Computer system, measurement method thereof and medium

Abstract

The invention discloses a computer system and a measurement method thereof and a medium, and the method comprises the steps: determining KPIs in a specific application scene, and setting a specific performance threshold value; the method comprises the following steps: performing data acquisition on a virtual machine node of a computer system by using an AI-based dynamic sampling technology; filtering and cleaning the collected data, and automatically marking abnormal data points through a K-means clustering algorithm; performing real-time performance prediction based on an LSTM neural network, identifying a performance bottleneck, and performing partial preliminary analysis at a virtual machine end by adopting an edge computing technology; according to a real-time data analysis result, KPI index parameters are dynamically adjusted, and a prediction result and an actual operation condition are fed back to the LSTM neural network model; and generating a chart and an analysis report based on natural language processing by using a data visualization tool, and displaying a measurement analysis result. According to the invention, through an intelligent feedback mechanism, monitoring index parameters are dynamically adjusted, a prediction result is compared with an actual operation condition, and a monitoring model is continuously optimized, so that system monitoring is more accurate and efficient.

Description

Technical Field

[0001] The present invention belongs to the field of computer applications, and more specifically, particularly relates to a computer system and a measurement method and medium thereof. Background Art

[0002] With the rapid development of my country's national economy, information construction and national defense construction, people's needs for high-security and high-performance information security application products are becoming more and more urgent. The research and development of information security application products based on trusted computing will promote the informatization process of my country's economic society and safeguard national information security. In order to meet the needs of users with relatively high requirements for information security, the existing technology provides a new generation of trusted computing security application products with trusted computer platforms, trusted basic input and output systems, trusted storage and virtual machine monitoring technologies. Such trusted computing security application products generally require trust measurement of computer systems to improve their security.

[0003] The amount of data generated by modern systems during operation is extremely large, especially in high-concurrency environments. Traditional measurement methods have difficulty processing such large amounts of data, resulting in inefficient performance monitoring. System performance is affected by many factors, including hardware, software, and networks, resulting in complex multi-dimensional characteristics of data, and traditional single indicators cannot fully reflect the system status. In key business scenarios, the system has extremely high requirements for the real-time performance data. Traditional measurement methods usually have delays and cannot capture performance bottlenecks in time, resulting in delayed response after problems occur. As load changes and user demand fluctuates, the system's performance indicators need to be adjusted dynamically. Existing methods often cannot adapt to these changes in real time. In complex data sets, abnormal data points are often submerged and difficult to identify. Existing methods lack an effective anomaly detection mechanism, resulting in the inability to detect potential problems in a timely manner. Traditional analysis methods rely on manual judgment, lack automation and intelligence, reduce efficiency, and may lead to human errors. Summary of the invention

[0004] The purpose of this section is to summarize some aspects of embodiments of the present invention and briefly introduce some preferred embodiments. Some simplifications or omissions may be made in this section and the specification abstract and the invention title of this application to avoid blurring the purpose of this section, the specification abstract and the invention title, and such simplifications or omissions cannot be used to limit the scope of the present invention.

[0005] In view of the above or existing problems of the measurement method of the computer system, the present invention is proposed.

[0006] In order to solve the above technical problems, the present invention provides the following technical solutions: An embodiment of the present invention provides a computer system measurement method, including: clarifying the core goal of measurement, determining KPIs in specific application scenarios, and setting specific performance thresholds; deploying intelligent agents, using AI-based dynamic sampling technology, automatically adjusting the data collection frequency according to the load conditions, and performing data collection on virtual machine nodes of the computer system; filtering and cleaning the collected data, removing repeated, abnormal or erroneous measurement data, and automatically marking abnormal data points through a K-means clustering algorithm; performing real-time performance prediction based on an LSTM neural network, identifying performance bottlenecks, providing early warnings, and using edge computing technology to perform partial preliminary analysis on the virtual machine side; dynamically adjusting KPI indicator parameters according to real-time data analysis results, and using an intelligent feedback loop to feed back the prediction results and actual operating conditions to the LSTM neural network model; using data visualization tools to generate charts and analysis reports based on natural language processing to display measurement analysis results.

[0007] As a preferred solution of the computer system measurement method of the present invention, the core goal of the measurement is clarified, the KPIs in a specific application scenario are determined, and specific performance thresholds are set, including: According to different application scenarios, the following KPIs and their performance thresholds are set: if in a Web application, the performance threshold of response time is ≤200ms, the performance threshold of page loading time is ≤2s, and the performance threshold of concurrent users is ≥1000 users; if in a database system, the performance threshold of query response time is ≤100ms, the performance threshold of transaction throughput is ≥500 transactions / s, and the performance threshold of lock waiting time is ≤50ms; if in a cloud service platform, the performance threshold of CPU utilization is ≤70%, the performance threshold of memory utilization is ≤75%, and the performance threshold of network latency is ≤50 ms.

[0008] As a preferred solution of the computer system measurement method of the present invention, the deployment of intelligent agents uses AI-based dynamic sampling technology to automatically adjust the data collection frequency according to the load conditions and collect data on the virtual machine nodes of the computer system, including: Choose a Prometheus agent, making sure it supports VM monitoring and is compatible with different virtualization platforms. Perform the installation steps on each VM node: download the installation package of the smart agent, install it through the command line or automated script, and configure the agent service to start automatically at boot; Collect basic performance data on the virtual machine node to generate a historical data set, use time series analysis to analyze the historical data, and identify the load pattern. The time series analysis expression is as follows: in, is the value at the current time point, is a constant term, is the autoregression coefficient, is the autoregressive order, is white noise; Train models based on system load and performance indicators, predict future load conditions, monitor current system load and response time, and use trained AI models to automatically adjust sampling frequency based on real-time load data: If the load is high, the sampling frequency is increased; if the load is low, the sampling frequency is decreased.

[0009] As a preferred solution of the measurement method of the computer system of the present invention, the filtering and cleaning of the collected data, removing repeated, abnormal or erroneous measurement data, and automatically marking abnormal data points by the K-means clustering algorithm include: First, perform data standardization. The standardization formula is: Where X is the original data, μ and σ are the mean and standard deviation respectively; Set the number of clusters K=2, use sklearn.cluster.KMeans to train the clustering model, use the standardized data as input, and train the model to obtain K cluster centers; Compute the Euclidean distance between each data point and the center point of its cluster: in, Represents data points To cluster center The Euclidean distance of Represents data points The jth eigenvalue of Represents the cluster center The j-th eigenvalue of ; Data points with larger distances are considered outliers and are marked as "abnormal". After training, the model is used to cluster new data and automatically mark abnormal data points.

[0010] As a preferred solution of the computer system measurement method of the present invention, the real-time performance prediction based on the LSTM neural network, identification of performance bottlenecks, provision of early warning, and edge computing technology are used to perform partial preliminary analysis on the virtual machine side, including: In the edge computing module, the collected data is preprocessed: duplicate and abnormal data are removed, and the data is normalized so that each performance indicator is on the same scale; Convert the data into the time series format required by LSTM and define the time step, such as using the past 10 time point data to predict the next time point: Xt=[xt-10,xt-9,…,xt-1] Among them, Xt is the input feature at the current time point; Run the LSTM model regularly on the virtual machine to predict the latest collected data and generate future performance indicators; compare the prediction results with the set performance threshold to identify potential performance bottlenecks: If the predicted CPU usage exceeds 80%, it is considered a performance bottleneck. Once a bottleneck is identified, the system immediately triggers an alarm.

[0011] As a preferred solution of the measurement method of the computer system described in the present invention, the method includes: dynamically adjusting the KPI indicator parameters according to the real-time data analysis results, using an intelligent feedback loop to feed back the prediction results and the actual operation status to the LSTM neural network model, including: continuing to collect real-time performance data from virtual machine nodes, calculating the current KPI indicator in real time based on the collected data, defining the initial KPI indicator and its threshold when the system starts, and regularly evaluating whether the current KPI indicator meets the set threshold; if a certain KPI indicator continues to approach or exceed the threshold, dynamically adjusting the threshold of the indicator, such as when the CPU usage rate exceeds 80% continuously, adjusting it to 85% to adapt to the new load situation; designing a feedback mechanism in the LSTM model to compare the prediction results with the actual operation status, and after each prediction, collecting the prediction results and the actual performance data to form a feedback data set, including the predicted KPI value, the actual KPI value and the current system status data.

[0012] As a preferred solution of the measurement method of the computer system described in the present invention, the use of data visualization tools to generate charts and analysis reports based on natural language processing to display measurement analysis results includes: using NLP technology to automatically generate analysis reports based on real-time data analysis results, the report containing data summaries, trend analysis, anomaly detection results and optimization suggestions; creating a user-friendly dashboard to centrally display performance monitoring data, visualization charts and analysis reports.

[0013] A computer system, comprising: The performance indicator definition module is used to clarify the core objectives of the measurement, determine the KPIs for specific application scenarios, and set specific performance thresholds; The data collection module is used to deploy intelligent agents, use AI-based dynamic sampling technology, automatically adjust the data collection frequency according to the load conditions, and collect data on the virtual machine nodes of the computer system; The data processing module is used to filter and clean the collected data, remove duplicate, abnormal or erroneous measurement data, and automatically mark abnormal data points through the K-means clustering algorithm; The performance prediction module is used to make real-time performance predictions based on the LSTM neural network, identify performance bottlenecks, provide early warnings, and use edge computing technology to perform some preliminary analysis on the virtual machine side; The adjustment and optimization module is used to dynamically adjust KPI indicator parameters based on real-time data analysis results, and use intelligent feedback loops to feed back the prediction results and actual operation status to the LSTM neural network model; The report generation module is used to generate charts and analysis reports based on natural language processing using data visualization tools to display the measurement analysis results.

[0014] A computing device, comprising: at least one processor, memory, and input-output unit; The memory is used to store a computer program, and the processor is used to call the computer program stored in the memory to execute the steps of the measurement method of the computer system.

[0015] A computer-readable storage medium includes instructions, which, when executed on a computer, cause the computer to perform the steps of a computer system measurement method.

[0016] The beneficial effects of the present invention are as follows: The present invention utilizes AI-based dynamic sampling technology to automatically adjust the data collection frequency according to changes in the system load, ensuring the timeliness and accuracy of key performance data without affecting system performance, and greatly reducing unnecessary data transmission and storage. By performing some preliminary analysis on the virtual machine node and sinking the preliminary data processing to the edge device, the pressure on the central server is reduced and the overall data processing efficiency is improved. Real-time performance data analysis and prediction are performed through the LSTM neural network, potential performance bottlenecks are identified in advance, and accurate early warning notifications are provided, so that system administrators can take timely countermeasures to avoid failures. Through the intelligent feedback mechanism, the monitoring index parameters are dynamically adjusted, the prediction results are compared with the actual operating conditions, and the monitoring model is continuously optimized to make system monitoring more accurate and efficient. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative labor.

[0018] Figure 1 A flowchart of a computer system measurement method provided by an embodiment of the present invention.

[0019] Figure 2 A schematic diagram of a computer system structure provided for an embodiment of the present invention.

[0020] Figure 3 A schematic structural diagram of a medium according to an embodiment of the present invention is schematically shown.

[0021] Figure 4 A schematic diagram of the structure of a computing device according to an embodiment of the present invention is schematically shown.

[0022] In the drawings, the same or corresponding reference numerals represent the same or corresponding parts. DETAILED DESCRIPTION

[0023] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the accompanying drawings.

[0024] In the following description, many specific details are set forth to facilitate a full understanding of the present invention, but the present invention may also be implemented in other ways different from those described herein, and those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0025] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The term "in one embodiment" that appears in different places in this specification does not necessarily refer to the same embodiment, nor does it refer to a separate or selective embodiment that is mutually exclusive with other embodiments.

[0026] Example 1 Reference below Figure 1 , Figure 1 This is a flow chart of a computer system measurement method provided by an embodiment of the present invention. It should be noted that the embodiments of the present invention can be applied to any applicable scenario.

[0027] Figure 1 The process of the computer system measurement method provided by one embodiment of the present invention includes: S1: Clarify the core objectives of the measurement, determine the KPIs for specific application scenarios, and set specific performance thresholds.

[0028] Preferably, the following KPIs and their performance thresholds are set according to different application scenarios: If in a web application, the performance threshold of response time is ≤200ms, and the page loading time is The performance threshold is ≤ 2s, and the performance threshold of concurrent users is ≥ 1000 users; If in a database system, the performance threshold of query response time is ≤100ms, the performance threshold of transaction throughput is ≥500 transactions / s, and the performance threshold of lock waiting time is ≤50ms; If in the cloud service platform, the performance threshold of CPU usage is ≤70%, the performance threshold of memory usage is ≤75%, and the performance threshold of network delay is ≤50 ms.

[0029] S2: Deploy intelligent agents, use AI-based dynamic sampling technology, automatically adjust the data collection frequency according to the load conditions, and collect data on the virtual machine nodes of the computer system.

[0030] Preferably, choose a Prometheus agent, make sure the agent supports virtual machine monitoring and is compatible with different virtualization platforms; Perform the installation steps on each VM node: download the installation package of the smart agent, install it through the command line or automated script, and configure the agent service to start automatically at boot; Collect basic performance data on the virtual machine node to generate a historical data set, use time series analysis to analyze the historical data, and identify the load pattern. The time series analysis expression is as follows: in, is the value at the current time point, is a constant term, is the autoregression coefficient, is the autoregressive order, is white noise; Train models based on system load and performance indicators, predict future load conditions, monitor current system load and response time, and use trained AI models to automatically adjust sampling frequency based on real-time load data: If the load is high, the sampling frequency is increased; if the load is low, the sampling frequency is decreased.

[0031] Furthermore, add the IP address of the virtual machine to the Prometheus configuration file so that it can crawl data regularly, and set the crawling frequency to every 5 seconds; After starting Prometheus, use its web interface to view monitoring data, including CPU usage, memory usage, network traffic, etc. Prometheus automatically stores the collected data as time series to ensure long-term storage and query of data. Use Prometheus' query function to analyze historical data and identify load patterns. For example, analyze the changes in CPU usage in the past hour to view its trends and fluctuations; Extract historical data from Prometheus and organize it into a format suitable for model training. Select LSTM neural network as the load prediction model and train it using training data to ensure that the model can accurately predict future load conditions. Track performance indicators such as CPU, memory, and response time in real time, and set thresholds to determine load status. For example, when CPU usage exceeds 70%, it is considered high load; Automatically adjust data sampling frequency based on real-time monitoring results: Under high load (over 70%), increase the sampling frequency to collect data every 5 seconds; At low load (less than 30%), reduce the sampling frequency to collect data every 60 seconds; Configure alarm rules in Prometheus to monitor high load conditions. Once the CPU usage exceeds the set threshold, the system will automatically trigger an alarm.

[0032] S3: Filter and clean the collected data to remove duplicate, abnormal or erroneous measurement data, and automatically mark abnormal data points through the K-means clustering algorithm.

[0033] Preferably, data standardization is performed first, and the standardization formula is: Where X is the original data, μ and σ are the mean and standard deviation respectively; Set the number of clusters K=2, use sklearn.cluster.KMeans to train the clustering model, use the standardized data as input, and train the model to obtain K cluster centers; Compute the Euclidean distance between each data point and the center point of its cluster: in, Represents data points To cluster center The Euclidean distance of Represents data points The jth eigenvalue of Represents the cluster center The j-th eigenvalue of ; Data points with larger distances are considered outliers and are marked as "abnormal". After training, the model is used to cluster new data and automatically mark abnormal data points.

[0034] Furthermore, suppose there is a set of multidimensional data sets, each data point contains multiple feature values, as shown below: The standardized data: For example, if feature 1 has a mean of 5.0 and a standard deviation of 0.5, and feature 2 has a mean of 3.2 and a standard deviation of 0.3, then the standardized data would be: Set the number of clusters K to 2, use the standardized data as input, train the K-Means clustering model, and get two cluster centers. Assuming that the standardized eigenvalue of a data point is [1.0, -1.0] and the cluster center is [0.5, 0.5], calculate the distance:

[0035] Set a threshold, for example, if the distance exceeds a certain value (such as 2.0), it is marked as an anomaly. According to the calculated distance, the data points with a larger distance are marked as "anomalies"; for example, if the distance of a data point is 2.5, which exceeds the set threshold, it is marked as "anomaly".

[0036] S4: Real-time performance prediction based on LSTM neural network, identifying performance bottlenecks, providing early warning, and using edge computing technology to perform some preliminary analysis on the virtual machine side.

[0037] Preferably, in the edge computing module, the collected data is preprocessed: duplicate and abnormal data are removed, and the data is normalized so that each performance indicator is on the same scale; Convert the data into the time series format required by LSTM and define the time step, such as using the past 10 time point data to predict the next time point: Xt=[xt-10,xt-9,…,xt-1] Among them, Xt is the input feature at the current time point; Run the LSTM model regularly on the virtual machine to predict the latest collected data and generate future performance indicators; compare the prediction results with the set performance threshold to identify potential performance bottlenecks: If the predicted CPU usage exceeds 80%, it is considered a performance bottleneck. Once a bottleneck is identified, the system immediately triggers an alarm.

[0038] Further, for example, suppose we have the following data points: Normalize all performance indicators to the same scale to facilitate the training and prediction of the LSTM model, using the Min-Max normalization formula: Among them, X′ is the normalized data, X is the original data, min(X) is the minimum value in the data set, and max(X) is the maximum value in the data set; Processed data: Use the data of the past 10 time points as the input features of the LSTM model. For example, for time point t=11, the input features are: X11=[CPU usage (t-10), CPU usage (t-9), …, CPU usage (t-1)]; Run the trained LSTM model on the virtual machine regularly (e.g., every minute) to predict the latest collected data and generate future performance indicators (e.g., CPU usage at the next time point). For example, suppose the model predicts that the CPU usage at time point t=12 is 85%; The performance threshold is set to 80%. If the predicted CPU usage exceeds this threshold, it is considered a potential performance bottleneck. The predicted CPU usage is compared with the threshold: the predicted result is 85%>80% (threshold), so a performance bottleneck is identified.

[0039] S5: Dynamically adjust KPI indicator parameters based on real-time data analysis results, and use intelligent feedback loops to feed back prediction results and actual operating conditions to the LSTM neural network model.

[0040] Preferably, real-time performance data is continuously collected from the virtual machine nodes, and the current KPI indicators are calculated in real time based on the collected data. Initial KPI indicators and their thresholds are defined when the system is started, and whether the current KPI indicators meet the set thresholds is regularly evaluated; If a KPI indicator continues to approach or exceed the threshold, dynamically adjust the threshold of the indicator. For example, if the CPU usage rate exceeds 80% continuously, adjust it to 85% to adapt to the new load situation. A feedback mechanism is designed in the LSTM model to compare the prediction results with the actual operation status. After each prediction, the prediction results and actual performance data are collected to form a feedback data set, including the predicted KPI value, Actual KPI values ​​and current system status data.

[0041] Furthermore, when the system starts, the initial KPI indicators are set as follows: the initial threshold of CPU usage is 80%, the initial threshold of memory usage is 75%, and the initial threshold of response time is 200ms. Based on the collected real-time data, the current KPI indicators are calculated; for example, the average CPU usage in the last 5 minutes is calculated. Check every minute whether the current KPI indicator meets the set threshold: if the CPU usage exceeds 80% for three consecutive times (15 seconds interval), it is marked as close to the threshold; if a KPI indicator exceeds the set threshold for 5 minutes, it enters the dynamic adjustment stage; If the CPU usage rate exceeds 80% continuously, adjust its threshold to 85%. Similarly, other KPIs can be processed similarly: if the memory usage rate threshold exceeds 75%, adjust it to 80%; if the response time threshold exceeds 200ms, adjust it to 220ms; record the adjusted new thresholds and notify related system components for subsequent monitoring; In the LSTM model, a feedback loop is designed to collect comparison data between each prediction result and the actual operation status. The prediction accuracy of the LSTM model is analyzed based on the feedback data set. If the prediction result deviates greatly from the actual result (for example, more than 5%), consider retraining the model and adjusting the parameters or training data set.

[0042] S6: Use data visualization tools to generate charts and natural language processing-based analysis reports to present the results of metric analysis.

[0043] Preferably, NLP technology is used to automatically generate an analysis report based on the real-time data analysis results, which includes data summary, trend analysis, anomaly detection results and optimization suggestions; Create user-friendly dashboards that centrally display performance monitoring data, visualization charts, and analytical reports.

[0044] After introducing the method of the exemplary embodiment of the present invention, next, refer to Figure 2 A computer system according to an exemplary embodiment of the present invention is described, the system comprising: The performance indicator definition module is used to clarify the core objectives of the measurement, determine the KPIs for specific application scenarios, and set specific performance thresholds; The data collection module is used to deploy intelligent agents, use AI-based dynamic sampling technology, automatically adjust the data collection frequency according to the load conditions, and collect data on the virtual machine nodes of the computer system; The data processing module is used to filter and clean the collected data, remove duplicate, abnormal or erroneous measurement data, and automatically mark abnormal data points through the K-means clustering algorithm; The performance prediction module is used to make real-time performance predictions based on the LSTM neural network, identify performance bottlenecks, provide early warnings, and use edge computing technology to perform some preliminary analysis on the virtual machine side; The adjustment and optimization module is used to dynamically adjust KPI indicator parameters based on real-time data analysis results, and use intelligent feedback loops to feed back the prediction results and actual operation status to the LSTM neural network model; The report generation module is used to generate charts and analysis reports based on natural language processing using data visualization tools to display the measurement analysis results.

[0045] After introducing the method and system of the exemplary embodiment of the present invention, next, reference is made to Figure 3 For a description of a computer-readable storage medium according to an exemplary embodiment of the present invention, please refer to Figure 3 , the computer-readable storage medium shown is a CD 30, on which a computer program (i.e., a program product) is stored. When the computer program is executed by the processor, it will implement the steps recorded in the above method implementation method, for example, clarify the core objectives of the measurement, determine the KPIs in a specific application scenario, and set a specific performance threshold; deploy intelligent agents, use AI-based dynamic sampling technology, automatically adjust the data collection frequency according to the load situation, and collect data on the virtual machine nodes of the computer system; filter and clean the collected data, remove duplicate, abnormal or erroneous measurement data, and automatically mark abnormal data points through the K-means clustering algorithm; perform real-time performance prediction based on the LSTM neural network, identify performance bottlenecks, provide early warnings, and use edge computing technology to perform some preliminary analysis on the virtual machine side; dynamically adjust KPI indicator parameters based on real-time data analysis results, and use intelligent feedback loops to feed back the prediction results and actual operation conditions to the LSTM neural network model; use data visualization tools to generate charts and analysis reports based on natural language processing to display the measurement analysis results; the specific implementation methods of each step will not be repeated here.

[0046] It should be noted that examples of the computer-readable storage medium may also include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other optical or magnetic storage media, which are not listed here one by one.

[0047] After introducing the method, system and medium of the exemplary embodiment of the present invention, next, reference is made to Figure 4 A computing device for measuring the computer system of an exemplary embodiment of the present invention.

[0048] Figure 4 A block diagram of an exemplary computing device 40 suitable for implementing embodiments of the present invention is shown, and the computing device 40 may be a computer system or a server. Figure 4 The computing device 40 shown is only an example and should not bring any limitation to the functionality and scope of use of the embodiments of the present invention.

[0049] like Figure 4 As shown, the components of the computing device 40 may include, but are not limited to: one or more processors or processing units 401 , a system memory 402 , and a bus 403 connecting different system components (including the system memory 402 and the processing unit 401 ).

[0050] The computing device 40 typically includes a variety of computer system readable media. These media can be any available media that can be accessed by the computing device 40, including volatile and non-volatile media, removable and non-removable media.

[0051] The system memory 402 may include computer system readable media in the form of volatile memory, such as random access memory (RAM) 4021 and / or cache memory 4022. The computing device 40 may further include other removable / non-removable, volatile / non-volatile computer system storage media. By way of example only, the ROM 4023 may be used to read and write non-removable, non-volatile magnetic media ( Figure 4 is not shown in the Figure 4 As shown in FIG. 4 , a disk drive for reading and writing a removable non-volatile disk (e.g., a “floppy disk”) and an optical disk drive for reading and writing a removable non-volatile optical disk (e.g., a CD-ROM, a DVD-ROM, or other optical media) may be provided. In these cases, each drive may be connected to bus 403 via one or more data medium interfaces. System memory 402 may include at least one program product having a set (e.g., at least one) of program modules that are configured to perform the functions of various embodiments of the present invention.

[0052] A program / utility 4025 having a set (at least one) of program modules 4024 may be stored, for example, in system memory 402, and such program modules 4024 include, but are not limited to, an operating system, one or more application programs, other program modules, and program data, each of which or some combination may include an implementation of a network environment. Program modules 4024 generally perform the functions and / or methods of the embodiments described herein.

[0053] The computing device 40 may also communicate with one or more external devices 404 (e.g., a keyboard, a pointing device, a display, etc.). Such communication may be performed via an input / output (I / O) interface 405. Furthermore, the computing device 40 may also communicate with one or more networks (e.g., a local area network (LAN), a wide area network (WAN), and / or a public network, such as the Internet) via a network adapter 406. Figure 4 As shown, the network adapter 406 communicates with other modules (such as the processing unit 401, etc.) of the computing device 40 via the bus 403. It should be understood that although Figure 4 Not shown, other hardware and / or software modules may be used in conjunction with computing device 40 .

[0054] The processing unit 401 executes various functional applications and data processing by running the programs stored in the system memory 402, for example, defining the core objectives of the measurement, determining the KPIs in a specific application scenario, and setting specific performance thresholds; deploying intelligent agents, using AI-based dynamic sampling technology, automatically adjusting the data collection frequency according to the load conditions, and collecting data on the virtual machine nodes of the computer system; Filter and clean the collected data to remove duplicate, abnormal or erroneous measurement data, and automatically mark abnormal data points through the K-means clustering algorithm; perform real-time performance prediction based on the LSTM neural network, identify performance bottlenecks, provide early warnings, and use edge computing technology to perform some preliminary analysis on the virtual machine side; dynamically adjust KPI indicator parameters based on real-time data analysis results, and use intelligent feedback loops to feed back prediction results and actual operating conditions to the LSTM neural network model; use data visualization tools to generate charts and analysis reports based on natural language processing to display measurement analysis results.

[0055] The specific implementation method of each step will not be repeated here. It should be noted that although several units / modules or sub-units / sub-modules of the synchronous escape wiring device based on multiple commodity flows are mentioned in the above detailed description, this division is merely exemplary and not mandatory. In fact, according to an embodiment of the present invention, the features and functions of two or more units / modules described above can be concretized in one unit / module. Conversely, the features and functions of one unit / module described above can be further divided to be concretized by multiple units / modules.

[0056] In the description of the present invention, it should be noted that the terms “first”, “second” and “third” are only used for descriptive purposes and cannot be understood as indicating or implying relative importance.

[0057] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0058] In the several embodiments provided by the present invention, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. The device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some communication interfaces, and the indirect coupling or communication connection of devices or units can be electrical, mechanical or other forms.

[0059] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0060] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0061] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a non-volatile computer-readable storage medium that can be executed by a processor. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, etc., various media that can store program codes.

[0062] Finally, it should be noted that the above-described embodiments are only specific implementations of the present invention, which are used to illustrate the technical solutions of the present invention, rather than to limit them. The protection scope of the present invention is not limited thereto. Although the present invention is described in detail with reference to the above-described embodiments, ordinary technicians in the field should understand that any technician familiar with the technical field can still modify the technical solutions recorded in the above-described embodiments within the technical scope disclosed by the present invention, or can easily think of changes, or make equivalent replacements for some of the technical features therein; and these modifications, changes or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the protection scope of the claims.

[0063] In addition, although the operations of the method of the present invention are described in a specific order in the drawings, this does not require or imply that the operations must be performed in this specific order, or that all the operations shown must be performed to achieve the desired results. Additionally or alternatively, some steps may be omitted, multiple steps may be combined into one step, and / or one step may be decomposed into multiple steps.

Claims

1. A computer system measurement method, characterized in that: include: Clarify the core objectives of the measurement, determine the KPIs for specific application scenarios, and set specific performance thresholds; Deploy intelligent agents, use AI-based dynamic sampling technology to automatically adjust the data collection frequency according to load conditions, and collect data on the virtual machine nodes of the computer system; Filter and clean the collected data to remove duplicate, abnormal or erroneous measurement data, and automatically mark abnormal data points through the K-means clustering algorithm; Based on LSTM neural network, real-time performance prediction is carried out to identify performance bottlenecks, provide early warning, and use edge computing technology to perform some preliminary analysis on the virtual machine side; Dynamically adjust KPI indicator parameters based on real-time data analysis results, and use intelligent feedback loops to feed back prediction results and actual operating conditions to the LSTM neural network model; Use data visualization tools to generate charts and natural language processing-based analysis reports to present metric analysis results.

2. The computer system measurement method according to claim 1, characterized in that: The core objectives of the above are to clearly define the measurement, determine the KPIs for specific application scenarios, and set specific performance thresholds, including: According to different application scenarios, set the following KPIs and their performance thresholds: If in a web application, the performance threshold of response time is ≤200ms, and the page loading time is The performance threshold is ≤ 2s, and the performance threshold of concurrent users is ≥ 1000 users; If in a database system, the performance threshold of query response time is ≤100ms, the performance threshold of transaction throughput is ≥500 transactions / s, and the performance threshold of lock waiting time is ≤50ms; If in the cloud service platform, the performance threshold of CPU usage is ≤70%, the performance threshold of memory usage is ≤75%, and the performance threshold of network delay is ≤50 ms.

3. The computer system measurement method according to claim 1, characterized in that: The deployment of intelligent agents uses AI-based dynamic sampling technology to automatically adjust the data collection frequency according to the load conditions and collect data on the virtual machine nodes of the computer system, including: Choose a Prometheus agent, making sure it supports VM monitoring and is compatible with different virtualization platforms. Perform the installation steps on each VM node: download the installation package of the smart agent, install it through the command line or automated script, and configure the agent service to start automatically at boot; Collect basic performance data on the virtual machine node to generate a historical data set, use time series analysis to analyze the historical data, and identify the load pattern. The time series analysis expression is as follows: in, is the value at the current time point, is a constant term, is the autoregression coefficient, is the autoregressive order, is white noise; Train models based on system load and performance indicators, predict future load conditions, monitor current system load and response time, and use trained AI models to automatically adjust sampling frequency based on real-time load data: If the load is high, the sampling frequency is increased; if the load is low, the sampling frequency is decreased.

4. The computer system measurement method according to claim 1, characterized in that: The collected data is filtered and cleaned to remove duplicate, abnormal or erroneous measurement data, and abnormal data points are automatically marked by the K-means clustering algorithm, including: First, perform data standardization. The standardization formula is: Where X is the original data, μ and σ are the mean and standard deviation respectively; Set the number of clusters K=2, use sklearn.cluster.KMeans to train the clustering model, use the standardized data as input, and train the model to obtain K cluster centers; Compute the Euclidean distance between each data point and the center point of its cluster: in, Represents data points To cluster center The Euclidean distance of Represents data points The jth eigenvalue of Represents the cluster center The j-th eigenvalue of ; Data points with larger distances are considered outliers and are marked as "abnormal". After training, the model is used to cluster new data and automatically mark abnormal data points.

5. The computer system measurement method according to claim 1, characterized in that: The LSTM neural network is used to predict real-time performance, identify performance bottlenecks, provide early warnings, and use edge computing technology to perform some preliminary analysis on the virtual machine side, including: In the edge computing module, the collected data is preprocessed: duplicate and abnormal data are removed, and the data is normalized so that each performance indicator is on the same scale; Convert the data into the time series format required by LSTM and define the time step, such as using the past 10 time point data to predict the next time point: X t =[xt-10,xt-9,…,xt-1] Among them, X t is the input feature at the current time point; Run the LSTM model regularly on the virtual machine to predict the latest collected data and generate future performance indicators; compare the prediction results with the set performance threshold to identify potential performance bottlenecks: If the predicted CPU usage exceeds 80%, it is considered a performance bottleneck. Once a bottleneck is identified, the system immediately triggers an alarm.

6. The computer system measurement method according to claim 1, characterized in that: The KPI indicator parameters are dynamically adjusted according to the real-time data analysis results, and the prediction results and actual operation conditions are fed back to the LSTM neural network model using an intelligent feedback loop, including: Continue to collect real-time performance data from virtual machine nodes, calculate the current KPI indicators in real time based on the collected data, define the initial KPI indicators and their thresholds when the system starts, and regularly evaluate whether the current KPI indicators meet the set thresholds; If a KPI indicator continues to approach or exceed the threshold, dynamically adjust the threshold of the indicator. For example, if the CPU usage rate exceeds 80% continuously, adjust it to 85% to adapt to the new load situation. A feedback mechanism is designed in the LSTM model to compare the prediction results with the actual operation status. After each prediction, the prediction results and actual performance data are collected to form a feedback data set, including the predicted KPI value, Actual KPI values ​​and current system status data.

7. The computer system measurement method according to claim 1, characterized in that: The data visualization tool is used to generate charts and analysis reports based on natural language processing to display the measurement analysis results, including: Using NLP technology, analysis reports are automatically generated based on real-time data analysis results. The reports include data summaries, trend analysis, anomaly detection results, and optimization suggestions. Create user-friendly dashboards that centrally display performance monitoring data, visualization charts, and analytical reports.

8. A computer system, characterized in that: include: The performance indicator definition module is used to clarify the core objectives of the measurement, determine the KPIs for specific application scenarios, and set specific performance thresholds; The data collection module is used to deploy intelligent agents, use AI-based dynamic sampling technology, automatically adjust the data collection frequency according to the load conditions, and collect data on the virtual machine nodes of the computer system; The data processing module is used to filter and clean the collected data, remove duplicate, abnormal or erroneous measurement data, and automatically mark abnormal data points through the K-means clustering algorithm; The performance prediction module is used to make real-time performance predictions based on the LSTM neural network, identify performance bottlenecks, provide early warnings, and use edge computing technology to perform some preliminary analysis on the virtual machine side; The adjustment and optimization module is used to dynamically adjust KPI indicator parameters based on real-time data analysis results, and use intelligent feedback loops to feed back the prediction results and actual operation status to the LSTM neural network model; The report generation module is used to generate charts and analysis reports based on natural language processing using data visualization tools to display the measurement analysis results.

9. A computing device, comprising: at least one processor, memory, and input-output unit; The memory is used to store a computer program, and the processor is used to call the computer program stored in the memory to execute the steps of the measurement method of the computer system as described in any one of claims 1 to 7.

10. A computer-readable storage medium comprising instructions, which, when executed on a computer, enables the computer to execute the steps of the computer system measurement method according to any one of claims 1 to 7.

Images