Cloud host resource demand prediction method and device, computer equipment and storage medium

Abstract

The invention relates to a cloud host resource demand prediction method and device, computer equipment and a storage medium. The method comprises the following steps: acquiring cloud host resource monitoring index data containing a CPU utilization rate and a memory utilization rate at a current moment, performing data annotation processing on the cloud host resource monitoring index data according to a first dimension, a second dimension and a third dimension to obtain first, second and third resource monitoring index data, and performing data annotation processing on the first, second and third resource monitoring index data through a PromQL expression. According to the first, second and third resource monitoring index data, obtaining first, second and third resource index data, and finally, according to the first, second and third resource monitoring index data, obtaining the first, second and third resource monitoring index data. And obtaining a predicted cloud host resource demand under the first dimension, a predicted cloud host resource demand under the second dimension and a predicted cloud host resource demand under the third dimension. Through data analysis and prediction of dimension division, the prediction accuracy of the cloud host resource demand is improved.

Description

Technical Field

[0001] This application relates to the field of cloud computing technology, and in particular to a method, apparatus, computer equipment, computer-readable storage medium, and computer program product for predicting cloud server resource demand. Background Technology

[0002] With the widespread application of cloud computing, the execution of various tasks is increasingly dependent on cloud server resources. Analyzing and predicting the historical situation of cloud server resources can yield the predicted cloud server resources needed for the next round. Based on the predicted cloud server resources, resource scheduling can be carried out in advance to ensure the smooth progress of various tasks.

[0003] However, current traditional methods for forecasting cloud server resource demand suffer from low forecast accuracy. Summary of the Invention

[0004] Therefore, it is necessary to provide a cloud resource demand forecasting method, apparatus, computer equipment, computer-readable storage medium, and computer program product that can improve the forecasting accuracy in response to the above-mentioned technical problems.

[0005] Firstly, this application provides a method for predicting cloud server resource demand, including:

[0006] Collect cloud server resource monitoring metrics data up to the current moment; cloud server resource monitoring metrics data include CPU utilization and memory utilization;

[0007] The cloud server resource monitoring indicator data are labeled according to the first dimension, the second dimension, and the third dimension to obtain the first resource monitoring indicator data, the second resource monitoring indicator data, and the third resource monitoring indicator data; the dimension range of the first dimension is greater than the dimension range of the second dimension, and the dimension range of the second dimension is greater than the dimension range of the third dimension.

[0008] Using PromQL expressions, first resource indicator data, second resource indicator data, and third resource indicator data are obtained based on first resource monitoring indicator data, second resource monitoring indicator data, and third resource monitoring indicator data. The resource indicator data includes maximum CPU utilization, average CPU utilization, maximum memory utilization, and average memory utilization. The resource indicator data can be any one of the first resource indicator data, second resource indicator data, and third resource indicator data.

[0009] Based on the first resource indicator data, the second resource indicator data, and the third resource indicator data, we obtain the predicted cloud server resource demand in the first dimension, the predicted cloud server resource demand in the second dimension, and the predicted cloud server resource demand in the third dimension.

[0010] In conjunction with the first aspect, in one embodiment, based on the first resource indicator data, the second resource indicator data, and the third resource indicator data, the predicted cloud server resource demand in the first dimension, the predicted cloud server resource demand in the second dimension, and the predicted cloud server resource demand in the third dimension are obtained, including:

[0011] Based on the first resource indicator data, the second resource indicator data, and the third resource indicator data, obtain the first sub-resource indicator data, the second sub-resource indicator data, and the third sub-resource indicator data for multiple historical time windows prior to the current moment;

[0012] Based on the first, second, and third sub-resource indicator data under each historical time window, generate corresponding first-time resource indicator data change charts, second-time resource indicator data change charts, and third-time resource indicator data change charts.

[0013] Based on the resource indicator data change charts at the first, second, and third time points, we obtain the predicted cloud server resource requirements in the first, second, and third dimensions.

[0014] In conjunction with the first aspect, in one embodiment, based on the first resource indicator data, the second resource indicator data, and the third resource indicator data, the first sub-resource indicator data, the second sub-resource indicator data, and the third sub-resource indicator data under multiple historical time windows prior to the current time are obtained, including:

[0015] The first sub-resource indicator data under multiple historical time windows prior to the current moment is obtained through an asynchronous query method using a thread pool;

[0016] The second sub-resource indicator data for multiple historical time windows prior to the current moment are obtained through an asynchronous query method using a thread pool.

[0017] The third sub-resource indicator data for multiple historical time windows prior to the current moment is obtained through an asynchronous query method using a thread pool.

[0018] In conjunction with the first aspect, in an exemplary embodiment, multiple historical time windows prior to the current moment are obtained through the following steps:

[0019] Retrieve the historical time series up to the current moment;

[0020] Based on the time slicing strategy, the historical time series is divided into time windows to obtain multiple historical time windows before the current moment.

[0021] In conjunction with the first aspect, in one embodiment, the method further includes:

[0022] The resource indicator data change graphs at the first, second, and third time points are sent to the user terminal; the user terminal is used to display the resource indicator data change graphs at the first, second, and third time points through a visual interface.

[0023] In conjunction with the first aspect, in one embodiment, for any current time resource indicator data change graph, where the current time resource indicator data change graph is any one of a first time resource indicator data change graph, a second time resource indicator data change graph, and a third time resource indicator data change graph, the method further includes:

[0024] Obtain pre-set resource indicator data thresholds;

[0025] If, in the current time resource indicator data change graph, there is a resource indicator data that is greater than or equal to the resource indicator data threshold, the resource indicator data will be identified as abnormal resource indicator data.

[0026] The system generates early warning information based on the abnormal resource indicator data and returns it to the user terminal. The user terminal is used to extract the abnormal resource indicator data from the early warning information, generate optimization suggestions based on the abnormal resource indicator data, highlight the abnormal resource indicators through a visual interface, and display the corresponding optimization suggestions.

[0027] Secondly, this application also provides a cloud server resource demand prediction device, comprising:

[0028] The data acquisition module is used to collect cloud server resource monitoring metrics data up to the current moment; the cloud server resource monitoring metrics data includes CPU utilization and memory utilization.

[0029] The data annotation module is used to annotate cloud host resource monitoring indicator data according to the first dimension, the second dimension, and the third dimension to obtain the first resource monitoring indicator data, the second resource monitoring indicator data, and the third resource monitoring indicator data; the dimension range of the first dimension is greater than the dimension range of the second dimension, and the dimension range of the second dimension is greater than the dimension range of the third dimension.

[0030] The data analysis module is used to obtain first resource indicator data, second resource indicator data, and third resource indicator data using PromQL expressions based on first resource monitoring indicator data, second resource monitoring indicator data, and third resource monitoring indicator data. The resource indicator data includes maximum CPU utilization, average CPU utilization, maximum memory utilization, and average memory utilization. The resource indicator data can be any one of the first resource indicator data, second resource indicator data, and third resource indicator data.

[0031] The demand forecasting module is used to obtain the predicted cloud server resource demand in the first dimension, the predicted cloud server resource demand in the second dimension, and the predicted cloud server resource demand in the third dimension based on the first resource indicator data, the second resource indicator data, and the third resource indicator data.

[0032] Thirdly, this application also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to perform the following steps:

[0033] Collect cloud server resource monitoring metrics data up to the current moment; cloud server resource monitoring metrics data include CPU utilization and memory utilization;

[0034] The cloud server resource monitoring indicator data are labeled according to the first dimension, the second dimension, and the third dimension to obtain the first resource monitoring indicator data, the second resource monitoring indicator data, and the third resource monitoring indicator data; the dimension range of the first dimension is greater than the dimension range of the second dimension, and the dimension range of the second dimension is greater than the dimension range of the third dimension.

[0035] Using PromQL expressions, first resource indicator data, second resource indicator data, and third resource indicator data are obtained based on first resource monitoring indicator data, second resource monitoring indicator data, and third resource monitoring indicator data. The resource indicator data includes maximum CPU utilization, average CPU utilization, maximum memory utilization, and average memory utilization. The resource indicator data can be any one of the first resource indicator data, second resource indicator data, and third resource indicator data.

[0036] Based on the first resource indicator data, the second resource indicator data, and the third resource indicator data, we obtain the predicted cloud server resource demand in the first dimension, the predicted cloud server resource demand in the second dimension, and the predicted cloud server resource demand in the third dimension.

[0037] Fourthly, this application also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, performs the following steps:

[0038] Collect cloud server resource monitoring metrics data up to the current moment; cloud server resource monitoring metrics data include CPU utilization and memory utilization;

[0039] The cloud server resource monitoring indicator data are labeled according to the first dimension, the second dimension, and the third dimension to obtain the first resource monitoring indicator data, the second resource monitoring indicator data, and the third resource monitoring indicator data; the dimension range of the first dimension is greater than the dimension range of the second dimension, and the dimension range of the second dimension is greater than the dimension range of the third dimension.

[0040] Using PromQL expressions, first resource indicator data, second resource indicator data, and third resource indicator data are obtained based on first resource monitoring indicator data, second resource monitoring indicator data, and third resource monitoring indicator data. The resource indicator data includes maximum CPU utilization, average CPU utilization, maximum memory utilization, and average memory utilization. The resource indicator data can be any one of the first resource indicator data, second resource indicator data, and third resource indicator data.

[0041] Based on the first resource indicator data, the second resource indicator data, and the third resource indicator data, we obtain the predicted cloud server resource demand in the first dimension, the predicted cloud server resource demand in the second dimension, and the predicted cloud server resource demand in the third dimension.

[0042] Fifthly, this application also provides a computer program product, including a computer program that, when executed by a processor, performs the following steps:

[0043] Collect cloud server resource monitoring metrics data up to the current moment; cloud server resource monitoring metrics data include CPU utilization and memory utilization;

[0044] The cloud server resource monitoring indicator data are labeled according to the first dimension, the second dimension, and the third dimension to obtain the first resource monitoring indicator data, the second resource monitoring indicator data, and the third resource monitoring indicator data; the dimension range of the first dimension is greater than the dimension range of the second dimension, and the dimension range of the second dimension is greater than the dimension range of the third dimension.

[0045] Using PromQL expressions, first resource indicator data, second resource indicator data, and third resource indicator data are obtained based on first resource monitoring indicator data, second resource monitoring indicator data, and third resource monitoring indicator data. The resource indicator data includes maximum CPU utilization, average CPU utilization, maximum memory utilization, and average memory utilization. The resource indicator data can be any one of the first resource indicator data, second resource indicator data, and third resource indicator data.

[0046] Based on the first resource indicator data, the second resource indicator data, and the third resource indicator data, we obtain the predicted cloud server resource demand in the first dimension, the predicted cloud server resource demand in the second dimension, and the predicted cloud server resource demand in the third dimension.

[0047] The aforementioned cloud server resource demand prediction method, apparatus, computer equipment, computer-readable storage medium, and computer program product collect cloud server resource monitoring indicator data, including CPU utilization and memory utilization, up to the current moment. The data is then labeled according to a first dimension, a second dimension, and a third dimension to obtain first resource monitoring indicator data, second resource monitoring indicator data, and third resource monitoring indicator data. The first dimension has a wider range than the second dimension, and the second dimension has a wider range than the third dimension. Using PromQL expressions, first resource indicator data, second resource indicator data, and third resource monitoring indicator data are obtained from these data. The resource indicator data includes maximum CPU utilization, average CPU utilization, maximum memory utilization, and average memory utilization. Each resource indicator can be any one of the first, second, and third resource indicator data. Finally, based on the first, second, and third resource indicator data, predicted cloud server resource demand under the first dimension, predicted cloud server resource demand under the second dimension, and predicted cloud server resource demand under the third dimension are obtained. By acquiring cloud server resource monitoring indicator data and tagging the collected data, performing data analysis according to the corresponding tags, and predicting cloud server resource demand in different dimensions based on the analysis results, the accuracy of cloud server resource demand prediction is improved through dimensional data analysis and prediction. Attached Figure Description

[0048] To more clearly illustrate the technical solutions in the embodiments of this application or related technologies, the drawings used in the description of the embodiments of this application or related technologies will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0049] Figure 1 This is a diagram illustrating the application environment of a cloud server resource demand prediction method in one embodiment.

[0050] Figure 2 This is a flowchart illustrating a cloud server resource demand prediction method in one embodiment;

[0051] Figure 3 This is a diagram of a multi-cloud management monitoring technology architecture in one embodiment;

[0052] Figure 4 This is a diagram of the multi-cloud management monitoring function architecture in another embodiment;

[0053] Figure 5 Here is a graph showing the monitoring indicators for cloud hosts in another embodiment;

[0054] Figure 6 This is a diagram showing resource usage from an organizational perspective in one embodiment.

[0055] Figure 7 Here is a diagram showing resource usage from a project perspective in another embodiment;

[0056] Figure 8 This is a detailed chart showing resource usage in one embodiment.

[0057] Figure 9 This is a structural block diagram of a cloud host resource demand prediction device in one embodiment;

[0058] Figure 10 This is an internal structural diagram of a computer device in one embodiment. Detailed Implementation

[0059] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0060] With the widespread adoption of cloud computing, enterprises are increasingly reliant on cloud server resources. Current technologies for analyzing cloud server resource usage often limit themselves to monitoring single dimensions or simple metrics, such as focusing only on the CPU utilization or memory usage of a single cloud server. This fails to provide a comprehensive understanding of the overall utilization of cloud server resources within a specific business or department. In complex business scenarios, it becomes difficult to accurately identify performance bottlenecks and resource waste points, leading to unreasonable resource allocation, increased operating costs, and compromised business stability and user experience. Therefore, it is necessary to understand the overall resource usage in a timely manner, accurately predict resource needs in advance based on business requirements, and optimize allocated resources accordingly: expanding capacity for businesses with insufficient cloud server resources and downsizing or reclaiming resources for businesses with low cloud server utilization.

[0061] The cloud server resource demand prediction method provided in this application embodiment can be applied to, for example, Figure 1In the application environment shown, user terminal 102 communicates with server 104 via a network. A data storage system can store the data that server 104 needs to process. The data storage system can be integrated onto server 104, or it can be located in the cloud or on other network servers. Server 104 collects cloud host resource monitoring metrics data up to the current moment from the data storage system. These metrics include CPU utilization and memory utilization. The cloud host resource metrics data are labeled according to a first, second, and third dimension to obtain first, second, and third resource monitoring metrics data. The first dimension has a wider range than the second dimension, and the second dimension has a wider range than the third dimension. Using PromQL expressions, the first, second, and third resource metrics data are obtained from these metrics. These resource metrics include maximum CPU utilization, average CPU utilization, maximum memory utilization, and average memory utilization. Each resource metric can be any one of the first, second, and third resource metrics. Finally, based on these three resource metrics, predicted cloud host resource requirements under the first, second, and third dimensions are obtained. The user terminal 102 can be, but is not limited to, various personal computers, laptops, smartphones, tablets, IoT devices, and portable wearable devices. IoT devices can include smart speakers, smart TVs, smart air conditioners, smart in-vehicle systems, and projection devices. Portable wearable devices can include smartwatches, smart bracelets, and head-mounted displays. Head-mounted displays can be virtual reality (VR) devices, augmented reality (AR) devices, and smart glasses. The server 104 can be a standalone physical server, a server cluster or distributed system composed of multiple physical servers, or a cloud server providing cloud computing services.

[0062] In one exemplary embodiment, such as Figure 2 As shown, a cloud server resource demand prediction method is provided, which is applied to... Figure 1 Taking server 104 as an example, the explanation includes the following steps S201 to S204. Wherein:

[0063] Step S201: Collect cloud host resource monitoring indicator data up to the current moment; cloud host resource monitoring indicator data includes CPU utilization and memory utilization.

[0064] Among them, cloud server resource monitoring metrics can be understood as data used to judge the usage of cloud server resources, which may include CPU utilization, memory utilization, and disk I / O, etc.

[0065] For example, server 104 can collect cloud host resource monitoring index data at preset time intervals and extract specific performance data from the target object according to preset collection parameters, such as CPU utilization, memory usage, disk I / O, etc.

[0066] Step S202: Label the cloud host resource monitoring indicator data according to the first dimension, the second dimension, and the third dimension to obtain the first resource monitoring indicator data, the second resource monitoring indicator data, and the third resource monitoring indicator data; the dimension range of the first dimension is greater than the dimension range of the second dimension, and the dimension range of the second dimension is greater than the dimension range of the third dimension.

[0067] The first dimension can be understood as the dimension with the largest coverage of the target object, which can include organizational and departmental dimensions. The second dimension can be understood as the dimension with the second largest coverage of the target object, which can include projects, etc. The third dimension can be understood as the dimension with the smallest coverage of the target object, which can include details, etc., that is, the instances contained under the project are used as dividing labels.

[0068] Optionally, server 104 labels the cloud host resource monitoring indicator data according to the first dimension, the second dimension, and the third dimension, wherein the coverage of the target objects corresponding to the first dimension, the second dimension, and the third dimension are successively reduced. The cloud host resource monitoring indicator data labeled with the first dimension is used as the first resource monitoring indicator data, the cloud host resource monitoring indicator data labeled with the second dimension is used as the second resource monitoring indicator data, and the cloud host resource monitoring indicator data labeled with the third dimension is used as the third resource monitoring indicator data.

[0069] Step S203: Using PromQL expressions, obtain the first resource indicator data, the second resource indicator data, and the third resource indicator data based on the first resource monitoring indicator data, the second resource monitoring indicator data, and the third resource monitoring indicator data. The resource indicator data includes the maximum CPU utilization, the average CPU utilization, the maximum memory utilization, and the average memory utilization. The resource indicator data can be any one of the first resource indicator data, the second resource indicator data, and the third resource indicator data.

[0070] Among them, PromQL expressions are a query language specifically designed for the Prometheus monitoring system, used to query time-series data stored in Prometheus; the first resource metric data can be understood as the maximum CPU utilization, average CPU utilization, maximum memory utilization, and average memory utilization in the first dimension. Similarly, the second resource metric data can be understood as the maximum CPU utilization, average CPU utilization, maximum memory utilization, and average memory utilization in the second dimension. The third resource metric data can be understood as the maximum CPU utilization, average CPU utilization, maximum memory utilization, and average memory utilization in the third dimension.

[0071] For example, server 104 performs data analysis and calculation on the first resource monitoring indicator data from the Prometheus monitoring system using PromQL expressions to obtain the corresponding first resource indicator data, performs data analysis and calculation on the second resource monitoring indicator data to obtain the corresponding second resource indicator data, and performs data analysis and calculation on the third resource monitoring indicator data to obtain the corresponding third resource indicator data.

[0072] Step S204: Based on the first resource indicator data, the second resource indicator data, and the third resource indicator data, obtain the predicted cloud host resource requirements in the first dimension, the second dimension, and the third dimension.

[0073] Among them, the predicted cloud server resource demand can be understood as the future usage of cloud server resources.

[0074] Optionally, server 104 draws a resource indicator data change graph for a first time based on the first resource indicator data, analyzes the graphical trend of the resource indicator data change graph for the first time, and obtains the predicted cloud host resource demand in the first dimension; draws a resource indicator data change graph for a second time based on the second resource indicator data, analyzes the graphical trend of the resource indicator data change graph for the second time, and obtains the predicted cloud host resource demand in the second dimension; and draws a resource indicator data change graph for a third time based on the third resource indicator data, analyzes the graphical trend of the resource indicator data change graph for the third time, and obtains the cloud host resource demand in the third dimension.

[0075] In the aforementioned cloud server resource demand prediction method, cloud server resource monitoring indicator data, including CPU utilization and memory utilization, up to the current moment, are collected. This data is then labeled according to a first dimension, a second dimension, and a third dimension to obtain first, second, and third resource monitoring indicator data. The first dimension has a wider range than the second dimension, and the second dimension has a wider range than the third dimension. Using PromQL expressions, first, second, and third resource indicator data are obtained based on these data. The resource indicator data includes maximum CPU utilization, average CPU utilization, maximum memory utilization, and average memory utilization. Each resource indicator can be any one of the first, second, or third resource indicator data. Finally, based on these three resource indicator data, predicted cloud server resource demand under the first dimension, predicted cloud server resource demand under the second dimension, and predicted cloud server resource demand under the third dimension are obtained. By acquiring cloud server resource monitoring indicator data and tagging the collected data, performing data analysis according to the corresponding tags, and predicting cloud server resource demand in different dimensions based on the analysis results, the accuracy of cloud server resource demand prediction is improved through dimensional data analysis and prediction.

[0076] In one embodiment, based on first resource indicator data, second resource indicator data, and third resource indicator data, the predicted cloud server resource demand in the first dimension, the predicted cloud server resource demand in the second dimension, and the predicted cloud server resource demand in the third dimension are obtained, including:

[0077] Based on the first, second, and third resource indicator data, obtain the first, second, and third sub-resource indicator data for multiple historical time windows prior to the current moment; according to the first, second, and third sub-resource indicator data for each historical time window, generate corresponding first-time resource indicator data change graphs, second-time resource indicator data change graphs, and third-time resource indicator data change graphs; based on the first-time resource indicator data change graphs, second-time resource indicator data change graphs, and third-time resource indicator data change graphs, obtain the predicted cloud host resource demand in the first dimension, the predicted cloud host resource demand in the second dimension, and the predicted cloud host resource demand in the third dimension.

[0078] Here, the historical time window can be understood as a smaller time segment obtained by dividing a long time period up to the current moment into segments. The first time resource indicator data change chart can be understood as a trend chart of cloud host resource indicator data changing over time in the first dimension. Similarly, the second time resource indicator data change chart can be understood as a trend chart of cloud host resource indicator data changing over time in the second dimension. The third time resource indicator data change chart can be understood as a trend chart of cloud host resource indicator data changing over time in the third dimension. The change trend chart can be understood as an image of different shapes used to display the trend of resources, which can include bar charts, line charts, and pie charts, etc.

[0079] Optionally, server 104 performs dynamic sharding queries based on the first resource indicator data, the second resource indicator data, and the third resource indicator data to obtain the first sub-resource indicator data, the second sub-resource indicator data, and the third sub-resource indicator data under multiple historical time windows before the current time. According to the first sub-resource indicator data, the second sub-resource indicator data, and the third sub-resource indicator data under each historical time window, corresponding first time resource indicator data change graphs, second time resource indicator data change graphs, and third time resource indicator data change graphs are generated. The server performs trend analysis on the first time resource indicator data change graphs, the second time resource indicator data change graphs, and the third time resource indicator data change graphs to obtain the predicted cloud host resource demand in the first dimension, the predicted cloud host resource demand in the second dimension, and the predicted cloud host resource demand in the third dimension.

[0080] Based on the aforementioned implementation method, by generating a corresponding time-based resource indicator data change graph, the changes in resource indicator data over time can be visualized intuitively, facilitating the analysis of change trends and thereby accelerating the acquisition of predicted cloud host resource requirements under different dimensions.

[0081] In one embodiment, based on first resource indicator data, second resource indicator data, and third resource indicator data, first sub-resource indicator data, second sub-resource indicator data, and third sub-resource indicator data under multiple historical time windows prior to the current time are obtained, including:

[0082] The system uses an asynchronous query method with a thread pool to retrieve the first sub-resource indicator data for multiple historical time windows prior to the current moment; it also uses an asynchronous query method with a thread pool to retrieve the second sub-resource indicator data for multiple historical time windows prior to the current moment; and it uses an asynchronous query method with a thread pool to retrieve the third sub-resource indicator data for multiple historical time windows prior to the current moment.

[0083] In this context, asynchronous querying via thread pool can be understood as synchronously querying data from different nodes, including different data and different time periods.

[0084] For example, server 104 uses an asynchronous query method with a thread pool to synchronously obtain the first sub-resource indicator data under multiple historical time windows before the current time, simultaneously obtain the second sub-resource indicator data under multiple historical time windows before the current time, and simultaneously obtain the third sub-resource indicator data under multiple historical time windows before the current time.

[0085] According to the above implementation method, by utilizing the thread pool exception query method, the synchronous query and acquisition of sub-data of different dimensions and different time periods is realized, thereby accelerating the query speed of sub-resource indicator data and thus accelerating the acquisition speed of predicting cloud host resource requirements.

[0086] In an exemplary embodiment, multiple historical time windows prior to the current moment are obtained through the following steps: obtaining the historical time series up to the current moment; dividing the historical time series into time windows according to the time slicing strategy to obtain multiple historical time windows prior to the current moment.

[0087] Optionally, server 104 obtains the historical time series up to the current moment, and divides the historical time series into multiple time intervals of equal length according to the time slicing strategy, thereby obtaining multiple historical time windows before the current moment.

[0088] Based on the aforementioned implementation method, the historical time series is divided into time intervals by using a time slicing strategy, thereby obtaining the corresponding historical time window. This provides a query location for querying the sub-resource indicator data under the corresponding time window, and also provides a data foundation for subsequently generating the corresponding time resource indicator data change chart.

[0089] In one embodiment, the method further includes: sending a first-time resource indicator data change graph, a second-time resource indicator data change graph, and a third-time resource indicator data change graph to a user terminal; the user terminal is used to display the first-time resource indicator data change graph, the second-time resource indicator data change graph, and the third-time resource indicator data change graph through a visual interface.

[0090] The visual interface can be understood as a display screen that converts computer language into natural language that users can understand and present it.

[0091] For example, server 104 sends the first time resource indicator data change graph, the second time resource indicator data change graph, and the third time resource indicator data change graph to user terminal 102. User terminal 102 displays the second time resource indicator data change graph, the second time resource indicator data change graph, and the third time resource indicator data change graph to the user through a visual interface.

[0092] According to the above implementation method, by sending the first time resource indicator data change chart, the second time resource indicator data change chart, and the third time resource indicator data change chart to the user terminal, and having the user terminal display the images through a visual interface, it is easier for users to clearly perceive the historical resource usage and learn about future resource demand information, so as to make advance manual intervention and adjustment.

[0093] In one embodiment, for any current time resource indicator data change graph, which is any one of a first time resource indicator data change graph, a second time resource indicator data change graph, and a third time resource indicator data change graph, the method further includes:

[0094] The system retrieves pre-set resource indicator data thresholds. If, in the current time resource indicator data change graph, there is a resource indicator data that is greater than or equal to the resource indicator data threshold, the resource indicator data is identified as abnormal resource indicator data. An early warning message is generated based on the abnormal resource indicator data and returned to the user terminal. The user terminal is used to extract the abnormal resource indicator data from the early warning message, generate optimization suggestions based on the abnormal resource indicator data, highlight the abnormal resource indicator through a visual interface, and display the corresponding optimization suggestions.

[0095] Among them, the resource indicator data threshold can be understood as the maximum utilization level of relevant resources under normal cloud server operation. Exceeding this level may cause the system to fail to operate normally and stably. The early warning information can be understood as the information used to alert to abnormal situations, which may include early warning text information and early warning signals. The optimization suggestions can be understood as the relevant processing measures used to deal with abnormal resource indicator data so that the system can return to normal operation, which may include increasing the number of CPU cores, expanding memory, etc.

[0096] Optionally, for any one of the resource indicator data change maps of the first time, the second time, and the third time, the server 104 obtains a pre-set resource indicator data threshold. If there is a resource indicator data in the current time resource indicator data change map that is greater than or equal to the resource indicator data threshold, the resource indicator data is identified as abnormal resource indicator data. An early warning message is generated based on the abnormal resource indicator data and returned to the user terminal 102. The user terminal 102 extracts the abnormal resource indicator data from the early warning message and generates optimization suggestions based on the abnormal resource indicator data. The abnormal resource indicator data is highlighted or highlighted in yellow through a visualization interface, and the corresponding optimization suggestions are displayed.

[0097] Based on the aforementioned implementation methods, potential resource bottlenecks or fault points can be quickly identified by real-time monitoring thresholds, and automatic alarms and suggestions can be provided to improve operational efficiency and system reliability.

[0098] In one exemplary embodiment, a specific implementation process of a cloud server resource demand prediction method is provided, applicable to, for example, Figure 3 and Figure 4 The architecture diagram shown below. Where:

[0099] The research focuses on unified monitoring and data collection capabilities for cloud resources, enabling cross-cloud platform operation, support for various resource types, and adaptation to multiple data collection methods (API interfaces (Application Programming Interface) and QGA (Quality Guarantee Analysis)) to meet diverse customer site requirements.

[0100] Multi-cloud management, based on Prometheus, enables comprehensive monitoring and alerting of cloud resources. Key design features are as follows:

[0101] ① It supports multiple cloud platforms and various resource types (including virtual machines, storage, networks, etc.), and supports multiple data acquisition methods (API interface, QGA) to adapt to different customer site needs.

[0102] ② It supports highly customizable monitoring settings and provides a wealth of configurable monitoring metrics to meet various complex monitoring needs.

[0103] ③Supports multi-dimensional (organization, project, resource) resource usage analysis.

[0104] The main key steps involved in the function:

[0105] I. Data Collection Implementation Steps:

[0106] 1. Cloud Platform API Call Data Collection: Multi-Cloud Management uses an SDK (Software Development Kit) for resource management. It uses an AccessKey for authentication to call the DescribeInstanceMonitorData monitoring interface to obtain cloud host resource monitoring data. Data is collected every 5 minutes and stored in Prometheus. Similar methods are used for other product resources, with the corresponding SDKs calling the API at set time intervals to obtain data and store it in Prometheus.

[0107] 2. Exporter Plugin Data Collection: In the Prometheus configuration file, add data collection task configurations for each Exporter, specifying parameters such as the target address and collection interval. For Linux cloud hosts, a NodeExporter data collection plugin is developed to extract specific performance data from the target object, such as CPU utilization, memory usage, and disk I / O, and expose this data in a Prometheus-supported format for the Prometheus server to crawl and store. Different types of Exporters are responsible for monitoring different types of metrics.

[0108] 3. Tag the collected monitoring indicator data: organization, resource set (project), cloud environment information, etc., store it in Prometheus database, and use PromQL expressions to retrieve the corresponding data based on the tag values ​​for analysis.

[0109] II. Data Analysis and Model Building:

[0110] 1. Statistical Analysis Implementation: Extract historical data on cloud server resource usage from Prometheus, assemble complex expressions using PromQL to calculate the average and maximum values ​​of indicators such as CPU utilization and memory utilization, and evaluate resource usage.

[0111] 2. The cloud server usage function involves scenarios with large query time spans and large amounts of data. If these scenarios are not handled properly and the corresponding indicator data is directly queried from Prometheus, it will lead to problems such as high load on Prometheus and long query time. Therefore, a dynamic sharding query method is adopted. This method divides a large time period query into multiple smaller time periods through an algorithm, and then uses a thread pool to asynchronously query the indicator data of multiple smaller time periods. Finally, the indicator data obtained from multiple smaller time periods are aggregated to obtain the corresponding cloud server usage indicator data.

[0112] 3. Trend Analysis Implementation: Use the Matplotlib library to draw time series line charts of resource indicators such as CPU utilization and network bandwidth. The horizontal axis represents time, and the vertical axis represents resource usage. Analyze resource usage trends and predict future resource demand.

[0113] III. Implementation of Visualization and Decision Support:

[0114] 1. Visual Interface Development: Develop a visual decision support system using the front-end framework Vue.js and the back-end framework Django, retrieving analysis results data from the database, such as... Figure 5 As shown, the cloud server resource usage, analysis results, and predicted trends are displayed in the form of bar charts, line charts, pie charts, etc.

[0115] 2. Intelligent early warning function implementation: Set resource usage thresholds. When CPU usage exceeds 85% or memory usage exceeds 95%, the system sends early warning information to relevant personnel via SMS and other means, highlights abnormal indicators on the visual interface, and provides optimization suggestions, such as increasing the number of CPU cores or expanding memory.

[0116] 3. For example Figure 6 As shown, the organizational dimension analysis (i.e., the first dimension mentioned above) displays the average and maximum values ​​of indicators such as CPU utilization and memory utilization of cloud hosts according to the organizational dimension.

[0117] 4. For example Figure 7 As shown, the project dimension analysis (i.e., the second dimension mentioned above) displays the average and maximum values ​​of indicators such as CPU utilization and memory utilization of cloud hosts according to the project dimension.

[0118] 5. For example Figure 8 As shown, the detailed analysis (i.e. the aforementioned third dimension) displays the average and maximum values ​​of indicators such as CPU utilization and memory utilization of cloud hosts, as well as detailed monitoring information of current cloud host indicators according to the detailed dimension.

[0119] Compared with the prior art, this application has the following technical advantages:

[0120] 1. Supports multi-dimensional aggregated analysis of organizations, projects, resources, etc., providing comprehensive basis for management and optimization, and improving the efficiency of cloud resource utilization and the accuracy of cost management.

[0121] 2. Utilize APIs, Exporters, and dynamic sharding queries to ensure efficient collection, storage, and querying of large datasets, resolve performance bottlenecks in queries spanning long time periods, and guarantee the response speed and stability of the monitoring system.

[0122] 3. Implement time series trend analysis using Matplotlib to help operations and maintenance personnel predict resource needs in advance. The charts are intuitive and visual, improving data readability and decision support efficiency.

[0123] 4. Real-time monitoring thresholds enable rapid detection of potential resource bottlenecks or fault points, automatic alarms and suggestions, and improved operation and maintenance efficiency and system reliability.

[0124] It should be understood that although the steps in the flowcharts of the embodiments described above are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the embodiments described above may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages of other steps.

[0125] Based on the same inventive concept, this application also provides a cloud server resource demand forecasting device for implementing the cloud server resource demand forecasting method described above. The solution provided by this device is similar to the solution described in the above method; therefore, the specific limitations in one or more cloud server resource demand forecasting device embodiments provided below can be found in the limitations of the cloud server resource demand forecasting method described above, and will not be repeated here.

[0126] In one exemplary embodiment, such as Figure 9 As shown, a cloud server resource demand forecasting device is provided, comprising: a data acquisition module 901, a data labeling module 902, a data analysis module 903, and a demand forecasting module 904, wherein:

[0127] The data acquisition module 901 is used to collect cloud host resource monitoring indicator data up to the current moment; the cloud host resource monitoring indicator data includes CPU utilization and memory utilization.

[0128] The data annotation module 902 is used to annotate the cloud host resource monitoring indicator data according to the first dimension, the second dimension, and the third dimension to obtain the first resource monitoring indicator data, the second resource monitoring indicator data, and the third resource monitoring indicator data; the dimension range of the first dimension is greater than the dimension range of the second dimension, and the dimension range of the second dimension is greater than the dimension range of the third dimension.

[0129] The data analysis module 903 is used to obtain first resource indicator data, second resource indicator data, and third resource indicator data through PromQL expressions based on first resource monitoring indicator data, second resource monitoring indicator data, and third resource monitoring indicator data. The resource indicator data includes maximum CPU utilization, average CPU utilization, maximum memory utilization, and average memory utilization. The resource indicator data can be any one of the first resource indicator data, second resource indicator data, and third resource indicator data.

[0130] The demand forecasting module 904 is used to obtain the predicted cloud host resource demand in the first dimension, the predicted cloud host resource demand in the second dimension, and the predicted cloud host resource demand in the third dimension based on the first resource indicator data, the second resource indicator data, and the third resource indicator data.

[0131] In one embodiment, the demand forecasting module 904 further includes a data acquisition submodule, a change graph generation submodule, and a demand forecasting submodule, wherein:

[0132] The data acquisition submodule is used to acquire first sub-resource indicator data, second sub-resource indicator data and third sub-resource indicator data under multiple historical time windows before the current moment, based on first resource indicator data, second resource indicator data and third resource indicator data.

[0133] The change graph generation submodule is used to generate corresponding first-time resource indicator data change graphs, second-time resource indicator data change graphs, and third-time resource indicator data change graphs according to the first, second, and third sub-resource indicator data under each historical time window.

[0134] The demand forecasting submodule is used to obtain the predicted cloud server resource demand in the first dimension, the predicted cloud server resource demand in the second dimension, and the predicted cloud server resource demand in the third dimension based on the resource indicator data change graphs in the first, second, and third time periods.

[0135] In one embodiment, the data acquisition submodule is further configured to obtain first sub-resource indicator data under multiple historical time windows prior to the current moment through an asynchronous query method using a thread pool; obtain second sub-resource indicator data under multiple historical time windows prior to the current moment through an asynchronous query method using a thread pool; and obtain third sub-resource indicator data under multiple historical time windows prior to the current moment through an asynchronous query method using a thread pool.

[0136] In an exemplary embodiment, the cloud host resource demand prediction device further includes a time window segmentation module, which is used to obtain historical time series up to the current moment; and to segment the historical time series into time windows according to a time slicing strategy to obtain multiple historical time windows before the current moment.

[0137] In one embodiment, the change graph generation submodule is further configured to send the first time resource indicator data change graph, the second time resource indicator data change graph, and the third time resource indicator data change graph to the user terminal; the user terminal is configured to display the first time resource indicator data change graph, the second time resource indicator data change graph, and the third time resource indicator data change graph through a visual interface.

[0138] In one embodiment, for any current-time resource indicator data change graph, which can be any one of a first-time resource indicator data change graph, a second-time resource indicator data change graph, and a third-time resource indicator data change graph, the change graph generation submodule is further used to obtain a pre-set resource indicator data threshold. If there is a resource indicator data in the current-time resource indicator data change graph that is greater than or equal to the resource indicator data threshold, the resource indicator data is determined to be abnormal resource indicator data. An early warning message is generated based on the abnormal resource indicator data and returned to the user terminal. The user terminal is used to extract the abnormal resource indicator data from the early warning message, generate optimization suggestions based on the abnormal resource indicator data, highlight the abnormal resource indicator through a visual interface, and display the corresponding optimization suggestions.

[0139] Each module in the aforementioned cloud server resource demand prediction device can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device, or stored in the memory of a computer device as software, so that the processor can call and execute the corresponding operations of each module.

[0140] In one exemplary embodiment, a computer device is provided, which may be a server, and its internal structure diagram may be as follows: Figure 10As shown, the computer device includes a processor, memory, input / output (I / O) interfaces, and a communication interface. The processor, memory, and I / O interfaces are connected via a system bus, and the communication interface is also connected to the system bus via the I / O interfaces. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and a database. The internal memory provides the environment for the operating system and computer programs stored in the non-volatile storage media. The database stores cloud host resource monitoring index data, including first resource monitoring index data, second resource monitoring index data, third resource monitoring index data, first resource index data, second resource index data, third resource index data, predicted cloud host resource requirements in the first dimension, predicted cloud host resource requirements in the second dimension, and predicted cloud host resource requirements in the third dimension. The I / O interfaces are used for exchanging information between the processor and external devices. The communication interface is used for communicating with external terminals via a network connection. When the computer program is executed by the processor, it implements a cloud host resource requirement prediction method.

[0141] Those skilled in the art will understand that Figure 10 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0142] In one exemplary embodiment, a computer device is provided, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the cloud host resource demand prediction method of the above embodiment.

[0143] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon, which, when executed by a processor, implements the cloud host resource demand prediction method of the above embodiment.

[0144] In one embodiment, a computer program product is provided, including a computer program that, when executed by a processor, implements the cloud host resource demand prediction method of the above embodiment.

[0145] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of the relevant data must comply with relevant regulations.

[0146] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile memory and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM). The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, artificial intelligence (AI) processors, etc., and are not limited to these.

[0147] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this application.

[0148] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this application should be determined by the appended claims.

Claims

1. A method for predicting cloud server resource demand, characterized in that, The method includes: Collect cloud server resource monitoring metrics data up to the current moment; the cloud server resource monitoring metrics data includes CPU utilization and memory utilization. The cloud host resource monitoring indicator data are labeled according to the first dimension, the second dimension, and the third dimension to obtain the first resource monitoring indicator data, the second resource monitoring indicator data, and the third resource monitoring indicator data; the dimension range of the first dimension is greater than the dimension range of the second dimension, and the dimension range of the second dimension is greater than the dimension range of the third dimension. Using PromQL expressions, first resource indicator data, second resource indicator data, and third resource indicator data are obtained based on the first resource monitoring indicator data, the second resource monitoring indicator data, and the third resource monitoring indicator data. The resource indicator data includes maximum CPU utilization, average CPU utilization, maximum memory utilization, and average memory utilization. The resource indicator data can be any one of the first resource indicator data, the second resource indicator data, and the third resource indicator data. Based on the first resource indicator data, the second resource indicator data, and the third resource indicator data, the predicted cloud host resource requirements under the first dimension, the predicted cloud host resource requirements under the second dimension, and the predicted cloud host resource requirements under the third dimension are obtained.

2. The method according to claim 1, characterized in that, The step of obtaining the predicted cloud server resource requirements in the first dimension, the second dimension, and the third dimension based on the first resource indicator data, the second resource indicator data, and the third resource indicator data includes: Based on the first resource indicator data, the second resource indicator data, and the third resource indicator data, obtain the first sub-resource indicator data, the second sub-resource indicator data, and the third sub-resource indicator data under multiple historical time windows prior to the current moment; Based on the first, second, and third sub-resource indicator data under each historical time window, generate corresponding first-time resource indicator data change charts, second-time resource indicator data change charts, and third-time resource indicator data change charts. Based on the first time-based resource indicator data change chart, the second time-based resource indicator data change chart, and the third time-based resource indicator data change chart, the predicted cloud host resource demand under the first dimension, the predicted cloud host resource demand under the second dimension, and the predicted cloud host resource demand under the third dimension are obtained.

3. The method according to claim 2, characterized in that, The step of obtaining first sub-resource indicator data, second sub-resource indicator data, and third sub-resource indicator data under multiple historical time windows prior to the current time, based on the first resource indicator data, the second resource indicator data, and the third resource indicator data, includes: The first sub-resource indicator data under multiple historical time windows prior to the current moment is obtained through an asynchronous query method using a thread pool; The second sub-resource indicator data under multiple historical time windows prior to the current moment are obtained through an asynchronous query method using a thread pool; The third sub-resource indicator data under multiple historical time windows prior to the current moment are obtained through an asynchronous query method using a thread pool.

4. The method according to any one of claims 1-3, characterized in that, The multiple historical time windows prior to the current moment are obtained through the following steps: Retrieve the historical time series up to the current moment; According to the time slicing strategy, the historical time series is divided into time windows to obtain multiple historical time windows before the current moment.

5. The method according to claim 2, characterized in that, The method further includes: The first time resource indicator data change graph, the second time resource indicator data change graph, and the third time resource indicator data change graph are sent to the user terminal; the user terminal is used to display the first time resource indicator data change graph, the second time resource indicator data change graph, and the third time resource indicator data change graph through a visual interface.

6. The method according to claim 5, characterized in that, For any current-time resource indicator data change chart, wherein the current-time resource indicator data change chart is any one of the first time resource indicator data change chart, the second time resource indicator data change chart, and the third time resource indicator data change chart, the method further includes: Obtain pre-set resource indicator data thresholds; If there is a resource indicator data in the current time resource indicator data change graph that is greater than or equal to the resource indicator data threshold, the resource indicator data will be identified as abnormal resource indicator data. The system generates an early warning message based on the abnormal resource indicator data and returns it to the user terminal. The user terminal is used to extract the abnormal resource indicator data from the early warning message, generate optimization suggestions based on the abnormal resource indicator data, highlight the abnormal resource indicator through the visualization interface, and display the corresponding optimization suggestions.

7. A cloud server resource demand prediction device, characterized in that, The device includes: The data acquisition module is used to collect cloud host resource monitoring indicator data up to the current moment; the cloud host resource monitoring indicator data includes CPU utilization and memory utilization. The data annotation module is used to annotate the cloud host resource monitoring indicator data according to a first dimension, a second dimension, and a third dimension to obtain first resource monitoring indicator data, second resource monitoring indicator data, and third resource monitoring indicator data; the dimension range of the first dimension is greater than the dimension range of the second dimension, and the dimension range of the second dimension is greater than the dimension range of the third dimension. The data analysis module is used to obtain first resource indicator data, second resource indicator data, and third resource indicator data using PromQL expressions based on the first resource monitoring indicator data, the second resource monitoring indicator data, and the third resource monitoring indicator data. The resource indicator data includes maximum CPU utilization, average CPU utilization, maximum memory utilization, and average memory utilization. The resource indicator data can be any one of the first resource indicator data, the second resource indicator data, and the third resource indicator data. The demand forecasting module is used to obtain the predicted cloud host resource demand in the first dimension, the predicted cloud host resource demand in the second dimension, and the predicted cloud host resource demand in the third dimension based on the first resource indicator data, the second resource indicator data, and the third resource indicator data.

8. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 6.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 6.

10. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 6.

Images