0043] In one embodiment, constructing a logistic regression model based on the training set and the validation set further includes:

[0044] Constructing the several basic models, and using the training set to train the several basic models to obtain several basic model parameters, including:

[0045] Constructing a linear regression model, and using a training set to train the linear regression model to obtain linear regression model parameters;

[0046] Constructing a support vector machine model, and using a training set to train the support vector machine model to obtain support vector machine model parameters;

[0047] Constructing a decision tree model, and using a training set to train the decision tree model to obtain decision tree model parameters;

[0048] According to the several basic model training parameters, the verification set is predicted by the several basic models to obtain verification prediction results, including:

[0049] Predicting the validation set using the linear regression model to obtain a linear regression prediction result;

[0050] Predicting the validation set by using the support vector machine model to obtain a support vector machine prediction result;

[0051] Predicting the validation set using the decision tree model to obtain a decision tree prediction result;

[0052] Constructing an original logistic regression model, and training the original logistic regression model according to the verification prediction result to obtain the logistic regression model, including:

[0053] Construct the original logistic regression model;

[0054] Integrate the linear regression prediction result, the support vector machine prediction result and the decision tree prediction result, train the original logistic regression model, and obtain logistic regression model parameters;

[0055] The logistic regression model is obtained according to the logistic regression model parameters.

[0056] In one embodiment, the first configuration parameter is modified according to the performance prediction model to obtain the second configuration parameter, including:

[0057] Obtaining the first configuration parameter, wherein the first configuration parameter includes at least one of the following: number of CPU cores, memory size, disk size, and network bandwidth;

[0058] Performing performance prediction on the first configuration parameter by using the performance prediction model to obtain a performance prediction result;

[0059] Acquire the user input parameters to obtain the user expected performance;

[0060] Performing performance satisfaction judgment based on the performance prediction result and the user expected performance;

[0061] In response to the performance satisfaction judgment result being not meeting expectations, the first configuration parameter is modified to obtain a second configuration parameter.

[0062] Modifying the first configuration parameter to obtain a second configuration parameter also includes:

[0063] According to the performance prediction results, the performance is compared with the user's expected performance to obtain the configuration items with poor performance;

[0064] Formulate a configuration correction plan for the configuration item with poor performance according to the difference between the performance prediction result and the user's expected performance;

[0065] According to the configuration correction scheme, combined with the first configuration parameter, a corrected first configuration parameter is obtained;

[0066] Performing performance verification on the modified first configuration parameter through the performance prediction model to obtain a performance verification result;

[0067] In response to the performance verification result satisfying the user's expected performance, the modified first configuration parameter is used as the second configuration parameter.

[0068] On the other hand, an initialization configuration device is provided, the device comprising:

[0069] A historical data acquisition module is used to obtain historical instance data based on historical cloud instances;

[0070] A configuration parameter recommendation module, used to obtain user input parameters, combine the historical instance data, perform intelligent recommendation, and obtain the first configuration parameter;

[0071] A performance prediction model building module is used to obtain a performance prediction model based on the historical instance data, logistic regression and hyperparameter tuning;

[0072] A configuration parameter correction module, used to correct the first configuration parameter according to the performance prediction model to obtain a second configuration parameter;

[0073] The initialization configuration module is used to implement the cloud instance initialization configuration according to the second configuration parameters through the cloud instance customization tool.

[0074] In another aspect, a computer device is provided, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the following steps are implemented:

[0075] According to the historical cloud instance, historical instance data is obtained;

[0076] Obtaining user input parameters, combining the historical instance data, performing intelligent recommendation, and obtaining first configuration parameters;

[0077] According to the historical instance data, a performance prediction model is obtained based on logistic regression and hyperparameter tuning;

[0078] According to the performance prediction model, the first configuration parameter is modified to obtain a second configuration parameter;

[0079] Through the cloud instance customization tool, the cloud instance initialization configuration is implemented according to the second configuration parameters.

[0080] In another aspect, a computer-readable storage medium is provided, on which a computer program is stored, and when the computer program is executed by a processor, the following steps are implemented:

[0081] According to the historical cloud instance, historical instance data is obtained;

[0082] Obtaining user input parameters, combining the historical instance data, performing intelligent recommendation, and obtaining first configuration parameters;

[0083] According to the historical instance data, a performance prediction model is obtained based on logistic regression and hyperparameter tuning;

[0084] According to the performance prediction model, the first configuration parameter is modified to obtain a second configuration parameter;

[0085] Through the cloud instance customization tool, the cloud instance initialization configuration is implemented according to the second configuration parameters.

[0086] The above-mentioned initialization configuration method, device, computer equipment and storage medium, by combining historical instance data with user input parameters for intelligent recommendation, can flexibly respond to different scenarios and users' personalized needs, avoid the rigidity of traditional fixed template configuration, and improve the accuracy and efficiency of cloud instance configuration; at the same time, by generating configuration parameters based on historical data and intelligent recommendation, it can be more in line with actual usage scenarios and performance requirements, can effectively avoid over-configuration or under-configuration of resources, and optimize the resource utilization efficiency of cloud instances; in addition, by introducing a performance prediction model, the configuration parameters are further revised after the initial configuration recommendation to ensure that the selected configuration can meet the user's performance expectations, which is conducive to improving the performance and stability of the system. BRIEF DESCRIPTION OF THE DRAWINGS

[0087] Figure 1 An application environment diagram of an initialization configuration method in an embodiment;

[0088] Figure 2 A schematic diagram of a flow chart of an initialization configuration method in an embodiment;

[0089] Figure 3 A structural block diagram of an initialization configuration device in an embodiment;

[0090] Figure 4 FIG. 4 is a diagram showing the internal structure of a computer device in one embodiment. DETAILED DESCRIPTION

[0091] 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.

[0092] The initialization configuration method provided by this 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 terminal 102 can send a cloud instance initialization configuration request to the server 104, and after receiving the request, the server 104 automatically configures the cloud instance through an initialization configuration method provided by the present application. Among them, the terminal 102 can be, but is not limited to, various personal computers, mobile devices, embedded devices or developer workstations, and the server 104 can be implemented by an independent server or a server cluster composed of multiple servers.

[0093] In one embodiment, Figure 2 As shown, an initialization configuration method is provided, which is applied to Figure 1 The server 104 in the example is used as an example to illustrate, and the following steps are included:

[0094] Step 201, obtaining historical instance data according to the historical cloud instance.

[0095] Among them, historical cloud instances refer to virtual computing resource units previously created and run on the cloud platform; historical instance data are structured or unstructured data extracted from historical cloud instances, including key information such as instance configuration, performance, and usage records.

[0096] Specifically, the log records of the historical cloud instance are extracted to obtain the original data of the historical instance; and the original data of the historical instance are preprocessed to obtain the historical instance data, wherein the data preprocessing includes data cleaning and standardization.

[0097] Step 202, obtaining the user input parameters, combining the historical instance data, performing intelligent recommendation, and obtaining the first configuration parameters.

[0098] The user input parameters include the service type currently required by the user and the performance expected by the user, which are manually input by the user or loaded according to a preset script.

[0099] Specifically, the user input parameters are obtained, combined with the historical instance data, and based on similarity calculation, the scene similarity is obtained; based on the scene similarity, a recommended reference instance group is analyzed, and the corresponding recommended reference configuration parameters are extracted, and comprehensive calculation is performed to obtain the first configuration parameters.

[0100] Step 203: Obtain a performance prediction model based on the historical instance data, logistic regression and hyperparameter tuning.

[0101] Among them, logistic regression is an algorithm for binary classification prediction, which is often used to determine whether a certain configuration can meet specific performance requirements; hyperparameter tuning improves the accuracy of the model by finding the optimal combination of model parameters.

[0102] Specifically, the preprocessed historical instance data is divided into training set, validation set and test set; the logistic regression model is constructed using the training set, and the random forest model is constructed based on the hyperparameter tuning method; the performance of the two models is evaluated according to the results of the validation set and the test set, and the model with the best performance is selected as the performance prediction model.

[0103] Step 204: modify the first configuration parameter according to the performance prediction model to obtain a second configuration parameter.

[0104] Specifically, the first configuration parameter is simulated and predicted through the performance prediction model to determine whether it can meet the user's expected performance; in response to the judgment result that it does not meet the user's expected performance, the configuration item with insufficient performance is identified, and corresponding adjustments and optimizations are made, and resources are reallocated until it is verified that the user's expected performance is met, and finally the optimized second configuration parameter is obtained.

[0105] Step 205: Implement cloud instance initialization configuration according to the second configuration parameters through the cloud instance customization tool.

[0106] Among them, the cloud instance customization tool is an automated configuration tool that is used to automatically complete the initialization operation of the cloud instance according to the given configuration parameters. Commonly used cloud instance customization tools include cloud-int and Ansible.

[0107] Specifically, a customized tool is used to load the second configuration parameters to create and initialize a cloud instance, wherein the creation and initialization of the cloud instance at least includes: allocating storage and computing resources, configuring network bandwidth, and installing an operating system and necessary software packages.

[0108] In the above initialization configuration method, by combining historical instance data with user input parameters for intelligent recommendation, it can flexibly respond to different scenarios and users' personalized needs, avoid the rigidity of traditional fixed template configuration, and improve the accuracy and efficiency of cloud instance configuration; at the same time, by generating configuration parameters based on historical data and intelligent recommendations, it can be more in line with actual usage scenarios and performance requirements, effectively avoid over- or under-configuration of resources, and optimize the resource utilization efficiency of cloud instances; in addition, by introducing a performance prediction model, the configuration parameters are further revised after the initial configuration recommendation to ensure that the selected configuration can meet the user's performance expectations, which is conducive to improving the performance and stability of the system.

[0109] In one embodiment, historical instance data is obtained according to the historical cloud instance, including:

[0110] Based on the system log, extract the log records of the historical cloud instance;

[0111] According to the log records, the original data of historical instances are obtained;

[0112] Performing data preprocessing on the original data of the historical instance to obtain the historical instance data, wherein the data preprocessing includes at least one of the following:

[0113] Filling missing values ​​in the original data of the historical instance;

[0114] Removing outliers from the original data of the historical instance;

[0115] The duplicate data in the original data of the historical embodiment is deleted.

[0116] Specifically, this embodiment helps to eliminate data noise and anomalies by extracting log records of historical instances from system logs and preprocessing them, which can significantly improve the accuracy and completeness of historical data, provide a reliable data foundation for subsequent intelligent recommendations and performance predictions, and ultimately improve the accuracy of configuration recommendations.

[0117] In one embodiment, the user input parameters are obtained, combined with the historical instance data, and intelligent recommendation is performed to obtain the first configuration parameters, including:

[0118] Acquiring the user input parameters, wherein the user input parameters include: the service required by the user and the performance expected by the user;

[0119] According to the historical instance data, historical installation services and historical performances are obtained;

[0120] According to the service required by the user, combined with the historical installed services, a service similarity calculation is performed to obtain a service similarity;

[0121] According to the user expected performance, combined with the historical performance, a performance similarity calculation is performed to obtain a performance similarity;

[0122] Combining the service similarity and the performance similarity to obtain scene similarity;

[0123] Sorting based on the scene similarity, selecting in turn a number of the historical cloud instances with the highest scene similarity, and obtaining a recommended reference instance group;

[0124] Acquire the historical instance data of the recommended reference instance group, extract the corresponding historical configuration parameters, and obtain the recommended reference configuration parameters;

[0125] A comprehensive calculation is performed based on the recommended reference configuration parameters to obtain the first configuration parameters.

[0126] It is worth noting that obtaining the user input parameters, combining the historical instance data, and performing intelligent recommendation to obtain the first configuration parameters specifically includes:

[0127] Get the user input parameters, where the service required by the user is recorded as the user required service vector U s , represents the specific service type required by the user, such as network server, database service; the user expected performance is recorded as the user expected performance vector U p , represents the performance expected by users, such as response time and throughput;

[0128] According to the historical instance data, the historical installation service is recorded as a historical installation service vector represents the service installed in the i-th historical instance; the historical performance is recorded as the historical performance vector represents the performance of the i-th historical instance;

[0129] According to the user required service vector U s , combined with the historical installation service vector Perform service similarity calculation to obtain service similarity, wherein the service similarity is obtained based on the following formula:

[0130]

[0131] in, represents the service similarity between the i-th historical instance and the user input parameter, represents the dot product operation between the user's required service vector and the historical installed service vector, ||U s || and They represent the modulo operation of the user required service vector and the historical installed service vector respectively;

[0132] According to the user expected performance vector U p , combined with the historical performance vector The performance similarity is calculated to obtain the performance similarity, wherein the performance similarity is obtained based on the following formula:

[0133]

[0134] in, represents the performance similarity between the i-th historical instance and the user input parameters, represents the dot product operation between the user's expected performance vector and the historical performance vector, ||U p || and They represent the modulo operation of the user's expected performance vector and the historical performance vector respectively;

[0135] The service similarity and the performance similarity are combined to obtain the scene similarity, wherein the scene similarity is obtained based on the following formula:

[0136]

[0137] Among them, Sim scene (U,H i ) represents the scenario similarity between the i-th historical instance and the user input parameter, α and β are the weights of service similarity and performance similarity, respectively, which are set according to the actual situation and satisfy α+β=1;

[0138] After calculating the scene similarity between each historical instance and the user input parameter, sort them based on the scene similarity, select the k historical cloud instances with the highest scene similarity in turn, and obtain the recommended reference instance group, which is recorded as the recommended reference instance set

[0139] Specifically, when selecting the k historical cloud instances with the highest scene similarity, preferably 3-8 historical instances whose scene similarity meets the scene similarity threshold are selected, wherein the similarity threshold is preferably 0.6-0.8, and can also be set according to actual conditions;

[0140] According to the recommended reference instance set Get its historical configuration parameters, including at least: CPU quantity, network bandwidth, and disk capacity. The configuration parameters of the i-th historical instance are recorded as C i The recommended reference configuration parameters obtained are recorded as the recommended reference configuration parameter set

[0141] According to the recommended reference configuration parameters, a comprehensive calculation is performed to obtain the first configuration parameters, wherein the first configuration parameters are obtained based on the following formula:

[0142]

[0143] Among them, C first The obtained first configuration parameters include at least: a first CPU configuration quantity, a first network bandwidth configuration size, and a first disk configuration capacity.

[0144] Specifically, this embodiment refines the accuracy of recommendations by performing similarity calculations based on user needs and in combination with historical instances, and is able to find configuration parameters that are closest to user needs based on historical instance data, which is conducive to providing personalized and accurate configuration recommendations; at the same time, the first configuration parameters are obtained through intelligent recommendations, which avoids the rigidity of traditional fixed modules and reduces manual intervention, making the generated configuration parameters more accurate, which is conducive to reducing the time required for subsequent corrections, thereby improving the efficiency and accuracy of cloud instance initialization configuration.

[0145] In one embodiment, a performance prediction model is obtained based on the historical instance data, logistic regression and hyperparameter tuning, including:

[0146] Acquiring historical instance data that has undergone the data preprocessing;

[0147] Performing feature engineering on the historical instance data to obtain a historical instance standard data set;

[0148] Dividing the historical instance standard data set into a training set, a validation set, and a test set according to a preset ratio;

[0149] Constructing a logistic regression model based on the training set and the validation set;

[0150] Based on the training set and combined with hyperparameter tuning, a random forest model is constructed;

[0151] According to the test set, model performance evaluation is performed on the logistic regression model and the random forest model respectively to obtain a logistic regression model performance evaluation result and a random forest model performance evaluation result;

[0152] In response to the performance evaluation result of the logistic regression model being better than the performance evaluation result of the random forest model, the logistic regression model is selected as the performance prediction model; otherwise, the random forest model is selected as the performance prediction model.

[0153] Specifically, this embodiment, through the construction of logistic regression and random forest models, can use historical data to make accurate performance predictions, and improve the accuracy of the model through hyperparameter tuning, so that the performance prediction model can effectively correct the optimization process of configuration parameters and improve the accuracy and stability of configuration parameters; at the same time, through model performance evaluation, it is ensured that the selected prediction model has higher accuracy, which is conducive to further improving the reliability of the prediction results.

[0154] In one embodiment, constructing a logistic regression model based on the training set and the validation set includes:

[0155] Constructing a plurality of basic models, and using the training set to train the plurality of basic models to obtain a plurality of basic model parameters;

[0156] According to the several basic model training parameters, the verification set is predicted by the several basic models to obtain a verification prediction result;

[0157] Constructing an original logistic regression model, and training the original logistic regression model according to the verification prediction result to obtain the logistic regression model;

[0158] Based on the training set, combined with hyperparameter tuning, a random forest model is constructed, including:

[0159] Build the original random forest model;

[0160] Selecting several key hyperparameters, and optimizing the several key hyperparameters based on grid search to obtain several preferred hyperparameters;

[0161] According to the several preferred hyperparameters and in combination with the training set, the original random forest model is trained to obtain the random forest model.

[0162] Specifically, this embodiment can obtain more comprehensive prediction results by constructing and training multiple basic models. On this basis, the prediction results of the basic models are used to train the logistic regression model, which helps to further improve the performance and prediction accuracy of the overall model. At the same time, a random forest model is also constructed in addition to the logistic regression model, and optimized based on hyperparameter tuning, which not only makes the model more flexible in processing complex data, but also further improves the accuracy of performance prediction.

[0163] In one embodiment, constructing a logistic regression model based on the training set and the validation set further includes:

[0164] Constructing the several basic models, and using the training set to train the several basic models to obtain several basic model parameters, including:

[0165] Constructing a linear regression model, and using a training set to train the linear regression model to obtain linear regression model parameters;

[0166] Constructing a support vector machine model, and using a training set to train the support vector machine model to obtain support vector machine model parameters;

[0167] Constructing a decision tree model, and using a training set to train the decision tree model to obtain decision tree model parameters;

[0168] According to the several basic model training parameters, the verification set is predicted by the several basic models to obtain verification prediction results, including:

[0169] Predicting the validation set using the linear regression model to obtain a linear regression prediction result;

[0170] Predicting the validation set by using the support vector machine model to obtain a support vector machine prediction result;

[0171] Predicting the validation set using the decision tree model to obtain a decision tree prediction result;

[0172] Constructing an original logistic regression model, and training the original logistic regression model according to the verification prediction result to obtain the logistic regression model, including:

[0173] Construct the original logistic regression model;

[0174] Integrate the linear regression prediction result, the support vector machine prediction result and the decision tree prediction result, train the original logistic regression model, and obtain logistic regression model parameters;

[0175] The logistic regression model is obtained according to the logistic regression model parameters.

[0176] Specifically, this embodiment integrates the prediction results of multiple basic models such as linear regression, support vector machine and decision tree to construct a logistic regression model, which can integrate the advantages of multiple models and obtain more accurate logistic regression model parameters, thereby improving the overall prediction performance and the reliability of configuration parameter recommendations.

[0177] In one embodiment, the first configuration parameter is modified according to the performance prediction model to obtain the second configuration parameter, including:

[0178] Obtaining the first configuration parameter, wherein the first configuration parameter includes at least one of the following: number of CPU cores, memory size, disk size, and network bandwidth;

[0179] Performing performance prediction on the first configuration parameter by using the performance prediction model to obtain a performance prediction result;

[0180] Acquire the user input parameters to obtain the user expected performance;

[0181] Performing performance satisfaction judgment based on the performance prediction result and the user expected performance;

[0182] In response to the performance satisfaction judgment result being not meeting expectations, the first configuration parameter is modified to obtain a second configuration parameter.

[0183] Modifying the first configuration parameter to obtain a second configuration parameter also includes:

[0184] According to the performance prediction results, the performance is compared with the user's expected performance to obtain the configuration items with poor performance;

[0185] Formulate a configuration correction plan for the configuration item with poor performance according to the difference between the performance prediction result and the user's expected performance;

[0186] According to the configuration correction scheme, combined with the first configuration parameter, a corrected first configuration parameter is obtained;

[0187] Performing performance verification on the modified first configuration parameter through the performance prediction model to obtain a performance verification result;

[0188] In response to the performance verification result satisfying the user's expected performance, the modified first configuration parameter is used as the second configuration parameter.

[0189] Specifically, this embodiment corrects the first configuration parameter through the performance prediction model, and can dynamically adjust the configuration to ensure that the user's actual performance requirements are met, avoiding the occurrence of over- or under-configuration; at the same time, this embodiment introduces a performance verification mechanism, and determines whether the corrected configuration meets expectations through verification results, thereby further optimizing the configuration parameters to achieve optimal performance, further improving the flexibility and accuracy of the configuration.

[0190] It should be understood that although Figure 2 The steps in the flowchart are shown in sequence as indicated by the arrows, but these steps are not necessarily executed in the order indicated by the arrows. Unless otherwise specified in this document, there is no strict order restriction for the execution of these steps, and these steps can be executed in other orders. Moreover, Figure 2 At least part of the steps may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily executed at the same time, but can be executed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but can be executed in turn or alternately with other steps or at least part of the sub-steps or stages of other steps.

[0191] In one embodiment, Figure 3 As shown, an initialization configuration device is provided, including: a historical data acquisition module, a configuration parameter recommendation module, a performance prediction model construction module, a configuration parameter correction module and an initialization configuration module, wherein:

[0192] A historical data acquisition module is used to obtain historical instance data based on historical cloud instances;

[0193] A configuration parameter recommendation module, used to obtain user input parameters, combine the historical instance data, perform intelligent recommendation, and obtain the first configuration parameter;

[0194] A performance prediction model building module is used to obtain a performance prediction model based on the historical instance data, logistic regression and hyperparameter tuning;

[0195] A configuration parameter correction module, used to correct the first configuration parameter according to the performance prediction model to obtain a second configuration parameter;

[0196] The initialization configuration module is used to implement the cloud instance initialization configuration according to the second configuration parameters through the cloud instance customization tool.

[0197] The historical data acquisition module is used to extract the log records of the historical cloud instance based on the system log; obtain the original data of the historical instance according to the log records; perform data preprocessing on the original data of the historical instance to obtain the historical instance data, wherein the data preprocessing includes at least one of the following: filling missing values ​​for numerical data in the original data of the historical instance; removing outliers in the original data of the historical instance; and deleting duplicate data in the original data of the historical instance.

[0198] The configuration parameter recommendation module is used to obtain the user input parameters, wherein the user input parameters include: user required services and user expected performance; according to the historical instance data, obtain historical installation services and historical performance performance; according to the user required services, combined with the historical installation services, perform service similarity calculation to obtain service similarity; according to the user expected performance, combined with the historical performance performance, perform performance similarity calculation to obtain performance similarity; combine the service similarity and the performance similarity to obtain scene similarity; sort based on the scene similarity, select several historical cloud instances with the highest scene similarity in turn, and obtain a recommended reference instance group; obtain several historical instance data of the recommended reference instance group, extract several corresponding historical configuration parameters, and obtain recommended reference configuration parameters; perform comprehensive calculation based on the recommended reference configuration parameters to obtain the first configuration parameters.

[0199] The performance prediction model construction module is used to obtain historical instance data that has undergone the data preprocessing; perform feature engineering on the historical instance data to obtain a historical instance standard data set; divide the historical instance standard data set into a training set, a validation set, and a test set according to a preset ratio; construct a logistic regression model based on the training set and the validation set; construct a random forest model based on the training set in combination with hyperparameter tuning; perform model performance evaluation on the logistic regression model and the random forest model based on the test set, respectively, to obtain a logistic regression model performance evaluation result and a random forest model performance evaluation result; in response to the logistic regression model performance evaluation result being better than the random forest model performance evaluation result, the logistic regression model is selected as the performance prediction model, otherwise, the random forest model is selected as the performance prediction model.

[0200] The configuration parameter correction module is used to obtain the first configuration parameter, wherein the first configuration parameter includes at least one of the following: number of CPU cores, memory size, disk size, and network bandwidth; perform performance prediction on the first configuration parameter through the performance prediction model to obtain a performance prediction result; obtain the user input parameter to obtain the user expected performance; perform performance satisfaction judgment based on the performance prediction result and the user expected performance; in response to the performance satisfaction judgment result being not meeting expectations, correct the first configuration parameter to obtain a second configuration parameter.

[0201] The initialization configuration module is used to use a customized tool to load the second configuration parameters to create and initialize a cloud instance, wherein the creation and initialization of the cloud instance at least includes: allocating storage and computing resources, configuring network bandwidth, and installing an operating system and necessary software packages.

[0202] For the specific definition of the initialization configuration device, please refer to the definition of the initialization configuration method above, which will not be repeated here. Each module in the above-mentioned initialization configuration device can be implemented in whole or in part by software, hardware and a combination thereof. The above-mentioned modules can be embedded in or independent of the processor in the computer device in the form of hardware, or can be stored in the memory of the computer device in the form of software, so that the processor can call and execute the operations corresponding to the above modules.

[0203] In one embodiment, a computer device is provided. The computer device may be a server, and its internal structure diagram may be as follows: Figure 4 As shown. The computer device includes a processor, a memory, a network interface and a database connected through a system bus. 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 initialization configuration data. The network 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, an initialization configuration method is implemented.

[0204] Those skilled in the art will understand that Figure 4 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.

[0205] In one embodiment, a computer device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the following steps when executing the computer program:

[0206] According to the historical cloud instance, historical instance data is obtained;

[0207] Obtaining user input parameters, combining the historical instance data, performing intelligent recommendation, and obtaining first configuration parameters;

[0208] According to the historical instance data, a performance prediction model is obtained based on logistic regression and hyperparameter tuning;

[0209] According to the performance prediction model, the first configuration parameter is modified to obtain a second configuration parameter;

[0210] Through the cloud instance customization tool, the cloud instance initialization configuration is implemented according to the second configuration parameters.

[0211] In one embodiment, when the processor executes the computer program, the processor further implements the following steps:

[0212] According to the historical cloud instance, historical instance data is obtained, including:

[0213] Based on the system log, extract the log records of the historical cloud instance;

[0214] According to the log records, the original data of historical instances are obtained;

[0215] Performing data preprocessing on the original data of the historical instance to obtain the historical instance data, wherein the data preprocessing includes at least one of the following:

[0216] Filling missing values ​​in the original data of the historical instance;

[0217] Removing outliers from the original data of the historical instance;

[0218] The duplicate data in the original data of the historical embodiment is deleted.

[0219] In one embodiment, when the processor executes the computer program, the processor further implements the following steps:

[0220] Obtaining user input parameters, combining the historical instance data, and performing intelligent recommendation to obtain first configuration parameters, including:

[0221] Acquiring the user input parameters, wherein the user input parameters include: the service required by the user and the performance expected by the user;

[0222] According to the historical instance data, historical installation services and historical performances are obtained;

[0223] According to the service required by the user, combined with the historical installed services, a service similarity calculation is performed to obtain a service similarity;

[0224] According to the user expected performance, combined with the historical performance, a performance similarity calculation is performed to obtain a performance similarity;

[0225] Combining the service similarity and the performance similarity to obtain scene similarity;

[0226] Sorting based on the scene similarity, selecting in turn a number of the historical cloud instances with the highest scene similarity, and obtaining a recommended reference instance group;

[0227] Acquire the historical instance data of the recommended reference instance group, extract the corresponding historical configuration parameters, and obtain the recommended reference configuration parameters;

[0228] A comprehensive calculation is performed based on the recommended reference configuration parameters to obtain the first configuration parameters.

[0229] In one embodiment, when the processor executes the computer program, the processor further implements the following steps:

[0230] According to the historical instance data, based on logistic regression and hyperparameter tuning, a performance prediction model is obtained, including:

[0231] Acquiring historical instance data that has undergone the data preprocessing;

[0232] Performing feature engineering on the historical instance data to obtain a historical instance standard data set;

[0233] Dividing the historical instance standard data set into a training set, a validation set, and a test set according to a preset ratio;

[0234] Constructing a logistic regression model based on the training set and the validation set;

[0235] Based on the training set and combined with hyperparameter tuning, a random forest model is constructed;

[0236] According to the test set, model performance evaluation is performed on the logistic regression model and the random forest model respectively to obtain a logistic regression model performance evaluation result and a random forest model performance evaluation result;

[0237] In response to the performance evaluation result of the logistic regression model being better than the performance evaluation result of the random forest model, the logistic regression model is selected as the performance prediction model; otherwise, the random forest model is selected as the performance prediction model;

[0238] Constructing a logistic regression model based on the training set and the validation set also includes:

[0239] Constructing a plurality of basic models, and using the training set to train the plurality of basic models to obtain a plurality of basic model parameters;

[0240] According to the several basic model training parameters, the verification set is predicted by the several basic models to obtain a verification prediction result;

[0241] Constructing an original logistic regression model, and training the original logistic regression model according to the verification prediction result to obtain the logistic regression model;

[0242] Based on the training set, combined with hyperparameter tuning, a random forest model is constructed, which also includes:

[0243] Build the original random forest model;

[0244] Selecting several key hyperparameters, and optimizing the several key hyperparameters based on grid search to obtain several preferred hyperparameters;

[0245] According to the several preferred hyperparameters and in combination with the training set, the original random forest model is trained to obtain the random forest model.

[0246] In one embodiment, when the processor executes the computer program, the processor further implements the following steps:

[0247] Constructing the several basic models, and using the training set to train the several basic models to obtain several basic model parameters, including:

[0248] Constructing a linear regression model, and using a training set to train the linear regression model to obtain linear regression model parameters;

[0249] Constructing a support vector machine model, and using a training set to train the support vector machine model to obtain support vector machine model parameters;

[0250] Constructing a decision tree model, and using a training set to train the decision tree model to obtain decision tree model parameters;

[0251] According to the several basic model training parameters, the verification set is predicted by the several basic models to obtain verification prediction results, including:

[0252] Predicting the validation set using the linear regression model to obtain a linear regression prediction result;

[0253] Predicting the validation set by using the support vector machine model to obtain a support vector machine prediction result;

[0254] Predicting the validation set using the decision tree model to obtain a decision tree prediction result;

[0255] Constructing an original logistic regression model, and training the original logistic regression model according to the verification prediction result to obtain the logistic regression model, including:

[0256] Construct the original logistic regression model;

[0257] Integrate the linear regression prediction result, the support vector machine prediction result and the decision tree prediction result, train the original logistic regression model, and obtain logistic regression model parameters;

[0258] The logistic regression model is obtained according to the logistic regression model parameters.

[0259] In one embodiment, when the processor executes the computer program, the processor further implements the following steps:

[0260] Obtaining the first configuration parameter, wherein the first configuration parameter includes at least one of the following: number of CPU cores, memory size, disk size, and network bandwidth;

[0261] Performing performance prediction on the first configuration parameter by using the performance prediction model to obtain a performance prediction result;

[0262] Acquire the user input parameters to obtain the user expected performance;

[0263] Performing performance satisfaction judgment based on the performance prediction result and the user expected performance;

[0264] In response to the performance satisfaction judgment result being not meeting expectations, the first configuration parameter is modified to obtain a second configuration parameter.

[0265] Modifying the first configuration parameter to obtain a second configuration parameter also includes:

[0266] According to the performance prediction results, the performance is compared with the user's expected performance to obtain the configuration items with poor performance;

[0267] Formulate a configuration correction plan for the configuration item with poor performance according to the difference between the performance prediction result and the user's expected performance;

[0268] According to the configuration correction scheme, combined with the first configuration parameter, a corrected first configuration parameter is obtained;

[0269] Performing performance verification on the modified first configuration parameter through the performance prediction model to obtain a performance verification result;

[0270] In response to the performance verification result satisfying the user's expected performance, the modified first configuration parameter is used as the second configuration parameter.

[0271] In one embodiment, a computer readable storage medium is provided, on which a computer program is stored, and when the computer program is executed by a processor, the following steps are implemented:

[0272] According to the historical cloud instance, historical instance data is obtained;

[0273] Obtaining user input parameters, combining the historical instance data, performing intelligent recommendation, and obtaining first configuration parameters;

[0274] According to the historical instance data, a performance prediction model is obtained based on logistic regression and hyperparameter tuning;

[0275] According to the performance prediction model, the first configuration parameter is modified to obtain a second configuration parameter;

[0276] Through the cloud instance customization tool, the cloud instance initialization configuration is implemented according to the second configuration parameters.

[0277] 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 memory, storage, database or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. As an illustration and not limitation, RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).

[0278] 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.

[0279] The above-mentioned 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 invention patent. 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 patent of the present application shall be subject to the attached claims.

Claims

1. An initialization configuration method, characterized in that: include: According to the historical cloud instance, historical instance data is obtained; Obtaining user input parameters, combining the historical instance data, performing intelligent recommendation, and obtaining first configuration parameters; According to the historical instance data, a performance prediction model is obtained based on logistic regression and hyperparameter tuning; According to the performance prediction model, the first configuration parameter is modified to obtain a second configuration parameter; Through the cloud instance customization tool, the cloud instance initialization configuration is implemented according to the second configuration parameters.

2. An initialization configuration method according to claim 1, characterized in that: According to the historical cloud instance, historical instance data is obtained, including: Based on the system log, extract the log records of the historical cloud instance; According to the log records, the original data of historical instances are obtained; Performing data preprocessing on the original data of the historical instance to obtain the historical instance data, wherein the data preprocessing includes at least one of the following: Filling missing values ​​in the original data of the historical instance; Removing outliers from the original data of the historical instance; The duplicate data in the original data of the historical embodiment is deleted.

3. An initialization configuration method according to claim 1, characterized in that: Obtaining user input parameters, combining the historical instance data, and performing intelligent recommendation to obtain first configuration parameters, including: Acquiring the user input parameters, wherein the user input parameters include: the service required by the user and the performance expected by the user; According to the historical instance data, historical installation services and historical performances are obtained; According to the service required by the user, combined with the historical installed services, a service similarity calculation is performed to obtain a service similarity; According to the user expected performance, combined with the historical performance, a performance similarity calculation is performed to obtain a performance similarity; Combining the service similarity and the performance similarity to obtain scene similarity; Sorting based on the scene similarity, selecting in turn a number of the historical cloud instances with the highest scene similarity, and obtaining a recommended reference instance group; Acquire the historical instance data of the recommended reference instance group, extract the corresponding historical configuration parameters, and obtain the recommended reference configuration parameters; A comprehensive calculation is performed based on the recommended reference configuration parameters to obtain the first configuration parameters.

4. An initialization configuration method according to claim 2, characterized in that: According to the historical instance data, based on logistic regression and hyperparameter tuning, a performance prediction model is obtained, including: Acquiring historical instance data that has undergone the data preprocessing; Performing feature engineering on the historical instance data to obtain a historical instance standard data set; Dividing the historical instance standard data set into a training set, a validation set, and a test set according to a preset ratio; Constructing a logistic regression model based on the training set and the validation set; Based on the training set and combined with hyperparameter tuning, a random forest model is constructed; According to the test set, model performance evaluation is performed on the logistic regression model and the random forest model respectively to obtain a logistic regression model performance evaluation result and a random forest model performance evaluation result; In response to the performance evaluation result of the logistic regression model being better than the performance evaluation result of the random forest model, the logistic regression model is selected as the performance prediction model; otherwise, the random forest model is selected as the performance prediction model.

5. An initialization configuration method according to claim 4, characterized in that: Constructing a logistic regression model based on the training set and the validation set, including: Constructing a plurality of basic models, and using the training set to train the plurality of basic models to obtain a plurality of basic model parameters; According to the several basic model training parameters, the verification set is predicted by the several basic models to obtain a verification prediction result; Constructing an original logistic regression model, and training the original logistic regression model according to the verification prediction result to obtain the logistic regression model; Based on the training set, combined with hyperparameter tuning, a random forest model is constructed, including: Build the original random forest model; Selecting several key hyperparameters, and optimizing the several key hyperparameters based on grid search to obtain several preferred hyperparameters; According to the several preferred hyperparameters and in combination with the training set, the original random forest model is trained to obtain the random forest model.

6. An initialization configuration method according to claim 5, characterized in that: Constructing a logistic regression model based on the training set and the validation set also includes: Constructing the several basic models, and using the training set to train the several basic models to obtain several basic model parameters, including: Constructing a linear regression model, and using a training set to train the linear regression model to obtain linear regression model parameters; Constructing a support vector machine model, and using a training set to train the support vector machine model to obtain support vector machine model parameters; Constructing a decision tree model, and using a training set to train the decision tree model to obtain decision tree model parameters; According to the several basic model training parameters, the verification set is predicted by the several basic models to obtain verification prediction results, including: Predicting the validation set using the linear regression model to obtain a linear regression prediction result; Predicting the validation set by using the support vector machine model to obtain a support vector machine prediction result; Predicting the validation set using the decision tree model to obtain a decision tree prediction result; Constructing an original logistic regression model, and training the original logistic regression model according to the verification prediction result to obtain the logistic regression model, including: Construct the original logistic regression model; Integrate the linear regression prediction result, the support vector machine prediction result and the decision tree prediction result, train the original logistic regression model, and obtain logistic regression model parameters; The logistic regression model is obtained according to the logistic regression model parameters.

7. An initialization configuration method according to claim 3, characterized in that: According to the performance prediction model, the first configuration parameter is modified to obtain a second configuration parameter, including: Obtaining the first configuration parameter, wherein the first configuration parameter includes at least one of the following: number of CPU cores, memory size, disk size, and network bandwidth; Performing performance prediction on the first configuration parameter by using the performance prediction model to obtain a performance prediction result; Acquire the user input parameters to obtain the user expected performance; Performing performance satisfaction judgment based on the performance prediction result and the user expected performance; In response to the performance satisfaction judgment result being not meeting expectations, the first configuration parameter is modified to obtain a second configuration parameter. Modifying the first configuration parameter to obtain a second configuration parameter also includes: According to the performance prediction results, the performance is compared with the user's expected performance to obtain the configuration items with poor performance; Formulate a configuration correction plan for the configuration item with poor performance according to the difference between the performance prediction result and the user's expected performance; According to the configuration correction scheme, combined with the first configuration parameter, a corrected first configuration parameter is obtained; Performing performance verification on the modified first configuration parameter through the performance prediction model to obtain a performance verification result; In response to the performance verification result satisfying the user's expected performance, the modified first configuration parameter is used as the second configuration parameter.

8. An initialization configuration device, characterized in that: The device comprises: A historical data acquisition module is used to obtain historical instance data based on historical cloud instances; A configuration parameter recommendation module, used to obtain user input parameters, combine the historical instance data, perform intelligent recommendation, and obtain the first configuration parameter; A performance prediction model building module is used to obtain a performance prediction model based on the historical instance data, logistic regression and hyperparameter tuning; A configuration parameter correction module, used to correct the first configuration parameter according to the performance prediction model to obtain a second configuration parameter; The initialization configuration module is used to implement the cloud instance initialization configuration according to the second configuration parameters through the cloud instance customization tool.

9. A computer device comprising a memory, a processor and a computer program stored in the memory and executable on the processor, characterized in that: 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