A cost optimization method, system, device and product based on cloud resource configuration

Abstract

The present invention belongs to the field of cloud computing technology, and its purpose is to provide a cost optimization method, system, device and product based on cloud resource configuration. The method includes: obtaining pre-processed cloud resource usage data of each cloud service instance, and performing cluster analysis on it to obtain k Clustering results are generated; regression analysis is performed on the pre-processed cloud resource usage data in each clustering result to obtain a resource usage forecast corresponding to each clustering result; time series forecasting is performed on the resource usage forecast corresponding to each clustering result to obtain a resource usage forecast corresponding to each clustering result in a specified future time period; and cloud resource allocation is performed for the cloud service instances in each clustering result based on the resource usage forecast corresponding to each clustering result in the specified future time period. The present invention can realize intelligent and automated management of cloud resources and ensure a balance between resource utilization and cost.

Description

Technical Field

[0001] The present invention belongs to the technical field of cloud computing, and specifically relates to a cost optimization method, system, device and product based on cloud resource configuration. Background Art

[0002] With the rapid development of cloud computing technology, more and more enterprises and organizations rely on cloud services to support their core businesses, enabling flexible resource management and efficient operations. Currently, cloud service providers typically use static and manual resource configuration methods to meet user needs. These methods often rely on experience and pre-defined rules or policies to allocate resources according to the user's initial needs when the cloud service instance is created. Manual configuration further requires administrators to continuously monitor and adjust.

[0003] However, in the process of using the existing technology, the inventors found that the existing technology has at least the following problems:

[0004] Workloads in cloud environments are often highly dynamic and uncertain, especially under bursty traffic. Existing static and manual resource allocation methods often fail to respond promptly to changes in resource demand. Over-allocation of resources can lead to idle resources and low resource utilization, which in turn increases unnecessary operating costs for the enterprise. Reducing resource allocation to reduce costs can lead to resource shortages for critical applications during peak periods, impacting service performance and user experience. Therefore, ensuring high resource utilization while minimizing costs remains a challenge that current technologies struggle to effectively address. Summary of the Invention

[0005] The present invention aims to solve the above technical problems at least to a certain extent, and provides a cost optimization method, system, device and product based on cloud resource configuration.

[0006] In order to achieve the above object, the present invention adopts the following technical solutions:

[0007] In a first aspect, the present invention provides a cost optimization method based on cloud resource configuration, comprising:

[0008] Obtaining cloud resource usage data of each cloud service instance, and preprocessing the cloud resource usage data of each cloud service instance to obtain preprocessed cloud resource usage data of each cloud service instance;

[0009] Perform cluster analysis on the pre-processed cloud resource usage data of all cloud service instances to obtain k clustering results; where k is a natural number greater than 1, and any clustering result includes the pre-processed cloud resource usage data of one or more cloud service instances, and each clustering result is bound to a different classification label;

[0010] Perform regression analysis on the pre-processed cloud resource usage data in each clustering result to obtain the resource usage prediction value corresponding to each clustering result;

[0011] Perform time series forecasting on the resource usage prediction values ​​corresponding to each clustering result to obtain the resource usage prediction values ​​corresponding to each clustering result in a specified future time period;

[0012] Based on the resource usage prediction value corresponding to each clustering result within a specified time period in the future, cloud resource configuration is performed for the cloud service instances in each clustering result.

[0013] In one possible design, cluster analysis is performed on the pre-processed cloud resource usage data of all cloud service instances to obtain k clustering results, including:

[0014] Obtaining the number k of clusters of the preprocessed cloud resource usage data;

[0015] Initialize k cluster centers, calculate the Euclidean distance of each cloud service instance's pre-processed cloud resource usage data to each cluster center, and assign the pre-processed cloud resource usage data of each cloud service instance to the cluster with the closest Euclidean distance to the cluster center;

[0016] Calculate the mean of the pre-processed cloud resource usage data of all cloud service instances in the cluster of each cluster center, and use the mean corresponding to each cluster center to update each cluster center to obtain k updated cluster centers;

[0017] Each cluster center is set as the corresponding updated cluster center, and the Euclidean distance of the pre-processed cloud resource usage data of each cloud service instance to the k cluster centers is recalculated until the mean change of the pre-processed cloud resource usage data of all cloud service instances in the cluster cluster of each cluster center is less than the threshold. At this time, the cluster clusters corresponding to the k cluster centers constitute k clustering results.

[0018] In one possible design, the pre-processed cloud resource usage data x of the i-th cloud service instance i To the jth cluster center μ j The Euclidean distance is:

[0019]

[0020] Among them, x im Represents the pre-processed cloud resource usage data x of the cloud service instance i The mth data point in ; M is the pre-processed cloud resource usage data x of the cloud service instance i The total number of data points in .

[0021] In one possible design, regression analysis is performed on the pre-processed cloud resource usage data in any clustering result to obtain a resource usage prediction value corresponding to the clustering result, including:

[0022] Building an initial regression model and obtaining historical cloud resource usage data that matches a classification label of any of the clustering results;

[0023] Training the initial regression model based on the historical cloud resource usage data to obtain a trained regression model;

[0024] The pre-processed cloud resource usage data in any clustering result is input into the trained regression model to obtain a resource usage prediction value corresponding to any clustering result.

[0025] In one possible design, the trained regression model is:

[0026] y=β0+β1x1+β2x2+…+β n x n +∈;

[0027] Among them, y is the target variable; x1, x2, ... and x n is the independent variable; ∈ is the preset error value; β1, β2, ... and β n is the regression coefficient.

[0028] In one possible design, a time series forecasting process is performed on the resource usage forecast value corresponding to any clustering result to obtain the resource usage forecast value corresponding to any clustering result within a specified future time period, including:

[0029] Building an initial time series model and obtaining historical cloud resource usage data that matches a classification label of any of the clustering results;

[0030] Training the initial time series model based on the historical cloud resource usage data to obtain a trained time series model;

[0031] The resource usage prediction value corresponding to any clustering result is input into the trained time series model to obtain the resource usage prediction value corresponding to any clustering result in a future specified time period t.

[0032] In one possible design, the trained time series model is:

[0033] y t =φ1y t-1 +φ2y t-2 +...+φ p y t-p +θ1∈t-1 +θ2∈ t-2 +...+θ q ∈ t-q +∈ t ;

[0034] Among them, y t y is the predicted value of resource usage corresponding to any clustering result in the future specified time period t; t-1 、y t-2 , ... and y t-p is the resource usage prediction value corresponding to any clustering result in the time periods t-1, t-2, ..., and tp; and are the autoregressive coefficients; θ1, θ2, … and θ q is the moving average coefficient; p is the autoregressive order; q is the moving average order; ∈ t-1 ,∈ t-2 ,…,∈ t-q and ∈ t is the random error in the time periods t-1, t-2, …, tq and t.

[0035] In a second aspect, the present invention provides a cost optimization system based on cloud resource configuration, for implementing any one of the above-mentioned cost optimization methods based on cloud resource configuration; the cost optimization system based on cloud resource configuration comprises:

[0036] An initial data acquisition module is used to acquire cloud resource usage data of each cloud service instance and preprocess the cloud resource usage data of each cloud service instance to obtain preprocessed cloud resource usage data of each cloud service instance;

[0037] a clustering analysis module, communicatively connected to the initial data acquisition module, configured to perform cluster analysis on the pre-processed cloud resource usage data of all cloud service instances to obtain k clustering results; wherein k is a natural number greater than 1, and any clustering result includes the pre-processed cloud resource usage data of one or more cloud service instances, and each clustering result is associated with a different classification label;

[0038] A regression analysis module, in communication with the cluster analysis module, for performing regression analysis on the pre-processed cloud resource usage data in each clustering result to obtain a resource usage prediction value corresponding to each clustering result;

[0039] A time series prediction module, in communication with the regression analysis module, is used to perform time series prediction processing on the resource usage prediction values ​​corresponding to each clustering result to obtain the resource usage prediction values ​​corresponding to each clustering result in a specified future time period;

[0040] The cloud resource configuration module is in communication with the time series prediction module and is used to configure cloud resources for the cloud service instances in each clustering result based on the resource usage prediction value corresponding to each clustering result in a specified future time period.

[0041] In a third aspect, the present invention provides an electronic device, comprising:

[0042] a memory for storing computer program instructions; and

[0043] A processor is configured to execute the computer program instructions to complete the operation of the cost optimization method based on cloud resource configuration as described in any one of the above.

[0044] In a fourth aspect, the present invention provides a computer program product, comprising a computer program or instructions, which, when executed by a computer, implements the cost optimization method based on cloud resource configuration as described in any one of the above.

[0045] The beneficial effects of the present invention are:

[0046] The present invention discloses a cost optimization method, system, device and product based on cloud resource configuration, which can realize intelligent and automated management of cloud resources, optimize cloud resource configuration costs during cloud resource configuration, and ensure the balance between resource utilization and cost. Specifically, during the implementation of the present invention, first, cloud resource usage data of each cloud service instance is obtained, and the cloud resource usage data of each cloud service instance is preprocessed to obtain the preprocessed cloud resource usage data of each cloud service instance; then, cluster analysis is performed on the preprocessed cloud resource usage data of all cloud service instances to obtain k clustering results; then, regression analysis is performed on the preprocessed cloud resource usage data in each clustering result to obtain resource usage prediction values ​​corresponding to each clustering result; then, time series prediction is performed on the resource usage prediction values ​​corresponding to each clustering result to obtain resource usage prediction values ​​corresponding to each clustering result in a specified future time period; finally, cloud resource configuration is performed on the cloud service instances in each clustering result based on the resource usage prediction values ​​corresponding to each clustering result in the specified future time period. In this process, through cluster analysis processing, cloud resource usage data can be divided into multiple categories, and instances with similar resource usage patterns can be classified into the same category, thereby forming different cloud service instance classification sets, which can provide more detailed input data for subsequent cloud resource demand forecasting and other analyses. Regression analysis processing can generate preliminary resource usage forecast data, and time series forecast processing can generate specific time series forecast results for future time periods, which can effectively deal with the uncertainty of workloads and ensure that the final resource allocation strategy can accurately match dynamically changing needs, thereby optimizing resource utilization and cost, and ensuring the optimal balance between resource utilization and cost. Based on the present invention, it is not only possible to improve resource utilization efficiency, reduce operating costs, maximize resource utilization efficiency and minimize costs, but also improve system stability and user experience, and meet the complex needs of modern cloud computing environments.

[0047] Other beneficial effects of the present invention will be further described in the specific embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] Figure 1 is a flow chart of the cost optimization method based on cloud resource configuration in Example 1;

[0049] Figure 2 This is a module block diagram of the cost optimization system based on cloud resource configuration in Example 2;

[0050] Figure 3 This is a module block diagram of the electronic device in Example 3. DETAILED DESCRIPTION

[0051] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the present invention will be briefly introduced below in conjunction with the drawings and the description of the embodiments or the prior art. Obviously, the following description of the structure of the drawings is only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative work. It should be noted that the description of these embodiments is used to help understand the present invention, but does not constitute a limitation of the present invention.

[0052] Example 1:

[0053] This embodiment discloses a cost optimization method based on cloud resource configuration, which can be executed by, but is not limited to, a computer device or a virtual machine with certain computing resources, such as a personal computer, a smart phone, a personal digital assistant, or a wearable device, or by a virtual machine.

[0054] like Figure 1 As shown, a cost optimization method based on cloud resource configuration may include, but is not limited to, the following steps:

[0055] S1. Obtain the cloud resource usage data of each cloud service instance, and pre-process the cloud resource usage data of each cloud service instance respectively to obtain the pre-processed cloud resource usage data of each cloud service instance; it should be noted that a cloud service instance is an independent computing resource unit or service instance provided to users on a cloud computing platform, which can be a virtual machine instance, a container instance or a database instance, etc., and is usually used to provide an isolated computing environment for different users or different applications, and therefore can be used to distinguish different users or different applications.

[0056] Specifically, in this embodiment, the cloud resource usage data includes usage data, performance indicator data, and cost information of the corresponding cloud service instance. Among them, the usage data is indicator data that directly reflects resource consumption and use, including but not limited to CPU (Central Processing Unit) usage rate, memory occupancy rate, network traffic and disk I / O (Input / Output) volume, etc., which can be used to measure the resource usage of the cloud service instance and is an important reference for cloud resource allocation and optimization in this embodiment. The performance indicator data is various indicators that reflect the performance of the cloud service instance, including but not limited to latency, throughput, response time and error rate, etc., which can be used to further evaluate the operating status and service quality of the system or application.

[0057] In this embodiment, the cloud resource usage data is preprocessed, such as performing data cleaning, outlier removal, invalid data screening, standardization, normalization, and time series resampling on the cloud resource usage data. Standardization and normalization can be used to eliminate the impact of different data scales. The preprocessing process can ensure the accuracy and effectiveness of subsequent data analysis.

[0058] S2. Perform cluster analysis on the pre-processed cloud resource usage data of all cloud service instances to obtain k clustering results; where k is a natural number greater than 1, any clustering result includes the pre-processed cloud resource usage data of one or more cloud service instances, and any clustering result includes the cluster cluster of the current cluster center. Each clustering result is bound to a different classification label, and the classification label is used to identify which clustering result each cloud service instance belongs to.

[0059] Specifically, in step S2, cluster analysis is performed on the pre-processed cloud resource usage data of all cloud service instances to obtain k clustering results, including:

[0060] S201. Obtain the number k of clusters of the pre-processed cloud resource usage data. Specifically, in this embodiment, the appropriate number of clusters of the pre-processed cloud resource usage data of all cloud service instances can be determined by the elbow method or the silhouette coefficient method, which is not limited here.

[0061] S202. Initialize k cluster centers, calculate the Euclidean distance of each cloud service instance's pre-processed cloud resource usage data to each cluster center, and assign the pre-processed cloud resource usage data of each cloud service instance to the cluster with the closest Euclidean distance to the cluster center;

[0062] In step S202 of this embodiment, the pre-processed cloud resource usage data x of the i-th cloud service instance i To the jth cluster center μ j The Euclidean distance is:

[0063]

[0064] Among them, x im Represents the pre-processed cloud resource usage data x of the cloud service instance i The mth data point in ; M is the pre-processed cloud resource usage data x of the cloud service instance i The total number of data points in ; it should be understood that j≤k.

[0065] S203. Calculate the mean of the pre-processed cloud resource usage data of all cloud service instances in the cluster of each cluster center, and use the mean corresponding to each cluster center to update each cluster center to obtain k updated cluster centers;

[0066] S204. Set each cluster center as the updated cluster center corresponding to it, and recalculate the Euclidean distance of the pre-processed cloud resource usage data of each cloud service instance to the k cluster centers, that is, return to step S202, until the mean change of the pre-processed cloud resource usage data of all cloud service instances in the cluster cluster of each cluster center is less than the threshold, that is, until the cluster center no longer changes significantly, at which time the cluster clusters corresponding to the k cluster centers constitute k clustering results.

[0067] S3. Perform regression analysis on the pre-processed cloud resource usage data in each clustering result to obtain the resource usage prediction value corresponding to each clustering result;

[0068] Specifically, in step S3, regression analysis is performed on the pre-processed cloud resource usage data in any clustering result to obtain a resource usage prediction value corresponding to any clustering result, including:

[0069] S301. Build an initial regression model and obtain historical cloud resource usage data that matches the classification label of any clustering result;

[0070] S302. Based on the historical cloud resource usage data, the initial regression model is trained to obtain a trained regression model;

[0071] In step S302 of this embodiment, the trained regression model is:

[0072] y=β0+β1x1+β2x2+…+β n x n +∈;

[0073] Where y is the target variable, which is the resource usage in this embodiment; x1, x2, ... and x n is the independent variable, which is time in this embodiment; ∈ is the preset error value; β1, β2, ... and β n is the regression coefficient.

[0074] Specifically, in this embodiment, the regression coefficient is obtained by using the least squares method or the gradient descent method, and the solution formula is:

[0075]

[0076] Wherein, X is the independent variable matrix in the historical cloud resource usage data, which is the time matrix in this embodiment; y′ is the target variable vector in the historical cloud resource usage data, which is the resource usage in this embodiment.

[0077] In addition, in this embodiment, during the training of the initial regression model, the R2 The fitting effect of the model is evaluated by indicators such as the mean square error and the mean square error, and the trained regression model is obtained when the preset conditions are met.

[0078] S303. Input the pre-processed cloud resource usage data in any clustering result into the trained regression model to obtain a resource usage prediction value corresponding to any clustering result.

[0079] S4. Perform time series forecasting on the resource usage prediction values ​​corresponding to each clustering result to obtain the resource usage prediction values ​​corresponding to each clustering result for a specified time period in the future;

[0080] Specifically, in step S4, a time series prediction process is performed on the resource usage prediction value corresponding to any clustering result to obtain a resource usage prediction value corresponding to any clustering result within a specified time period in the future, including:

[0081] S401. Build an initial time series model and obtain historical cloud resource usage data that matches the classification label of any clustering result;

[0082] S402. Based on the historical cloud resource usage data, the initial time series model is trained to obtain a trained time series model;

[0083] In this embodiment, the trained time series model adopts an ARIMA (Autoregressive Integrated Moving Average) model. Specifically, the trained time series model is:

[0084] y t =φ1y t-1 +φ2y t-2 +...+φ p y t-p +θ1∈ t-1 +θ2∈ t-2 +...+θ q ∈ t-q +∈ t ;

[0085] Among them, y t y is the predicted value of resource usage corresponding to any clustering result in the future specified time period t; t-1 、y t-2 , ... and y t-p is the resource usage prediction value corresponding to any clustering result in the time periods t-1, t-2, ..., and tp; and are the autoregressive coefficients; θ1, θ2, … and θ qis the moving average coefficient; p is the autoregressive order; q is the moving average order; ∈ t-1 ,∈ t-2 ,…,∈ t-q and ∈ t is the random error in the time periods t-1, t-2, …, tq and t.

[0086] Specifically, in this embodiment, the moving average coefficient and the autoregressive coefficient are estimated by maximum likelihood estimation or Bayesian method.

[0087] S403. Input the resource usage prediction value corresponding to any clustering result into the trained time series model to obtain the resource usage prediction value corresponding to any clustering result in a future specified time period t.

[0088] S5. Based on the predicted resource usage value corresponding to each clustering result within a specified future time period, perform cloud resource configuration for the cloud service instances in each clustering result.

[0089] This embodiment can realize intelligent and automated management of cloud resources, optimize cloud resource configuration costs during cloud resource configuration, and ensure a balance between resource utilization and costs. Specifically, during the implementation of this embodiment, first, cloud resource usage data of each cloud service instance is obtained, and the cloud resource usage data of each cloud service instance is preprocessed to obtain preprocessed cloud resource usage data of each cloud service instance; then, cluster analysis is performed on the preprocessed cloud resource usage data of all cloud service instances to obtain k clustering results; then, regression analysis is performed on the preprocessed cloud resource usage data in each clustering result to obtain resource usage prediction values ​​corresponding to each clustering result; then, time series prediction is performed on the resource usage prediction values ​​corresponding to each clustering result to obtain resource usage prediction values ​​corresponding to each clustering result in a specified future time period; finally, cloud resource configuration is performed on the cloud service instances in each clustering result based on the resource usage prediction values ​​corresponding to each clustering result in the specified future time period. In this process, through cluster analysis processing, cloud resource usage data can be divided into multiple categories, and instances with similar resource usage patterns can be classified into the same category, thereby forming different cloud service instance classification sets to provide more detailed input data for subsequent cloud resource demand forecasting and other analyses. Regression analysis processing can generate preliminary resource usage forecast data, and time series forecast processing can generate specific time series forecast results for future time periods, which can help to effectively deal with the uncertainty of workloads and ensure that the final resource allocation strategy can accurately match dynamically changing needs, thereby optimizing resource utilization and cost, and ensuring the optimal balance between resource utilization and cost. Based on this embodiment, it is not only possible to improve resource utilization efficiency, reduce operating costs, maximize resource utilization efficiency and minimize costs, but also improve system stability and user experience, and meet the complex needs of modern cloud computing environments.

[0090] Example 2:

[0091] This embodiment discloses a cost optimization system based on cloud resource configuration, which is used to implement the cost optimization method based on cloud resource configuration in Example 1; Figure 2 As shown, the cost optimization system based on cloud resource configuration includes:

[0092] An initial data acquisition module is used to acquire cloud resource usage data of each cloud service instance and preprocess the cloud resource usage data of each cloud service instance to obtain preprocessed cloud resource usage data of each cloud service instance;

[0093] a clustering analysis module, communicatively connected to the initial data acquisition module, configured to perform cluster analysis on the pre-processed cloud resource usage data of all cloud service instances to obtain k clustering results; wherein k is a natural number greater than 1, and any clustering result includes the pre-processed cloud resource usage data of one or more cloud service instances, and each clustering result is associated with a different classification label;

[0094] A regression analysis module, in communication with the cluster analysis module, for performing regression analysis on the pre-processed cloud resource usage data in each clustering result to obtain a resource usage prediction value corresponding to each clustering result;

[0095] A time series prediction module, in communication with the regression analysis module, is used to perform time series prediction processing on the resource usage prediction values ​​corresponding to each clustering result to obtain the resource usage prediction values ​​corresponding to each clustering result in a specified future time period;

[0096] The cloud resource configuration module is in communication with the time series prediction module and is used to configure cloud resources for the cloud service instances in each clustering result based on the resource usage prediction value corresponding to each clustering result in a specified future time period.

[0097] It should be noted that the working process, working details and technical effects of the cost optimization system based on cloud resource configuration provided in this embodiment 2 can be found in embodiment 1 and will not be repeated here.

[0098] Example 3:

[0099] Based on the embodiment 1 or 2, this embodiment discloses an electronic device, which may be a smart phone, a tablet computer, a laptop computer or a desktop computer. The electronic device may be called a user terminal, a portable terminal, a desktop terminal, etc. Figure 3 As shown, the electronic equipment includes:

[0100] a memory for storing computer program instructions; and

[0101] A processor is configured to execute the computer program instructions to complete the operations of the cost optimization method based on cloud resource configuration as described in any one of the embodiments 1.

[0102] Specifically, the processor 301 may include one or more processing cores, such as a 4-core processor, an 8-core processor, etc. The processor 301 may be implemented in at least one hardware form of DSP (Digital Signal Processing), FPGA (Field-Programmable Gate Array), and PLA (Programmable Logic Array). The processor 301 may also include a main processor and a coprocessor. The main processor is a processor for processing data in the awake state, also known as a CPU (Central Processing Unit); the coprocessor is a low-power processor for processing data in the standby state. In some embodiments, the processor 301 may be integrated with a GPU (Graphics Processing Unit), which is responsible for rendering and drawing the content to be displayed on the display screen.

[0103] The memory 302 may include one or more computer-readable storage media, which may be non-transitory. The memory 302 may also include high-speed random access memory and non-volatile memory, such as one or more disk storage devices and flash memory storage devices. In some embodiments, the non-transitory computer-readable storage medium in the memory 302 is used to store at least one instruction, which is executed by the processor 301 to implement the cost optimization method based on cloud resource configuration provided in Example 1 of the present application.

[0104] In some embodiments, the terminal may optionally include a communication interface 303 and at least one peripheral device. The processor 301, memory 302, and communication interface 303 may be connected via a bus or signal lines. Each peripheral device may be connected to the communication interface 303 via a bus, signal lines, or circuit boards. Specifically, the peripheral device may include at least one of a radio frequency circuit 304, a display screen 305, and a power supply 306.

[0105] The communication interface 303 can be used to connect at least one I / O (Input / Output)-related peripheral device to the processor 301 and the memory 302. In some embodiments, the processor 301, the memory 302, and the communication interface 303 are integrated on the same chip or circuit board; in some other embodiments, any one or two of the processor 301, the memory 302, and the communication interface 303 can be implemented on separate chips or circuit boards, which is not limited in this embodiment.

[0106] The radio frequency circuit 304 is used to receive and transmit RF (Radio Frequency) signals, also known as electromagnetic signals. The radio frequency circuit 304 communicates with a communication network and other communication devices via electromagnetic signals.

[0107] The display screen 305 is used to display a UI (User Interface). The UI may include graphics, text, icons, videos, or any combination thereof.

[0108] The power supply 306 is used to supply power to various components in the electronic device.

[0109] Example 4:

[0110] Based on any one of Examples 1 to 3, this embodiment discloses a computer program product, including a computer program or instructions, which, when executed by a computer, implements the cost optimization method based on cloud resource configuration as described in any one of Example 1.

[0111] Obviously, those skilled in the art will appreciate that the modules or steps of the present invention described above can be implemented using a general-purpose computing device. They can be centralized on a single computing device or distributed across a network of multiple computing devices. Alternatively, they can be implemented using program code executable by a computing device, which can then be stored in a storage device and executed by the computing device. Alternatively, they can be fabricated into separate integrated circuit modules, or multiple modules or steps can be fabricated into a single integrated circuit module for implementation. Thus, the present invention is not limited to any specific combination of hardware and software.

[0112] Finally, it should be noted that the above embodiments are intended only to illustrate the technical solutions of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art will appreciate that modifications may be made to the technical solutions described in the above embodiments, or that some of the technical features may be replaced with equivalents. Such modifications or replacements do not deviate from the spirit and scope of the technical solutions of the various embodiments of the present invention.

Claims

1. A cost optimization method based on cloud resource configuration, characterized in that: include: Obtaining cloud resource usage data of each cloud service instance, and preprocessing the cloud resource usage data of each cloud service instance to obtain preprocessed cloud resource usage data of each cloud service instance; Perform cluster analysis on the pre-processed cloud resource usage data of all cloud service instances to obtain k clustering results; where k is a natural number greater than 1, and any clustering result includes the pre-processed cloud resource usage data of one or more cloud service instances, and each clustering result is bound to a different classification label; Perform regression analysis on the pre-processed cloud resource usage data in each clustering result to obtain the resource usage prediction value corresponding to each clustering result; Perform time series forecasting on the resource usage prediction values ​​corresponding to each clustering result to obtain the resource usage prediction values ​​corresponding to each clustering result in a specified future time period; Based on the predicted resource usage values ​​corresponding to each clustering result within a specified future time period, cloud resource allocation is performed for the cloud service instances in each clustering result; Performing regression analysis on the pre-processed cloud resource usage data in any clustering result to obtain a resource usage prediction value corresponding to any clustering result, including: Building an initial regression model and obtaining historical cloud resource usage data that matches a classification label of any of the clustering results; Training the initial regression model based on the historical cloud resource usage data to obtain a trained regression model; Inputting the pre-processed cloud resource usage data in any clustering result into the trained regression model to obtain a resource usage prediction value corresponding to any clustering result; The regression model after training is: y=β0+β1x1+β2x2+…+β n x n +∈; Among them, y is the target variable; x1, x2, ... and x n is the independent variable; ∈ is the preset error value; β1, β2, ... and β n is the regression coefficient.

2. A cost optimization method based on cloud resource configuration according to claim 1, characterized in that: Perform cluster analysis on the pre-processed cloud resource usage data of all cloud service instances to obtain k clustering results, including: Obtaining the number k of clusters of the preprocessed cloud resource usage data; Initialize k cluster centers, calculate the Euclidean distance of each cloud service instance's pre-processed cloud resource usage data to each cluster center, and assign the pre-processed cloud resource usage data of each cloud service instance to the cluster with the closest Euclidean distance to the cluster center; Calculate the mean of the pre-processed cloud resource usage data of all cloud service instances in the cluster of each cluster center, and use the mean corresponding to each cluster center to update each cluster center to obtain k updated cluster centers; Each cluster center is set as the corresponding updated cluster center, and the Euclidean distance of the pre-processed cloud resource usage data of each cloud service instance to the k cluster centers is recalculated until the mean change of the pre-processed cloud resource usage data of all cloud service instances in the cluster cluster of each cluster center is less than the threshold. At this time, the cluster clusters corresponding to the k cluster centers constitute k clustering results.

3. The cost optimization method based on cloud resource configuration according to claim 1, characterized in that: Preprocessed cloud resource usage data x of the i-th cloud service instance i To the jth cluster center μ j The Euclidean distance is: Among them, x im Represents the pre-processed cloud resource usage data x of the cloud service instance i The mth data point in ; M is the pre-processed cloud resource usage data x of the cloud service instance i The total number of data points in .

4. The cost optimization method based on cloud resource configuration according to claim 1, characterized in that: Performing time series forecasting on the resource usage forecast value corresponding to any clustering result to obtain the resource usage forecast value corresponding to any clustering result within a specified future time period, including: Building an initial time series model and obtaining historical cloud resource usage data that matches a classification label of any of the clustering results; Training the initial time series model based on the historical cloud resource usage data to obtain a trained time series model; The resource usage prediction value corresponding to any clustering result is input into the trained time series model to obtain the resource usage prediction value corresponding to any clustering result in a future specified time period t.

5. The cost optimization method based on cloud resource configuration according to claim 4, characterized in that: The trained time series model is: y t =φ1y t-1 +φ2y t-2 +…+φ p y t-p +θ1∈ t-1 +θ2∈ t-2 +…+θ q ∈ t-q +∈ t ; Among them, y t y is the predicted value of resource usage corresponding to any clustering result in the future specified time period t; t-1 、y t-2 , ... and y t-p is the resource usage prediction value corresponding to any clustering result in the time periods t-1, t-2, ..., and tp; and are the autoregressive coefficients; θ1, θ2, … and θ q is the moving average coefficient; p is the autoregressive order; q is the moving average order; ∈ t-1 ,∈ t-2 ,…,∈ t-q and ∈ t is the random error in the time periods t-1, t-2, …, tq and t.

6. A cost optimization system based on cloud resource configuration, characterized in that: Used to implement the cost optimization method based on cloud resource configuration as described in any one of claims 1 to 5; The cost optimization system based on cloud resource configuration includes: An initial data acquisition module is used to acquire cloud resource usage data of each cloud service instance and preprocess the cloud resource usage data of each cloud service instance to obtain preprocessed cloud resource usage data of each cloud service instance; a clustering analysis module, communicatively connected to the initial data acquisition module, configured to perform cluster analysis on the pre-processed cloud resource usage data of all cloud service instances to obtain k clustering results; wherein k is a natural number greater than 1, and any clustering result includes the pre-processed cloud resource usage data of one or more cloud service instances, and each clustering result is associated with a different classification label; A regression analysis module, in communication with the cluster analysis module, for performing regression analysis on the pre-processed cloud resource usage data in each clustering result to obtain a resource usage prediction value corresponding to each clustering result; A time series prediction module, in communication with the regression analysis module, is used to perform time series prediction processing on the resource usage prediction values ​​corresponding to each clustering result to obtain the resource usage prediction values ​​corresponding to each clustering result in a specified future time period; The cloud resource configuration module is in communication with the time series prediction module and is used to configure cloud resources for the cloud service instances in each clustering result based on the resource usage prediction value corresponding to each clustering result in a specified future time period.

7. An electronic device, characterized in that: include: a memory for storing computer program instructions; as well as, A processor is configured to execute the computer program instructions to thereby complete the operation of the cost optimization method based on cloud resource configuration as described in any one of claims 1 to 5.

8. A computer program product comprising a computer program or instructions, characterized in that When the computer program or the instruction is executed by a computer, the cost optimization method based on cloud resource configuration according to any one of claims 1 to 5 is implemented.

Images