Resource demand prediction method and device, cloud platform operation and maintenance system and equipment

Abstract

The embodiment of the invention provides a resource demand prediction method and device, and a cloud platform operation and maintenance system and equipment, and relates to the technical field of computers, and the method comprises the steps: for each sample service system in a plurality of sample service systems, obtaining an index value of the sample service system for each index in a first index set at a first historical moment, the resource size of the preset resource allocated at the first historical moment is determined; taking each index in the first index set as an independent variable, taking a preset resource as a dependent variable, and constructing a multiple regression model by utilizing the index value of each index in the first index set and the corresponding resource size at a first historical moment to obtain a resource prediction model; and processing the index values of the target service system for the indexes in the first index set by using the resource prediction model to obtain the resource size which needs to be allocated to the target service system for the preset resource. Therefore, resource waste and performance bottleneck can be avoided to a certain extent.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of computer, in particular to a resource demand prediction method and device, a cloud platform operation and maintenance system and equipment. BACKGROUND

[0002] With the rapid development of cloud computing, big data and virtualization technology, the allocation and efficient use of resources have become one of the core problems in information technology infrastructure management. For example, in practical applications, it is necessary to allocate appropriate CPU, memory and other system resources to business systems to ensure the stability and efficiency of services.

[0003] However, in the prior art, resources are often allocated to business systems based on experience, which can easily lead to resource waste or performance bottlenecks. SUMMARY

[0004] The purpose of the embodiments of the present application is to provide a resource demand prediction method, device, cloud platform operation and maintenance system and equipment to avoid resource waste and performance bottlenecks to some extent. The specific technical solutions are as follows:

[0005] In a first aspect, the embodiments of the present application provide a resource demand prediction method, which comprises:

[0006] For each sample business system in a plurality of different types of sample business systems, the index values of each index in a first index set and the resource size of a preset resource allocated at a first historical time are obtained for the sample business system; wherein the first index set comprises: an index representing attribute information of a business system, and / or an index representing the running state of a business system;

[0007] Using the index values of each index in the first index set and the corresponding resource size at the first historical time, a multiple regression model is constructed with each index in the first index set as the independent variable and the preset resource as the dependent variable, to obtain a resource prediction model;

[0008] The index values of each index in the first index set for a target business system are processed using the resource prediction model to obtain the resource size of the preset resource that needs to be allocated to the target business system.

[0009] Optionally, each index in the first index set is determined by the following steps:

[0010] For each sample business system, the index values of each index in a second index set and the resource size of a preset resource allocated at a second historical time are obtained for the sample business system; wherein the first index set belongs to the second index set;

[0011] For each index in the second index set, a correlation coefficient representing a degree of correlation between the index and the preset resource is calculated based on the index values of the indexes in the second index set and the corresponding resource sizes at the second historical time;

[0012] An index is selected from the second index set based on the calculated correlation coefficients, and the first index set is obtained.

[0013] Optionally, the selecting of the index from the second index set based on the calculated correlation coefficients to obtain the first index set comprises:

[0014] selecting an index from the second index set whose corresponding correlation coefficient satisfies a preset screening condition to obtain the first index set;

[0015] The preset screening condition comprises at least one of the following:

[0016] the correlation degree represented by the corresponding correlation coefficient is not less than a preset degree threshold;

[0017] the correlation degree represented by the corresponding correlation coefficient conforms to a business logic between the index and the preset resource;

[0018] there is no multicollinearity between the index and other indexes in the second index set.

[0019] Optionally, the multivariate regression model is constructed based on the index values of the indexes in the first index set and the corresponding resource sizes at the first historical time, with the indexes in the first index set as independent variables and the preset resource as a dependent variable, comprising:

[0020] the index values of the indexes in the first index set and the corresponding resource sizes at the first historical time are preprocessed;

[0021] The preprocessing comprises at least one of the following:

[0022] missing values are filled in;

[0023] abnormal values are removed;

[0024] the multivariate regression model is constructed based on the preprocessed index values and the corresponding resource sizes, with the indexes in the first index set as independent variables and the preset resource as a dependent variable.

[0025] Optionally, the indexes of the attribute information of the business system include at least one of the following: a system business type, a business deployment type, whether there is a demilitarized zone (DMZ) service, and a service type; and the indexes of the running state of the business system include at least one of the following: a total number of registered users, a number of daily active users, a number of concurrent users at a peak, a number of services started, a total data volume of business files, a year-on-year increase of the total data volume of business files, a total data volume of databases, and a year-on-year increase of the total data volume of databases.

[0026] The preset resource indicates a CPU memory capacity and / or a number of CPU cores.

[0027] Optionally, the method further includes:

[0028] The resource prediction model is updated by using the index values of the first set of indexes and the resource size of the preset resource actually allocated to the target business system.

[0029] In a second aspect, an embodiment of the present application provides a cloud platform operation and maintenance system, including: a resource demand prediction agent;

[0030] The resource demand prediction agent is configured to execute any of the methods described above.

[0031] Optionally, the system further includes:

[0032] A work assignment agent, a monitoring sentinel agent, an asset inspection agent, and an alarm diagnosis agent.

[0033] The work assignment agent is configured to receive task information of a current operation and maintenance task input by a user, generate an operation and maintenance operation command based on the task information, and issue the operation and maintenance operation command to a device indicated by the task information for execution, or call the monitoring sentinel agent to execute an operation and maintenance task indicated by the task information, or call the asset inspection agent to execute an operation and maintenance task indicated by the task information.

[0034] The monitoring sentinel agent is configured to use a large language model to analyze monitoring index data of a monitored device indicated by the task information, and generate an analysis result in a natural language form.

[0035] The asset inspection agent is configured to use an inspection script of a to-be-inspected device indicated by the task information to perform inspection on the to-be-inspected device.

[0036] The alarm diagnosis agent is configured to use a large language model to analyze and diagnose received original alarm data, and obtain an alarm diagnosis result.

[0037] In a third aspect, an embodiment of the present application provides a device for constructing a resource prediction model, the device comprising:

[0038] a data obtaining module configured to, for each sample business system in a plurality of different types of sample business systems, obtain an index value of each index in a first index set for the sample business system at a first historical time and a resource size of a preset resource allocated at the first historical time, wherein the first index set comprises an index representing attribute information of the business system and / or an index representing a running state of the business system;

[0039] a model constructing module configured to, taking each index in the first index set as an independent variable and taking the preset resource as a dependent variable, construct a multiple regression model by using the index value of each index in the first index set at the first historical time and the corresponding resource size, to obtain the resource prediction model;

[0040] a demand prediction module configured to process the index value of each index in the first index set for a target business system by using the resource prediction model, to obtain a resource size of the preset resource that needs to be allocated to the target business system.

[0041] Optionally, the device further comprises:

[0042] an index set determining module configured to, for each sample business system, obtain an index value of each index in a second index set for the sample business system at a second historical time and a resource size of a preset resource allocated at the second historical time, wherein the first index set belongs to the second index set;

[0043] for each index in the second index set, calculate a correlation coefficient representing a correlation degree between the index and the preset resource by using the index value of each index in the second index set at the second historical time and the corresponding resource size;

[0044] selecting an index from the second index set based on the calculated correlation coefficients, to obtain the first index set.

[0045] Optionally, the index set determining module is specifically configured to select an index from the second index set for which a corresponding correlation coefficient satisfies a preset screening condition, to obtain the first index set;

[0046] wherein the preset screening condition comprises at least one of the following:

[0047] the correlation degree represented by the corresponding correlation coefficient is not less than a preset degree threshold;

[0048] the correlation degree represented by the corresponding correlation coefficient conforms to a business logic between the index and the preset resource;

[0049] There is no multicollinearity between the second set of indicators and other indicators.

[0050] Optionally, the construction module is specifically configured to preprocess the indicator values and the corresponding resource sizes of the first set of indicators at the first historical time point.

[0051] The preprocessing includes at least one of the following:

[0052] Filling in missing values;

[0053] Eliminating abnormal values;

[0054] Using the preprocessed indicator values and the corresponding resource sizes, a multiple regression model is constructed with the indicators in the first set of indicators as independent variables and the preset resource as dependent variable.

[0055] Optionally, the indicators representing the attribute information of the business system include at least one of the following: system business type, business deployment type, whether there is a DMZ service, and service type; and the indicators representing the running state of the business system include at least one of the following: total number of registered users, number of daily active users, number of concurrent users at peak, number of services started, total amount of business files, year-on-year increase of total amount of business files, total amount of databases, and year-on-year increase of total amount of databases.

[0056] The preset resource represents CPU memory capacity and / or CPU core number.

[0057] Optionally, the apparatus further includes:

[0058] The module updating module is configured to update the resource prediction model using the indicator values of the target business system for the indicators in the first set of indicators and the resource sizes of the preset resource actually allocated to the target business system.

[0059] In a fourth aspect, an embodiment of the present application provides an electronic device, including a processor, a communication interface, a memory and a communication bus, wherein the processor, the communication interface and the memory complete mutual communication through the communication bus;

[0060] The memory is used to store a computer program.

[0061] The processor is used to execute the program stored on the memory, and implement the method steps of any of the above aspects.

[0062] In a fifth aspect, an embodiment of the present application provides a computer readable storage medium, the computer readable storage medium stores a computer program, and the computer program is executed by a processor to implement the method steps of any of the above aspects.

[0063] Sixthly, embodiments of this application provide a computer program product containing instructions that, when run on a computer, cause the computer to perform any of the methods described above.

[0064] Beneficial effects of the embodiments in this application:

[0065] The resource demand prediction method provided in this application provides the following steps for each sample business system among multiple sample business systems of different types: First, it obtains the index values ​​of each index in a first index set for that sample business system at a first historical time, and the resource size of the preset resources allocated at that first historical time. The first index set includes: indicators representing the attribute information of the business system, and / or, indicators representing the operating status of the business system. Using each index in the first index set as the independent variable and the preset resources as the dependent variable, a multiple regression model is constructed using the index values ​​of each index in the first index set at the first historical time and the corresponding resource sizes to obtain a resource prediction model. The resource prediction model is then used to process the index values ​​of each index in the first index set for the target business system to obtain the resource size that needs to be allocated to the target business system for the preset resources.

[0066] Based on the above processing, since the resource size allocated at historical time points reflects the actual demand of the sample business system for the preset resources, and the indicator values ​​for each metric at historical time points reflect the actual state of the sample business system (including actual attributes and / or actual operating status), a resource prediction model can be constructed using the indicator values ​​and corresponding resource sizes at historical time points. This model can reflect the correlation between the actual state of the business system and the required resource size. Accordingly, the resource size obtained by processing the indicator values ​​of the target business system using the resource prediction model is adapted to the actual state of the target business system. Allocating resources to the target business system according to this resource size can, to a certain extent, avoid resource waste and performance bottlenecks.

[0067] Of course, implementing any product or method of this application does not necessarily require achieving all of the advantages described above at the same time. Attached Figure Description

[0068] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other embodiments can be obtained based on these drawings.

[0069] Figure 1 A flowchart illustrating a resource demand forecasting method provided in an embodiment of this application;

[0070] Figure 2This application provides a schematic diagram of the structure of a cloud platform operation and maintenance system.

[0071] Figure 3 This is a schematic diagram of the structure of a resource demand forecasting device provided in an embodiment of this application;

[0072] Figure 4 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Detailed Implementation

[0073] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art based on this application are within the scope of protection of this application.

[0074] Service providers can offer users different functional services based on their business systems, which can be categorized into different types according to their functions. For example, there are business systems for providing big data services, business systems for providing office applications, and business systems for providing sales services.

[0075] In existing technologies, resources are often allocated to business systems based on experience, which can easily lead to resource waste or performance bottlenecks.

[0076] To address the aforementioned technical problems, embodiments of this application provide a resource demand prediction method. This method can be applied to electronic devices, which can predict the resource demands of a business system based on the method provided in this application, thereby obtaining the amount of resources that need to be allocated to the business system.

[0077] See Figure 1 , Figure 1 This application provides a flowchart illustrating a resource demand forecasting method, which includes the following steps:

[0078] S101: For each of the multiple sample business systems of different types, obtain the indicator values ​​of each indicator in the first indicator set for that sample business system at the first historical time, as well as the resource size of the preset resources allocated at the first historical time.

[0079] The first set of indicators includes: indicators that characterize the attribute information of the business system, and / or indicators that characterize the operating status of the business system.

[0080] S102: Using each indicator in the first indicator set as the independent variable and the preset resource as the dependent variable, a multiple regression model is constructed using the indicator values ​​and corresponding resource sizes of each indicator in the first historical moment to obtain the resource prediction model.

[0081] S103: Use the resource prediction model to process the indicator values ​​of each indicator in the first indicator set for the target business system, and obtain the amount of resources that need to be allocated to the target business system for the preset resources.

[0082] Based on the above processing, since the resource size allocated at historical time points reflects the actual demand of the sample business system for the preset resources, and the indicator values ​​for each metric at historical time points reflect the actual state of the sample business system (including actual attributes and / or actual operating status), a resource prediction model can be constructed using the indicator values ​​and corresponding resource sizes at historical time points. This model can reflect the correlation between the actual state of the business system and the required resource size. Accordingly, the resource size obtained by processing the indicator values ​​of the target business system using the resource prediction model is adapted to the actual state of the target business system. Allocating resources to the target business system according to this resource size can, to a certain extent, avoid resource waste and performance bottlenecks.

[0083] Regarding step S101, since the resource size required by a business system is related to its own type, in order to make the constructed resource prediction model adaptable to different types of business systems, that is, in order to make the constructed resource prediction model able to predict the resource size required by different types of business systems, the indicator values ​​and corresponding resource sizes can be obtained for multiple different types of business systems (i.e. sample business systems).

[0084] Understandably, the metric values ​​of the sample business system reflect its true state, and the corresponding resource size is the actual resource size allocated to the sample business system. Therefore, the metric values ​​and corresponding resource sizes of a sample business system are well-matched. In other words, the resources represented by the resource size of the sample business system, when processed under the true state represented by the corresponding metric values, will not result in resource waste or performance bottlenecks.

[0085] Multiple sample business systems can include business systems of the same type; that is, multiple sample business systems can be selected for the same type. Furthermore, to further improve the effectiveness of the resource prediction model, the selected sample business systems can include all existing types of business systems.

[0086] The first set of indicators contains multiple indicators, and indicators that represent the attribute information of the business system can be called attribute indicators. It can be understood that the indicator values ​​of attribute indicators represent the attributes of the business system itself, that is, they represent the inherent state of the business system, which does not change over time.

[0087] Metrics that indicate the operational status of a business system can be called operational metrics. It's understandable that the value of the same operational metric for the same business system will change at different times; it's not fixed. Therefore, resource prediction models based on these metrics can adapt to dynamic changes in business load.

[0088] In one embodiment, the metrics (i.e., attribute metrics) that characterize the attribute information of a business system include at least one of the following: system business type, business deployment type, presence of isolated zone services, and service type.

[0089] System business type represents the type of services provided by a business system. For example, according to system business type, business systems can be divided into: business systems that provide basic platform services, business systems that provide big data services, business systems that provide office applications, and business systems that provide sales services, etc.

[0090] Based on the business deployment model, business systems can be divided into: business systems based on IaaS (Infrastructure as a Service) + PaaS (Platform as a Service), business systems based on PaaS, and business systems based on IaaS, etc.

[0091] An isolation zone is a logical or physical segment located between an organization's internal network and external networks. It is used to house public servers such as web (World Wide Web) servers and mail servers, protecting the security of internal network data through isolation between internal and external networks.

[0092] Based on service type, business systems can be divided into: computational business systems and non-computational business systems; non-computational business systems can also be called in-memory business systems. Computational business systems handle tasks with high data processing pressure, such as financial and scientific analysis. Whether a business system is a computational business system can be determined based on the amount of data computation during processing. For example, if the amount of data computation performed by a business processing system exceeds a preset computation threshold within a certain time period, then the business system can be identified as a computational business system.

[0093] The metrics characterizing the operational status of a business system (i.e., operational metrics) include at least one of the following: total number of registered users, number of daily active users, peak concurrent users, number of services activated, total data volume of business files, month-on-month increase in total data volume of business files, total data volume of the database, and month-on-month increase in total data volume of the database.

[0094] For any given business system, the more registered users it has, the greater its load and the more resources it will require. Therefore, the total number of registered users can be used as an indicator to build a resource prediction model. For example, the total number of registered users can be calculated by counting the total number of users who registered in a sample business system from its launch to the first historical moment.

[0095] For any given business system, a higher daily active user count will result in a greater system load and require more resources. Therefore, the number of daily active users can be used as a metric to build a resource prediction model. For example, one could count the number of daily active users of a sample business system from its launch to the first historical moment.

[0096] For any business system, a higher peak concurrent user count leads to greater system load and requires more resources. Therefore, peak concurrent user count can be used as a metric to build a resource prediction model. For example, the peak concurrent user count can be calculated by counting the maximum number of concurrent user accesses the system needs to handle during the peak period from the system's launch to the first historical moment. Alternatively, the peak concurrent user count can be estimated based on the average concurrent user count.

[0097] A single business system can provide multiple services; however, at any given time, only some services may be active. For a given business system, the more services it has active, the greater its load and the more resources it needs to allocate. Therefore, the number of active services can be used as an indicator to build a resource prediction model.

[0098] Business files refer to files generated during the business process of a business system. For example, business files may include images, videos, and log files. The total data volume of business files at the first historical moment can be the total data volume of newly generated business files within a period prior to the first historical moment (which can be called the statistical period). For example, if the first historical moment is the end of a month, the duration of the statistical period can be one month; or, if the first historical moment is the end of a quarter, the statistical period can also be one quarter.

[0099] Based on the aforementioned statistical period, the month-on-month increase in the total data volume of business files can be determined. For example, if the first historical point is the end of the month and the statistical period is one month, then the month-on-month increase in the total data volume of business files newly generated in the latest month can be calculated relative to the total data volume of business files newly generated in the previous month.

[0100] The total data volume of the business system's database at the first historical moment can be defined as the total amount of new data added to the business system's associated databases within a statistical period prior to the first historical moment. Associated databases are the DB (Database) middleware involved in the business system's business processing, including relational databases and non-relational databases.

[0101] Based on the aforementioned statistical period, the month-on-month increase in the total database data volume can be determined. For example, if the first historical point is the end of the month and the statistical period is one month, then the month-on-month increase in the total database data volume can be calculated by comparing the increase in the total database data volume added in the latest month to the increase in the total database data volume added in the previous month.

[0102] Preset resources may include computing resources and / or storage resources. Computing resources may include CPU memory capacity and / or the number of CPU cores; storage resources may be disk space.

[0103] In the embodiments of this application, the preset resources can be one type or multiple types. It is understood that if there are multiple preset resources, then for each type of preset resource, a resource prediction model can be constructed according to the resource demand prediction method provided in this embodiment, and the amount of resources that need to be allocated to the target business system for that type of preset resource can be determined. The processes of constructing the resource prediction models corresponding to each preset resource do not affect each other.

[0104] Regarding step S102, the multiple regression model is a mathematical model that analyzes the relationship between multiple independent variables and one dependent variable, and is suitable for explaining the correlation between variables.

[0105] The multiple regression model is expressed as formula (1):

[0106] (1)

[0107] Indicates the dependent variable. Indicates the first One independent variable, Represents the regression coefficient. This indicates the error term.

[0108] In this embodiment, for each sample business system, the index values ​​of each index in the first index set for that sample business system can be used as independent variables in the above formula, and the corresponding resource size can be used as the dependent variable, forming an equation. Multiple equations can be constructed based on multiple sample business systems. Then, the multiple equations are solved to obtain the regression coefficients and error terms, i.e., to obtain the resource prediction model for predicting resource size. For example, the regression coefficients and error terms can be obtained by solving using the least squares method.

[0109] In one embodiment, the indicators in the first set of indicators can be indicators that have an impact on the size of resources, as determined by technical personnel.

[0110] However, if the first set of indicators is determined solely by technical personnel, some irrelevant indicators may be introduced due to the personal subjective influence of the technical personnel. Therefore, in another embodiment, the first set of indicators can also be determined through the following steps.

[0111] Step 1: For each sample business system, obtain the indicator values ​​of each indicator in the second indicator set for that sample business system at the second historical time, as well as the resource size of the preset resources allocated at the second historical time.

[0112] The first set of indicators belongs to the second set of indicators.

[0113] Step 2: For each indicator in the second indicator set, use the indicator value and corresponding resource size of each indicator in the second historical time to calculate the correlation coefficient, which represents the degree of correlation between the indicator and the preset resource.

[0114] Step 3: Select indicators from the second set of indicators based on the calculated correlation coefficients to obtain the first set of indicators.

[0115] In the embodiments of this application, the second historical moment may be the same as or different from the first historical moment.

[0116] The first set of indicators belongs to the second set of indicators. This means that the second set of indicators, in addition to the indicators in the first set, may also include other indicators. The indicators in the second set can be those determined by technical personnel that influence the size of the resources.

[0117] In step 2 above, the correlation between the indicator and the preset resource is calculated; that is, the correlation between the indicator value and the size of the preset resource. The correlation coefficient can be the Pearson correlation coefficient, or other correlation coefficients such as the Kendall rank correlation coefficient.

[0118] For example, the correlation coefficient falls within the range of [-1, 1]. A correlation coefficient of 1 indicates that the size of the indicator is completely positively correlated with the size of the preset resource; a correlation coefficient of -1 indicates that the size of the indicator is completely negatively correlated with the size of the preset resource; and a correlation coefficient of 0 indicates that the size of the indicator is not linearly correlated with the size of the preset resource.

[0119] Since the correlation coefficient can reflect the degree of correlation between the indicator and the preset resources, a portion of the indicators can be selected from the second set of indicators to obtain the first set of indicators.

[0120] In one implementation, a first indicator can be obtained by directly selecting a pre-set number of indicators with relatively high correlation coefficients from the second indicator set.

[0121] Based on the above processing, indicators that are highly correlated with preset resources can be selected. Subsequently, resource prediction models can be built based on these indicators, which can reduce the amount of data processing and avoid the influence of some irrelevant indicators, thereby improving the accuracy of resource prediction models.

[0122] In one embodiment, indicators whose correlation coefficients satisfy preset screening conditions are selected from the second indicator set to obtain the first indicator set.

[0123] The preset filtering criteria include at least one of the following:

[0124] The correlation coefficients representing the degree of correlation are not less than a preset threshold.

[0125] The correlation coefficient represents the degree of correlation that aligns with the business logic between the indicators and the preset resources;

[0126] There is no multicollinearity between it and other indicators in the second indicator set.

[0127] Since the correlation coefficient calculated in the above manner may be positive or negative, the degree of correlation represented by the correlation coefficient can be expressed by the absolute value of the correlation coefficient. For example, if the correlation coefficient ranges from [-1, 1], then the preset threshold can be 0.2, or it could be 0.3.

[0128] By selecting indicators according to preset thresholds, the first set of indicators consists of those that have a significant impact on resource size, which can further improve the accuracy of resource prediction models and thus improve the accuracy of resource demand prediction.

[0129] It's understandable that there's a business logic behind the relationship between metrics and resource size. For example, from a business logic perspective, a larger "number of activated services" corresponds to a greater resource requirement; that is, the business logic between the "number of activated services" metric and resources indicates a positive correlation between their values. If the correlation coefficient of the "number of activated services" metric is positive, then it can be determined that the correlation coefficient aligns with the business logic. Conversely, if the correlation coefficient of the "number of activated services" metric is negative, then it can be determined that the correlation coefficient aligns with the business logic.

[0130] Furthermore, for any two indicators, if the correlation coefficients representing the degree of correlation between these two indicators do not conform to business logic, it indicates that there exists an indicator among these two indicators whose corresponding correlation coefficients do not conform to business logic. For each of these two indicators (which can be called the indicator to be processed), it can be determined whether the correlation coefficients representing the degree of correlation between this indicator to be processed and other indicators (other indicators besides these two indicators to be processed) conform to business logic. If they do not conform, it can be determined that the correlation coefficient corresponding to this indicator to be processed does not conform to business logic.

[0131] For example, for any two indicators (including indicator A and indicator B), if business logic determines that indicator A has a greater impact on resources than indicator B, and if the correlation coefficient of indicator A is not greater than the correlation coefficient of indicator B, it indicates that the correlation coefficients of indicators A and B do not conform to business logic. Indicators A and B are then treated as indicators to be processed, in order to identify the indicators whose correlation coefficients do not conform to business logic.

[0132] In this way, the resource forecasting model accurately reflects the business logic between indicators and resource size, further improving the accuracy of the resource forecasting model, thereby improving the accuracy of resource demand forecasting.

[0133] Multicollinearity indicates a linear correlation between independent variables, meaning that one independent variable can be a linear combination of one or more other independent variables. If multicollinearity exists, the matrix will not be invertible when calculating the model parameters (including regression coefficients and error terms) of a multiple regression model, which will reduce the effectiveness of the model.

[0134] Therefore, based on the constructed multiple regression model, the VIF (Variance Inflation Factor) of each indicator can be calculated. The VIF of an indicator can quantify the degree of collinearity between that indicator and other indicators; the larger the value, the more severe the collinearity problem.

[0135] Calculate the variance inflation factor for each indicator, and then determine the indicators whose variance inflation factor is less than a preset inflation threshold to obtain the first indicator set.

[0136] Based on the above processing, indicators that exhibit multicollinearity with other indicators can be filtered out, thereby improving the effectiveness of resource prediction models.

[0137] In one embodiment, indicators can be screened according to the above three conditions so that the indicators in the first indicator set simultaneously meet the above three conditions, so as to improve the effectiveness and accuracy of the resource prediction model as much as possible.

[0138] In one embodiment, multiple different multiple regression models can be constructed using combinations of the above-mentioned different screening conditions. Furthermore, measurement indicators (including: (Determination coefficient), RMSE (Root Mean Square Error), and MAE (Mean Absolute Error) are used to determine the resource prediction model used to predict resource size.

[0139] The coefficient of determination is used to evaluate the overall explanatory power of the model, ranging from 0 to 1. The closer the value is to 1, the better the model fits the data. The root mean square error represents the average absolute distance between the predicted value and the true value. The smaller the value, the closer the model's prediction is to the actual situation, and the stronger its predictive power. The mean absolute error represents the average absolute error between the predicted value and the true value. The smaller the value, the closer the model's prediction is to the actual situation, and the stronger its predictive power.

[0140] For example, the index values ​​of each constructed multiple regression model for the measurement index can be calculated separately, and the best-performing multiple regression model can be selected as the resource prediction model using the calculated index values.

[0141] Based on the above processing, the accuracy of resource prediction models can be measured, and the interpretability and stability of prediction results can be improved.

[0142] In one embodiment, to further improve the accuracy of the resource prediction model, at least one of the following preprocessing steps can be performed on the obtained index values ​​and corresponding resource sizes for each index in the first index set: filling in missing values ​​and removing abnormal values.

[0143] Then, using each indicator in the first indicator set as the independent variable and the preset resource as the dependent variable, a multiple regression model is constructed using the preprocessed indicator values ​​and the corresponding resource sizes.

[0144] In this embodiment of the application, since there may be data anomalies and data loss when collecting index values ​​and corresponding resource sizes, after the data is collected, missing data values ​​in the collected data can be filled in and abnormal data values ​​can be removed.

[0145] For example, median imputation can be applied to numerical indicator values, and mode imputation or "unknown" category can be set for categorical indicator values.

[0146] In addition, since the collected data comes from different business systems, and the units of the same data may not be the same in different business systems, the units of the data can be standardized.

[0147] Understandably, categorical indicators (such as system business type or business deployment type) can be encoded to obtain corresponding numerical features (such as 0 / 1) to facilitate the construction of multivariate regression models and preserve the characteristics of the original data as much as possible. For example, OneHot encoding can be used to encode these types of indicator values.

[0148] Regarding step S103, when it is necessary to allocate preset resources to a business system (i.e., the target business system), the indicator values ​​of the target business system for each indicator in the first indicator set can be determined. Then, the obtained indicator values ​​can be input into a pre-built resource prediction model, which is used to predict the resource size required by the business system for the preset resources, thereby obtaining the corresponding resource size. That is, using the indicator values ​​of the target business system for each indicator in the first indicator set as the values ​​of the independent variables in the resource prediction model, the corresponding dependent variable value is calculated as the resource size, which is the preset resource size to be allocated to the target business system. Subsequently, the determined resource size is used as a reference to allocate preset resources to the target business system.

[0149] Based on the above processing, since the resource prediction model can reflect the correlation between the actual state of the business system and the required resource size, the determined resource size is compatible with the state of the target business system. Allocating resources to the target business system according to this resource size can, to a certain extent, avoid resource waste and performance bottlenecks during business processing.

[0150] In addition, the above processing method can complete the prediction calculation in batches in a short time based on the indicator values ​​of the business system, and automatically output the corresponding resource allocation suggestions. This enables rapid and scientific prediction without the need for human experience judgment, which can avoid the lag in resource allocation to a certain extent, improve the level of intelligent and automated management of the business system, reduce operation and maintenance costs, and enhance the stability and reliability of the business system.

[0151] In one embodiment, when actually allocating preset resources to the target business system, fine-tuning may be made based on the determined resource size. Additionally, after preset resources have been allocated to the target business system, the size of the preset resources may also be fine-tuned based on business processing conditions. Subsequently, the adjusted resource size and the target business system's metric values ​​can be used as new training data to construct a new resource prediction model. This model can be adjusted in a timely manner to ensure reliability and accuracy in the actual production environment.

[0152] Based on the same inventive concept, this application also provides a cloud platform operation and maintenance system, which includes a resource demand prediction intelligent agent for executing any of the resource demand prediction methods in the above embodiments.

[0153] An intelligent agent is a proxy capable of perceiving its environment and taking actions to achieve specific goals. It can be software, hardware, or a system, possessing autonomy, adaptability, and interactivity. Intelligent agents perceive changes in the environment (e.g., through sensors or data input), make judgments and decisions based on their learned knowledge and algorithms, and then execute actions to influence the environment or achieve predetermined goals. For example, an autonomous decision-making and execution unit built on a Large Language Model (LLM) for a specific task possesses data understanding, intent recognition, task planning, and result feedback capabilities.

[0154] Based on the system provided in this application embodiment, the resource size obtained by processing the indicator values ​​of the target business system using the resource prediction model is adapted to the actual state of the target business system. Allocating resources to the target business system according to this resource size can avoid resource waste and performance bottlenecks to a certain extent.

[0155] In one embodiment, in addition to the resource demand prediction agent, the cloud platform operation and maintenance system also includes: a work assignment agent, a monitoring sentinel agent, an asset inspection agent, and an alarm diagnosis agent.

[0156] See Figure 2 , Figure 2 This is a schematic diagram of the structure of a cloud platform operation and maintenance system provided in an embodiment of this application. The cloud platform operation and maintenance system includes: a work assignment intelligent agent 110, a monitoring sentinel intelligent agent 120, an asset inspection intelligent agent 130, a resource demand prediction intelligent agent 140, and an alarm diagnosis intelligent agent 160.

[0157] The work assignment agent 110 is used to receive the task information of the current operation and maintenance task input by the user; generate operation and maintenance operation commands based on the task information and send them to the device indicated by the task information for execution, or call the monitoring sentinel agent 120 to execute the operation and maintenance task indicated by the task information, or call the asset inspection agent 130 to execute the operation and maintenance task indicated by the task information.

[0158] The monitoring sentinel agent 120 is used to perform data analysis on the monitoring index data of the monitored equipment indicated by the task information using a large language model, and generate analysis results in natural language form.

[0159] The asset inspection intelligent agent 130 is used to inspect the equipment to be inspected by utilizing the inspection script of the equipment to be inspected as indicated by the task information.

[0160] Alarm diagnosis agent 160 is used to parse and diagnose the received raw alarm data using a large language model to obtain alarm diagnosis results.

[0161] Resource demand prediction agent 140 is used to execute any of the resource demand prediction methods in the above embodiments to determine the amount of resources that need to be allocated to the business system.

[0162] The large language model involved in this application refers to a pre-trained artificial intelligence model with a large number of parameters and strong generalization and reasoning capabilities, used to support the cognitive and decision-making processes of intelligent agents.

[0163] In one embodiment, the alarm diagnosis agent 160 implements intelligent alarm diagnosis through the following steps:

[0164] Step 1: Receive the raw alarm data output by the cloud platform's alarm system;

[0165] like Figure 2 As shown, the alarm diagnostic agent 160 can receive raw alarm data generated by the monitoring tool 170 of the ICOP (Intelligent Cloud Operation Platform) platform in the form of abnormal indicators during the monitoring process.

[0166] Step 2: Based on the alarm diagnosis knowledge base, asset database, and parsing prompts, use a large language model to parse the raw alarm data and obtain the parsing results. The alarm diagnosis knowledge base includes at least: different alarm scenarios and their corresponding troubleshooting work chain templates, a historical fault case library, and maintenance documents for each device in the cloud platform. Among them, the historical fault case library includes at least: the faulty device for each historical case, the alarm scenario of the faulty device, and the root cause of the alarm. The parsing results include at least: the target device that triggered the alarm, the device information of the target device, obtained based on the search results of the asset database, and the target troubleshooting work chain corresponding to the alarm scenario of the target device, obtained based on the search results of the alarm diagnosis knowledge base.

[0167] Step 3: Based on the original alarm data, the search results, parsing results and diagnostic prompts from the alarm diagnosis knowledge base, use the large language model to perform alarm diagnosis and obtain the alarm diagnosis results of each target device output by the large language model. The alarm diagnosis results should include at least the root cause analysis results of the alarm.

[0168] In one embodiment, the work assignment agent 110 is configured to receive a natural language description of the current maintenance task input by the user and a list of target devices corresponding to the current maintenance task selected by the user; obtain the target agent assigned in the description of the current maintenance task; if the target agent is itself, use a large language model to generate one or more maintenance operation commands for each target device in the target device list based on the description of the current maintenance task and the list of target devices; issue one or more maintenance operation commands to the corresponding devices for execution; if the target agent is a monitoring sentinel agent 120, invoke the monitoring sentinel agent 120 to execute the current maintenance task; if the target agent is an asset inspection agent 130, invoke the asset inspection agent 130 to execute the current maintenance task.

[0169] The monitoring sentinel agent 120 is used to acquire one or more monitoring indicator data of the monitored devices, and to perform data analysis on the monitoring data of each monitored device using a large language model to generate analysis results in natural language form; the monitored devices are the devices in the operation and maintenance equipment list, or the monitored devices specified by the user.

[0170] The asset inspection intelligent agent 130 is used to obtain the inspection script corresponding to each device to be inspected; based on each inspection script, it performs inspection on the device to be inspected; the device to be inspected is the device in the maintenance equipment list, or the device in the inspection requirements input by the user through natural language description.

[0171] The cloud platform operation and maintenance system in this embodiment integrates a work assignment intelligent agent 110, a monitoring sentinel intelligent agent 120, an asset inspection intelligent agent 130, a resource demand prediction intelligent agent 140, and an alarm diagnosis intelligent agent 160, realizing intelligent alarm diagnosis, automatic operation and maintenance, and intelligent prediction of resource demand, further improving the integration and intelligence of the cloud platform operation and maintenance system.

[0172] In one embodiment, such as Figure 2 As shown, the cloud platform operation and maintenance system also includes: a human-computer interaction interface 150, a monitoring tool 170, and a unified asset management module 180. For example... Figure 2 As shown, the human-computer interaction interface 150, the monitoring tool 170, and the unified asset management module 180 belong to the ICOP platform and can be implemented based on relevant existing technologies.

[0173] The human-machine interface 150 is used for interaction between the system and the user. Specifically, the work assignment agent 110 can receive task information of the current operation and maintenance task input by the user through the human-machine interface 150. The monitoring sentinel agent 120 can receive monitoring commands triggered by the user and the specified monitored equipment through the human-machine interface 150; the asset inspection agent 130 can receive inspection requirements input by the user in natural language through the human-machine interface 150.

[0174] Monitoring tool 170 obtains monitoring data from device 100 for analysis by monitoring sentinel agent 120. Unified asset management module 180 stores information on all devices in the system and manages all devices, providing information on devices to be inspected to the asset inspection agent 130. The indicator values ​​of each indicator in the indicator set required by resource demand prediction agent 140 can be obtained from the resource information of each asset device in unified asset management module 180.

[0175] In one embodiment, the work assignment agent 110 includes: a task receiving module, a task parsing module, an agent invocation module, a command generation module, and a command issuing module.

[0176] The task receiving module is used to receive the natural language description of the current operation and maintenance task input by the user and the list of target devices corresponding to the current operation and maintenance task selected by the user.

[0177] The task parsing module is used to input the natural language description of the current operation and maintenance task, the target device list, and the preset parsing prompts into the large language model; obtain the parsing results returned by the large language model; the parsing results shall at least include the target agent, task keywords, device type, and device manufacturer.

[0178] The agent invocation module is used to send the parsing results to the monitoring sentinel agent 120 when the target agent is the monitoring sentinel agent, and invoking the monitoring sentinel agent 120 to execute the current operation and maintenance task; when the target agent is the asset inspection agent 130, it sends the parsing results to the asset inspection agent 130, and invoking the asset inspection agent 130 to execute the current operation and maintenance task; and receives the completion status returned by the monitoring sentinel agent 120 or the asset inspection agent 130 after completing the current operation and maintenance task.

[0179] The command generation module is used to match the parsing results with the correspondence between multiple devices and operation and maintenance command templates stored in the knowledge base to obtain the matching operation and maintenance command templates. The parsing results, the knowledge base, and preset search-related prompts are input into the large language model, enabling the large language model to search the knowledge base and return the target prior knowledge corresponding to the current operation and maintenance task. The target prior knowledge includes: historical operation and maintenance tasks and their corresponding historical operation and maintenance commands. The parsing results, target prior knowledge, operation and maintenance command templates, and preset command selection prompts are input into the large language model, obtaining one or more operation and maintenance commands corresponding to the current operation and maintenance task output by the large language model.

[0180] The command issuance module is used to issue one or more operation and maintenance commands to the corresponding target devices to execute the corresponding operation and maintenance operations.

[0181] In one embodiment, the monitoring sentinel agent 120 includes: a data acquisition module, a time-series data preprocessing module, a time-series analysis module, an anomaly analysis module, and a result output module.

[0182] The data acquisition module is used to acquire time-series data of one or more monitoring indicators of the monitored device.

[0183] The data preprocessing module is used to input time series data of different devices and indicator types into the corresponding pre-trained machine learning models. Each machine learning model is used to classify a set of time series data for each indicator under the device type based on the fluctuation characteristics of the time series data segments in each preset sliding window to obtain the fluctuation type of each time series data segment. Based on the fluctuation type, it is determined whether the time series data segment is abnormal.

[0184] The data time series analysis module is used to input the abnormal time series data segment, the fluctuation type of the time series data segment, the relevant equipment labels, and the preset diagnostic prompt words into the big language model for abnormal time series data segments; and obtain the diagnostic results and time series statistical results of the abnormal equipment output by the big language model.

[0185] The anomaly analysis module is used to input abnormal time-series data segments, diagnostic results and time-series statistical results of abnormal equipment, a preset anomaly pattern library and anomaly analysis prompt words into the large language model; enabling the large language model to perform end-to-end anomaly discrimination and semantic attribution, and output anomaly analysis results.

[0186] The results output module is used to output the anomaly analysis results from the large language model to the user in natural language form, and to retrieve the knowledge base and output the historical data and corresponding solutions, or relevant content from the standard operating procedure document to the user.

[0187] In one embodiment, the asset inspection intelligent agent 130 includes: an inspection strategy matching module, an inspection script acquisition module, an inspection impact judgment module, an inspection script execution module, an inspection risk assessment module, and an inspection result archiving module.

[0188] The inspection strategy matching module is used to obtain the equipment function type of the equipment to be inspected corresponding to the current inspection task, and obtain the inspection strategy corresponding to the equipment to be inspected based on the preset correspondence between equipment function type and inspection strategy.

[0189] The inspection script acquisition module is used to obtain inspection scripts for the equipment to be inspected, including: obtaining a first inspection script for the equipment to be inspected; or obtaining a second inspection script uploaded by the user based on the on-site situation; or, inputting the inspection requirements described by the user in natural language, the knowledge base, and the preset inspection script generation prompts into the large language model to obtain a third inspection script generated by the large language model; and binding the obtained inspection scripts with the equipment to be inspected to form one or more inspection task units.

[0190] The inspection impact judgment module is used to input the equipment to be inspected in each inspection unit, the preset asset topology knowledge graph, and the preset asset impact assessment prompts into the big language model. This allows the big language model to query the upstream and downstream equipment of the equipment to be inspected in the asset topology knowledge graph, and evaluate the impact of executing the inspection script in the current inspection unit on business services based on the upstream and downstream equipment of the equipment to be inspected. If the impact reaches a preset level, the module will output recommended adjustment information for the inspection strategy.

[0191] The inspection script execution module is used to inspect the equipment to be inspected based on the inspection strategy corresponding to the current operation and maintenance task or the adjusted inspection strategy, and to obtain the inspection results by using the inspection scripts in each inspection task unit.

[0192] The inspection risk assessment module is used to input the current inspection task, inspection results, knowledge base and preset evaluation prompts into the big language model, and obtain an inspection report output by the big language model that includes inspection conclusions, risk assessment and actionable suggestions.

[0193] The inspection results archiving module is used to store the current inspection tasks and inspection reports in the knowledge base.

[0194] Based on the same inventive concept, this application also provides a resource demand prediction device, see [link to relevant documentation]. Figure 3 , Figure 3 This application provides a schematic diagram of the structure of a resource demand forecasting device; the device includes:

[0195] The data acquisition module 301 is used to acquire, for each of the multiple sample business systems of different types, the indicator values ​​of each indicator in the first indicator set at a first historical moment, and the resource size of the preset resources allocated at the first historical moment; wherein, the first indicator set includes: indicators representing the attribute information of the business system, and / or, indicators representing the operating status of the business system.

[0196] The model building module 302 is used to construct a multiple regression model using each indicator in the first indicator set as the independent variable and the preset resource as the dependent variable, and to obtain a resource prediction model by using the indicator values ​​and corresponding resource sizes of each indicator in the first indicator set at the first historical moment.

[0197] The demand forecasting module 303 is used to process the indicator values ​​of each indicator in the first indicator set for the target business system using the resource forecasting model, so as to obtain the amount of resources that need to be allocated to the target business system for the preset resources.

[0198] Optionally, the device further includes:

[0199] The indicator set determination module is used to obtain, for each sample business system, the indicator value of each indicator in the second indicator set at a second historical time, and the resource size of the preset resources allocated at the second historical time; wherein, the first indicator set belongs to the second indicator set.

[0200] For each indicator in the second indicator set, a correlation coefficient representing the degree of correlation between the indicator and the preset resource is calculated using the indicator value and the corresponding resource size for each indicator in the second indicator set at the second historical time.

[0201] Based on the calculated correlation coefficients, indicators are selected from the second set of indicators to obtain the first set of indicators.

[0202] Optionally, the indicator set determination module is specifically used to select indicators whose correlation coefficients meet preset screening conditions from the second indicator set to obtain the first indicator set;

[0203] The preset filtering conditions include at least one of the following:

[0204] The correlation coefficients representing the degree of correlation are not less than a preset threshold.

[0205] The correlation coefficient represents the degree of correlation that aligns with the business logic between the indicator and the preset resource.

[0206] There is no multicollinearity between it and other indicators in the second indicator set.

[0207] Optionally, the construction module 302 is specifically used to preprocess the indicator values ​​and corresponding resource sizes of each indicator in the first indicator set at the first historical moment.

[0208] The preprocessing includes at least one of the following:

[0209] Fill in the missing values;

[0210] Remove outlier values;

[0211] Using each indicator in the first indicator set as the independent variable and the preset resource as the dependent variable, a multiple regression model is constructed using the preprocessed indicator values ​​and the corresponding resource size.

[0212] Optionally, the indicators characterizing the attribute information of the business system include at least one of the following: system business type, business deployment type, presence of DMZ zone service and service type; the indicators characterizing the operating status of the business system include at least one of the following: total number of registered users, number of daily active users, peak concurrent users, number of services enabled, total data volume of business files, month-on-month increase in total data volume of business files, total data volume of the database, and month-on-month increase in total data volume of the database.

[0213] The default resources represent CPU memory capacity and / or the number of CPU cores.

[0214] Optionally, the device further includes:

[0215] The module update module is used to update the resource prediction model by using the indicator values ​​of each indicator in the first indicator set of the target business system and the actual resource size of the preset resources allocated to the target business system.

[0216] This application also provides an electronic device, such as... Figure 4 As shown, it includes a processor 401, a communication interface 402, a memory 403, and a communication bus 404, wherein the processor 401, the communication interface 402, and the memory 403 communicate with each other through the communication bus 404.

[0217] Memory 403 is used to store computer programs;

[0218] When processor 401 executes the program stored in memory 403, it performs the following steps:

[0219] For each of the multiple sample business systems of different types, obtain the index values ​​of each index in the first index set for the sample business system at the first historical moment, as well as the resource size of the preset resources allocated at the first historical moment; wherein, the first index set includes: indexes representing the attribute information of the business system, and / or, indexes representing the operating status of the business system.

[0220] Using each indicator in the first indicator set as the independent variable and the preset resource as the dependent variable, a multiple regression model is constructed using the indicator values ​​and corresponding resource sizes of each indicator in the first indicator set at the first historical moment to obtain a resource prediction model.

[0221] The resource prediction model is used to process the indicator values ​​of each indicator in the first indicator set for the target business system to obtain the amount of resources that need to be allocated to the target business system for the preset resources.

[0222] The communication bus mentioned in the above electronic devices can be a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus, etc. This communication bus can be divided into address bus, data bus, control bus, etc. For ease of illustration, only one thick line is used to represent it in the diagram, but this does not mean that there is only one bus or one type of bus.

[0223] The communication interface is used for communication between the aforementioned electronic devices and other devices.

[0224] The memory may include random access memory (RAM) or non-volatile memory (NVM), such as at least one disk storage device. Optionally, the memory may also be at least one storage device located remotely from the aforementioned processor.

[0225] The processors mentioned above can be general-purpose processors, including central processing units (CPUs), network processors (NPs), etc.; they can also be digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components.

[0226] In another embodiment provided in this application, a computer-readable storage medium is also provided, which stores a computer program that, when executed by a processor, implements the steps of any of the above-described resource demand prediction methods.

[0227] In another embodiment provided in this application, a computer program product containing instructions is also provided, which, when run on a computer, causes the computer to execute any of the resource demand prediction methods described in the above embodiments.

[0228] In the above embodiments, implementation can be achieved entirely or partially through software, hardware, firmware, or any combination thereof. When implemented using software, it can be implemented entirely or partially as a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, all or part of the processes or functions described in the embodiments of this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., coaxial cable, fiber optic, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that integrates one or more available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium (e.g., solid state disk (SSD)).

[0229] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0230] The various embodiments in this specification are described in a related manner. Similar or identical parts between embodiments can be referred to mutually. Each embodiment focuses on describing the differences from other embodiments. In particular, the embodiments of apparatus, systems, electronic devices, computer-readable storage media, and computer program products are basically similar to the method embodiments, and therefore the descriptions are relatively simple; relevant parts can be referred to the descriptions of the method embodiments.

[0231] The above description is merely a preferred embodiment of this application and is not intended to limit the scope of protection of this application. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application are included within the scope of protection of this application.

Claims

1. A resource demand forecasting method, characterized in that, The method includes: For each of the multiple sample business systems of different types, obtain the index values ​​of each index in the first index set for the sample business system at the first historical moment, as well as the resource size of the preset resources allocated at the first historical moment; wherein, the first index set includes: indexes representing the attribute information of the business system, and / or, indexes representing the operating status of the business system. Using each indicator in the first indicator set as the independent variable and the preset resource as the dependent variable, a multiple regression model is constructed using the indicator values ​​and corresponding resource sizes of each indicator in the first indicator set at the first historical moment to obtain a resource prediction model. The resource prediction model is used to process the indicator values ​​of each indicator in the first indicator set for the target business system to obtain the amount of resources that need to be allocated to the target business system for the preset resources.

2. The method according to claim 1, characterized in that, The indicators in the first indicator set are determined through the following steps: For each sample business system, obtain the indicator values ​​of each indicator in the second indicator set for the sample business system at the second historical time, as well as the resource size of the preset resources allocated at the second historical time; wherein, the first indicator set belongs to the second indicator set; For each indicator in the second indicator set, a correlation coefficient representing the degree of correlation between the indicator and the preset resource is calculated using the indicator value and the corresponding resource size for each indicator in the second indicator set at the second historical time. Based on the calculated correlation coefficients, indicators are selected from the second set of indicators to obtain the first set of indicators.

3. The method according to claim 2, characterized in that, The step of selecting indicators from the second indicator set based on the calculated correlation coefficient to obtain the first indicator set includes: The first indicator set is obtained by selecting indicators whose correlation coefficients meet the preset screening conditions from the second indicator set; The preset filtering conditions include at least one of the following: The correlation coefficients representing the degree of correlation are not less than a preset threshold. The correlation coefficient represents the degree of correlation that aligns with the business logic between the indicator and the preset resource. There is no multicollinearity between it and other indicators in the second indicator set.

4. The method according to claim 1, characterized in that, Using each indicator in the first indicator set as the independent variable and the preset resource as the dependent variable, a multiple regression model is constructed using the indicator values ​​and corresponding resource sizes of each indicator in the first indicator set at a first historical moment, including: Preprocess the indicator values ​​and corresponding resource sizes of each indicator in the first indicator set at the first historical moment; The preprocessing includes at least one of the following: Fill in the missing values; Remove outlier values; Using each indicator in the first indicator set as the independent variable and the preset resource as the dependent variable, a multiple regression model is constructed using the preprocessed indicator values ​​and the corresponding resource size.

5. The method according to claim 1, characterized in that, The indicators characterizing the attribute information of the business system include at least one of the following: system business type, business deployment type, whether there is an isolated service and the service type; the indicators characterizing the operating status of the business system include at least one of the following: total number of registered users, number of daily active users, peak concurrent users, number of services that have been started, total data volume of business files, month-on-month increase in total data volume of business files, total data volume of the database, and month-on-month increase in total data volume of the database. The default resources represent CPU memory capacity and / or the number of CPU cores.

6. The method according to any one of claims 1-5, characterized in that, The method further includes: The resource prediction model is updated using the indicator values ​​of each indicator in the first indicator set for the target business system, as well as the actual resource size of the preset resources allocated to the target business system.

7. A cloud platform operation and maintenance system, characterized in that, include: Resource demand prediction agent; The resource demand prediction agent is used to perform the method according to any one of claims 1-6.

8. The system according to claim 7, characterized in that, The system also includes: Work assignment agent, monitoring sentinel agent, asset inspection agent, and alarm diagnosis agent; The work assignment agent is used to receive task information of the current operation and maintenance task input by the user; generate operation and maintenance operation commands based on the task information and send them to the device indicated by the task information for execution, or call the monitoring sentinel agent to execute the operation and maintenance task indicated by the task information, or call the asset inspection agent to execute the operation and maintenance task indicated by the task information. The monitoring sentinel agent is used to perform data analysis on the monitoring index data of the monitored equipment indicated by the task information using a large language model, and generate analysis results in natural language form. The asset inspection intelligent agent is used to inspect the equipment to be inspected using the inspection script of the equipment to be inspected indicated by the task information. The alarm diagnosis agent is used to parse and diagnose the received raw alarm data using a large language model to obtain alarm diagnosis results.

9. A resource demand forecasting device, characterized in that, The device includes: The data acquisition module is used to acquire, for each of the multiple sample business systems of different types, the indicator values ​​of each indicator in the first indicator set at a first historical moment, and the resource size of the preset resources allocated at the first historical moment; wherein, the first indicator set includes: indicators representing the attribute information of the business system, and / or, indicators representing the operating status of the business system. The model building module is used to construct a multiple regression model using each indicator in the first indicator set as the independent variable and the preset resource as the dependent variable, and to construct a resource prediction model by using the indicator values ​​and corresponding resource sizes of each indicator in the first indicator set at the first historical moment. The demand forecasting module is used to process the indicator values ​​of each indicator in the first indicator set for the target business system using the resource forecasting model, so as to obtain the amount of resources that need to be allocated to the target business system for the preset resources.

10. An electronic device, characterized in that, It includes a processor, a communication interface, a memory, and a communication bus, wherein the processor, the communication interface, and the memory communicate with each other through the communication bus; Memory, used to store computer programs; A processor, when executing a program stored in memory, implements the steps of the method described in any one of claims 1-6.

11. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, implements the steps of the method described in any one of claims 1-6.

Images