Cloud storage system performance test method and system based on large model

Generate diverse test data and complex workloads through large-scale pre-training models, which solves the problem that traditional cloud storage performance testing methods cannot fully simulate practical application scenarios, improves the accuracy and comprehensiveness of the test, and provides detailed performance reports. Helps optimize cloud storage systems.

CN119088628APending Publication Date: 2024-12-06SHANDONG LANGCHAO YUNTOU INFORMATION TECH CO LTD
View PDF 0 Cites 2 Cited by

Patent Information

Application Number
CN202411196570.2
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-29
Publication Date
2024-12-06

AI Technical Summary

Technical Problem

Traditional cloud storage performance testing methods cannot generate diverse test data and complex workloads, and it is difficult to fully simulate practical application scenarios, resulting in insufficient accuracy and comprehensiveness of test results, and it is difficult to cope with rapidly changing application needs.

Method used

Using large-scale pre-trained models to generate diverse test data and complex workloads, analyze data and workload characteristics, develop generation strategies to ensure the diversity and representation of data and workloads, and Perform tests on cloud storage systems and record performance metrics for analysis.

Benefits of technology

It improves the accuracy and comprehensiveness of performance testing, can simulate practical application scenarios more realistically, provide detailed performance reports, help optimize and improve cloud storage systems, and improve the quality and user experience of cloud storage services.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119088628A_ABST
    Figure CN119088628A_ABST
Patent Text Reader

Abstract

The invention discloses a cloud storage system performance test method and system based on a large model, belongs to the technical field of cloud storage, and aims to solve the technical problems of how to generate diversified test data and complex workload, comprehensively simulate an actual application scene and improve the accuracy and comprehensiveness of performance test. According to the technical scheme, the method comprises the following steps: generating data: generating diversified test data by utilizing a large model pre-training model; generating a workload: defining an operation mode, and generating a corresponding workload to simulate the operation mode in a practical application scene; test execution: executing the generated test data and workload on the cloud storage system, and recording performance indexes; and performance analysis: analyzing the recorded performance indexes, and generating a detailed performance report. The system comprises a data generation module, a workload generation module, a test execution module and a performance analysis module.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of cloud storage technology, and in particular to a cloud storage system performance testing method and system based on a large model. Background Art

[0002] With the rapid development of cloud computing technology, cloud storage systems, as an important part of cloud computing, have been widely used in various data storage and management scenarios. Cloud storage systems support enterprises and individual users to store, manage and retrieve large amounts of data by providing high-availability, high-reliability and high-scalability data storage services. However, with the rapid growth of data volume and the increasing complexity of application scenarios, the performance of cloud storage systems has become one of the focuses of users' attention.

[0003] Traditional cloud storage performance testing methods are usually based on predefined test data and workloads. These methods mainly include the following:

[0004] Benchmarking: By running a series of standardized test cases, the performance of the cloud storage system under different loads is measured. These test cases usually cover basic storage operations, such as reading, writing, and deleting. Although benchmarking can provide a certain performance reference, its limitation is that it cannot fully reflect the complex operations and data characteristics in actual application scenarios.

[0005] Simulation Testing: The performance of cloud storage systems is evaluated by simulating data and operation modes in actual application scenarios. Simulation testing can better reflect the performance in specific application scenarios, but because the generation of simulated data and workloads relies on experience and assumptions, the accuracy and comprehensiveness of the test results are still limited.

[0006] Real Data Analysis: Evaluate the performance of cloud storage systems by analyzing data and operation logs generated in actual applications. This method can provide very realistic performance data, but it is challenging in data collection and processing, and it is difficult to reproduce specific test scenarios.

[0007] The above traditional performance testing methods have the following problems and challenges in practical applications:

[0008] Insufficient data diversity: The test data generated by traditional methods are usually of a single type and cannot fully cover the various data types (such as text, images, videos, etc.) and data distribution characteristics that exist in actual applications.

[0009] Insufficient workload complexity: The workload patterns generated by traditional methods are often relatively simple and cannot simulate the complex operations and high-concurrency scenarios in actual applications.

[0010] Insufficient accuracy and comprehensiveness of test results: Since the generation of data and workloads relies on predefined rules and assumptions, the test results of traditional methods are limited in accuracy and comprehensiveness, and it is difficult to reflect the performance in real application scenarios.

[0011] Difficult to cope with rapidly changing application requirements: With the rapid changes in application requirements and the continuous development of technology, cloud storage systems are facing more and more new performance challenges. Traditional performance testing methods are difficult to adapt and adjust in a timely and effective manner.

[0012] Therefore, how to generate diverse test data and complex workloads, fully simulate actual application scenarios, and improve the accuracy and comprehensiveness of performance testing is a technical problem that needs to be solved urgently. Summary of the invention

[0013] The technical task of the present invention is to provide a cloud storage system performance testing method and system based on a large model to solve the problem of how to generate diversified test data and complex workloads, comprehensively simulate actual application scenarios, and improve the accuracy and comprehensiveness of performance testing.

[0014] The technical task of the present invention is achieved in the following way: a cloud storage system performance testing method based on a large model, the method is as follows:

[0015] Generate data: Use large model pre-training models (such as the Hairuo large model) to generate diverse test data;

[0016] Generate workloads: define the operation mode and generate corresponding and complex workloads to simulate the operation mode in actual application scenarios;

[0017] Test execution: Execute the generated test data and workload on the cloud storage system and record performance metrics;

[0018] Performance analysis: Analyze the recorded performance indicators and generate detailed performance reports.

[0019] As a preference, the generated data is specifically as follows:

[0020] Define data types: Based on actual business scenarios, define the text, image, video, and audio data types that need to be generated; specifically: first identify and classify the data requirements in the actual application scenario; then list all possible data types and create a detailed description document for each type; finally determine the characteristics of each data type: text data requires different lengths and languages, and image data requires different resolutions and color depths;

[0021] Analyze data characteristics: Use large-scale pre-trained models to analyze the data characteristics of data distribution, size, and format in actual application scenarios; specifically: first analyze the existing data set through the pre-trained model to extract feature distribution, data format, and size information; then use statistical analysis methods to quantify and classify data characteristics; finally, write an analysis report that describes the characteristics of each data type in detail;

[0022] Formulate data generation strategy: Based on the results of data feature analysis, formulate data generation strategy to ensure that the generated data is diverse and representative. Specifically, first formulate diversified data generation plans based on the analysis results to ensure wide coverage of data; then formulate data generation rules and parameters to ensure that the generated data meets the expected distribution and characteristics; finally, use the pre-trained model to configure the generation strategy, which includes parameter adjustment and data enhancement technology;

[0023] Generate data: call the pre-trained model to generate diversified test data and store the data in the cloud storage system; specifically: first use the pre-trained model to generate data according to the strategy; then verify the quality of the generated data to ensure that the generated data meets the defined characteristics; then use the cloud storage system (such as AWS S3, Google Cloud Storage or Azure BlobStorage) to upload the generated data to the cloud storage system; finally, set appropriate access control and backup policies to ensure the security and high availability of data in the cloud storage system.

[0024] As a preferred method, the workload is generated as follows:

[0025] Define operation modes: Define multiple operation modes such as read, write, update, and delete, as well as combined operations and high-concurrency scenarios according to actual application scenarios; specifically: first, through in-depth analysis of actual applications, define detailed descriptions of basic operations and combined operations; then determine the operation mode in high-concurrency scenarios and write detailed operation procedures;

[0026] Analyze workload characteristics: Use large-scale pre-trained models to analyze workload characteristics in actual applications. Workload characteristics include operation frequency, operation sequence, and operation dependency. Specifically, first collect and analyze the operation logs of existing systems to extract workload characteristics. Then use pre-trained models to perform feature analysis to identify frequency, sequence, and dependency. Finally, write an analysis report to describe the detailed characteristics of the workload.

[0027] Workload generation strategy: Based on the results of workload characteristic analysis, a workload generation strategy is formulated to ensure that the generated workload is complex and representative. Specifically, a workload generation strategy covering multiple operation modes is first formulated; then the parameters and rules for generating workloads are determined to ensure their complexity and representativeness; finally, the strategy is adjusted to suit different test requirements and scenarios.

[0028] Generate workload: call the pre-trained model to generate complex workloads, and apply the workloads to the cloud storage system; specifically: first use the pre-trained model to generate workload scripts and tasks; then verify the generated workload to ensure that it meets the expected characteristics; then deploy the workload tasks to the cloud storage system and prepare for execution; finally, configure the relevant parameters of the cloud storage system (such as AWS S3, Google Cloud Storage, or Azure Blob Storage) to ensure that the workload can be executed correctly and record operation logs.

[0029] Preferably, the test is performed as follows:

[0030] Test environment preparation: Configure the test environment of the cloud storage system, which includes hardware resources, software configuration and network settings. Specifically, first configure the corresponding cloud storage resources (such as computing nodes and storage devices) according to the test requirements; then install and configure the necessary software, including the operating system, database and other dependent components; then set up the network environment of the cloud storage system to ensure the stability and scalability of the test environment; finally, use the cloud service provider's management console or CLI tool to configure and manage resources;

[0031] Test task scheduling: Schedule test tasks according to the predefined test plan to ensure that the test tasks are executed in the set order and time. Specifically, first write a test plan to describe the order and time schedule of the test tasks in detail; then use automation tools (such as AWS Batch, Google Cloud Scheduler, or Azure Logic Apps) to schedule test tasks to ensure that the tasks are executed as planned; finally, monitor the progress of tasks in real time and adjust the task scheduling in a timely manner;

[0032] Test data loading: Load the generated test data into the cloud storage system and ensure the integrity and consistency of the data. Specifically, first migrate the generated data from the local or other storage system to the cloud storage; then verify the integrity and consistency of the loaded data to ensure that the data is correct; finally, use scripts or automation tools (such as AWS Data Sync, Google Transfer Appliance, or Azure Data Box) to automate the data loading process to improve efficiency.

[0033] Workload execution: Execute the generated workload on the cloud storage system and record the operation logs and performance indicators in real time. Specifically, start the workload task first and monitor its execution. Then use cloud monitoring tools (such as AWS CloudWatch, Google Stackdriver, or Azure Monitor) to record the details of each operation. Finally, collect performance indicators, including response time, throughput, and resource utilization.

[0034] Exception handling: During the test process, any abnormalities in the monitoring system status are promptly handled to ensure the stability and continuity of the test process; specifically: first monitor the system status in real time and use the alarm system to detect abnormalities; then develop a detailed exception handling process to ensure rapid response and repair; finally, record the abnormal situation and handling process for subsequent analysis and improvement.

[0035] Preferably, the performance analysis is as follows:

[0036] Data collection: Obtain operation logs and performance indicators from the test execution process, including response time, throughput, and resource utilization. Specifically, first use data collection tools (such as AWS CloudWatch Logs, Google Cloud Logging, or Azure Log Analytics) to collect operation logs and performance indicators; then ensure the integrity and accuracy of the data to prevent data loss or tampering; finally, collect data regularly for continuous monitoring and analysis;

[0037] Data processing: pre-process the collected data; specifically: first clean the data to remove invalid or erroneous data; then convert the data format to adapt the data format to the requirements of the analysis tool; finally aggregate the data to generate necessary statistical indicators;

[0038] Performance analysis: Based on the preprocessed data, we conduct performance analysis to evaluate the performance of the cloud storage system under different workloads and identify performance bottlenecks and optimization points. Specifically, we first use analytical tools and models to evaluate system performance (such as AWS QuickSight, Google Data Studio, or Azure Power BI); then identify performance bottlenecks, which include long response time and low throughput; and finally, we make optimization suggestions to help improve the system.

[0039] Report generation: Based on the analysis results, a performance report is written to describe the test background in detail, and then a detailed performance report is generated; the detailed performance report includes the test background, test methods, test results, performance evaluation and optimization suggestions.

[0040] A cloud storage system performance testing system based on a large model, the system comprising:

[0041] The data generation module is used to generate diversified test data using a large model pre-training model (such as the Hairuo large model);

[0042] The workload generation module is used to define the operation mode and generate corresponding and complex workloads to simulate the operation mode in the actual application scenario;

[0043] A test execution module, used to execute the generated test data and workload on the cloud storage system and record performance indicators;

[0044] The performance analysis module is used to analyze the recorded performance indicators and generate detailed performance reports.

[0045] Preferably, the data generation module comprises:

[0046] The data type definition submodule is used to define the text, image, video and audio data types that need to be generated according to the actual business scenarios; the working process of the data type definition submodule is as follows:

[0047] ① Identify and classify data requirements in actual application scenarios;

[0048] ② List all possible data types and create detailed description documents for each type;

[0049] ③ Determine the characteristics of each data type: text data requires different lengths and languages, image data requires different resolutions and color depths;

[0050] The data characteristic analysis submodule is used to analyze the data characteristics of data distribution, size and format in actual application scenarios using a large-scale pre-trained model; the working process of the data characteristic analysis submodule is as follows:

[0051] ① Analyze existing data sets through pre-trained models to extract feature distribution, data format and size information;

[0052] ② Use statistical analysis methods to quantify and classify data characteristics;

[0053] ③Write an analysis report, describing in detail the characteristics of each data type;

[0054] The data generation strategy formulation submodule is used to formulate data generation strategies based on the data characteristic analysis results to ensure that the generated data is diverse and representative. The working process of the data generation strategy formulation submodule is as follows:

[0055] ① Develop diversified data generation plans based on the analysis results to ensure extensive data coverage;

[0056] ② Formulate rules and parameters for data generation to ensure that the generated data conforms to the expected distribution and characteristics;

[0057] ③ Use pre-trained models to configure the generation strategy, which includes parameter adjustment and data enhancement techniques;

[0058] The data generation submodule is used to call the pre-trained model to generate diversified test data and store the data in the cloud storage system. The working process of the data generation submodule is as follows:

[0059] ①Use pre-trained models to generate data according to the strategy;

[0060] ②Verify the quality of the generated data to ensure that the generated data meets the defined characteristics;

[0061] ③Use cloud storage systems (such as AWS S3, Google Cloud Storage, or Azure Blob Storage) to upload the generated data to a cloud storage system;

[0062] ④ Set up appropriate access control and backup strategies to ensure the security and high availability of data in the cloud storage system;

[0063] The workload generation module comprises:

[0064] The operation mode definition submodule is used to define multiple operation modes of reading, writing, updating and deleting as well as combined operations and high concurrency scenarios according to actual application scenarios; the working process of the operation mode definition submodule is as follows:

[0065] ① Through in-depth analysis of actual applications, define detailed descriptions of basic operations and combined operations;

[0066] ②Determine the operation mode in high-concurrency scenarios and write detailed operation procedures;

[0067] The workload characteristics analysis submodule is used to analyze the workload characteristics in actual applications using a large-scale pre-trained model; the workload characteristics include operation frequency, operation sequence, and operation dependency; the specific working process of the workload characteristics analysis submodule is as follows:

[0068] ① Collect and analyze the operation logs of existing systems to extract workload characteristics;

[0069] ② Use pre-trained models for feature analysis to identify frequency, sequence, and dependencies;

[0070] ③ Write an analysis report describing the detailed characteristics of the workload;

[0071] The workload generation strategy submodule is used to formulate a workload generation strategy based on the workload characteristic analysis results to ensure that the generated workload is complex and representative. The working process of the workload generation strategy submodule is as follows:

[0072] ① Develop workload generation strategies that cover multiple operating modes;

[0073] ② Determine the parameters and rules for generating workloads to ensure their complexity and representativeness;

[0074] ③Adjust strategies to suit different testing needs and scenarios;

[0075] The workload generation submodule is used to call the pre-trained model to generate complex workloads and apply the workloads to the cloud storage system. The working process of the workload generation submodule is as follows:

[0076] ① Generate workload scripts and tasks using pre-trained models;

[0077] ② Verify the generated workload to ensure it meets the expected characteristics;

[0078] ③Deploy the workload tasks to the cloud storage system and prepare for execution;

[0079] ④Configure relevant parameters of the cloud storage system (such as AWS S3, Google Cloud Storage, or Azure Blob Storage) to ensure that the workload can be executed correctly and record operation logs.

[0080] Preferably, the test execution module includes:

[0081] The test environment preparation submodule is used to configure the test environment of the cloud storage system. The test environment of the cloud storage system includes hardware resources, software configuration and network settings. The working process of the test environment preparation submodule is as follows:

[0082] ①Configure corresponding cloud storage resources (such as computing nodes and storage devices) according to test requirements;

[0083] ②Install and configure necessary software, including operating system, database and other dependent components;

[0084] ③ Set up the network environment of the cloud storage system to ensure the stability and scalability of the test environment;

[0085] ④ Use the cloud service provider’s management console or CLI tools to configure and manage resources;

[0086] The test task scheduling submodule is used to schedule test tasks according to the predefined test plan to ensure that the test tasks are executed in the set order and time; the working process of the test task scheduling submodule is as follows:

[0087] ① Write a test plan, describing in detail the sequence and timing of test tasks;

[0088] ② Use automation tools (such as AWS Batch, Google Cloud Scheduler, or Azure LogicApps) to schedule test tasks and ensure that tasks are executed as planned;

[0089] ③ Monitor task progress in real time and adjust task scheduling in a timely manner;

[0090] The test data loading submodule is used to load the generated test data into the cloud storage system and ensure the integrity and consistency of the data. The working process of the test data loading submodule is as follows:

[0091] ① Migrate the generated data from local or other storage systems to cloud storage;

[0092] ②Verify the integrity and consistency of data loading to ensure that the data is correct;

[0093] ③ Use scripts or automation tools (such as AWS Data Sync, Google Transfer Appliance, or Azure Data Box) to automate the data loading process and improve efficiency;

[0094] The workload execution submodule is used to execute the generated workload on the cloud storage system and record the operation logs and performance indicators in real time. The working process of the workload execution submodule is as follows:

[0095] ① Start workload tasks and monitor their execution;

[0096] ② Use cloud monitoring tools (such as AWS CloudWatch, Google Stackdriver, or Azure Monitor) to record detailed information for each operation;

[0097] ③ Collect performance indicators, including response time, throughput and resource utilization;

[0098] The exception handling submodule is used to handle abnormal situations in the monitoring system status in a timely manner during the test process to ensure the stability and continuity of the test process; the working process of the exception handling submodule is as follows:

[0099] ① Monitor system status in real time and use the alarm system to detect abnormal situations;

[0100] ② Develop a detailed exception handling process to ensure rapid response and repair;

[0101] ③Record abnormal situations and handling processes for subsequent analysis and improvement;

[0102] The performance analysis module includes:

[0103] The data acquisition submodule is used to obtain operation logs and performance indicators from the test execution process. The operation logs and performance indicators include response time, throughput and resource utilization. The working process of the data acquisition submodule is as follows:

[0104] ① Use data collection tools (such as AWS CloudWatch Logs, Google Cloud Logging, or AzureLog Analytics) to collect operation logs and performance metrics;

[0105] ② Ensure the integrity and accuracy of data and prevent data loss or tampering;

[0106] ③ Collect data regularly for continuous monitoring and analysis;

[0107] The data processing submodule is used to pre-process the collected data; the working process of the data processing submodule is as follows:

[0108] ① Clean the data and remove invalid or erroneous data;

[0109] ②Convert the data format to adapt it to the requirements of the analysis tool;

[0110] ③ Aggregate data and generate necessary statistical indicators;

[0111] The performance analysis submodule is used to perform performance analysis based on preprocessed data, evaluate the performance of the cloud storage system under different workloads, and identify performance bottlenecks and optimization points. The working process of the performance analysis submodule is as follows:

[0112] ① Use analytical tools and models to evaluate system performance (such as AWS QuickSight, Google Data Studio, or Azure Power BI);

[0113] ②Identify performance bottlenecks, including long response time and low throughput;

[0114] ③Propose optimization suggestions to help improve the system;

[0115] The report generation submodule is used to write a performance report based on the analysis results, generate a detailed description of the test background, and then generate a detailed performance report; wherein the detailed performance report includes the test background, test methods, test results, performance evaluation and optimization suggestions.

[0116] An electronic device comprising: a memory and at least one processor;

[0117] Wherein, the memory stores computer-executable instructions;

[0118] The at least one processor executes the computer-executable instructions stored in the memory, so that the at least one processor executes the cloud storage system performance testing method based on the large model as described above.

[0119] A computer-readable storage medium stores computer-executable instructions. When a processor executes the computer-executable instructions, the cloud storage system performance testing method based on a large model as described above is implemented.

[0120] The cloud storage system performance testing method and system based on a large model of the present invention have the following advantages:

[0121] (1) The present invention uses a large-scale pre-trained model to generate diverse test data and complex workloads, fully simulates actual application scenarios, improves the accuracy and comprehensiveness of performance testing, and provides strong support for the optimization and improvement of cloud storage systems;

[0122] (ii) The present invention utilizes large-scale pre-trained models (such as the Hairuo large model) to generate diverse test data and complex workloads to simulate real application scenarios, thereby improving the accuracy and comprehensiveness of performance testing. It is suitable for cloud storage service providers, data center operation and maintenance personnel, and enterprise IT departments to perform performance evaluation and optimization of cloud storage systems, and can provide more realistic and valuable performance test results.

[0123] (III) The present invention significantly improves the accuracy and comprehensiveness of cloud storage system performance testing by using large-scale pre-trained models to generate diverse test data and complex workloads, fully simulating actual application scenarios. Through detailed performance analysis reports, the present invention provides strong support for the optimization and improvement of cloud storage systems, and helps to improve the quality of cloud storage services and user experience. BRIEF DESCRIPTION OF THE DRAWINGS

[0124] The present invention is further described below in conjunction with the accompanying drawings.

[0125] Attached Figure 1 It is a structural diagram of the cloud storage system performance testing system based on the large model. DETAILED DESCRIPTION

[0126] The cloud storage system performance testing method and system based on a large model of the present invention are described in detail below with reference to the accompanying drawings and specific embodiments.

[0127] Embodiment 1:

[0128] This embodiment provides a cloud storage system performance testing method based on a large model, the method is as follows:

[0129] S1. Generate data: Use a large model pre-training model (such as the Hairuo large model) to generate diverse test data;

[0130] S2. Generate workload: define the operation mode and generate corresponding and complex workloads to simulate the operation mode in actual application scenarios;

[0131] S3, test execution: execute the generated test data and workload on the cloud storage system and record performance metrics;

[0132] S4. Performance analysis: Analyze the recorded performance indicators and generate a detailed performance report.

[0133] The generated data in step S1 of this embodiment are specifically as follows:

[0134] S101. Define data types: Based on the actual business scenario, define the data types of text, image, video and audio that need to be generated; the details are as follows:

[0135] S10101. Identify and classify data requirements in actual application scenarios;

[0136] S10102. List all possible data types and create detailed description documents for each type;

[0137] S10103. Determine the characteristics of each data type: text data requires different lengths and languages, and image data requires different resolutions and color depths;

[0138] S102. Analyze data characteristics: Use a large-scale pre-trained model to analyze the data distribution, size, and format characteristics of data in actual application scenarios; the details are as follows:

[0139] S10201. Analyze the existing data set through the pre-trained model to extract feature distribution, data format and size information;

[0140] S10202. Use statistical analysis methods to quantify and classify data characteristics;

[0141] S10203. Write an analysis report that describes in detail the characteristics of each data type;

[0142] S103. Formulate data generation strategy: Based on the data characteristics analysis results, formulate a data generation strategy to ensure that the generated data is diverse and representative; the details are as follows:

[0143] S10301. Develop diversified data generation plans based on analysis results to ensure extensive data coverage;

[0144] S10302. Formulate rules and parameters for data generation to ensure that the generated data conforms to the expected distribution and characteristics;

[0145] S10303. Use a pre-trained model to configure a generation strategy, where the configuration generation strategy includes parameter adjustment and data enhancement technology;

[0146] S104, generate data: call the pre-trained model to generate diversified test data, and store the data in the cloud storage system; the details are as follows:

[0147] S10401. Generate data according to the strategy using the pre-trained model;

[0148] S10402. Verify the quality of generated data to ensure that the generated data meets the defined characteristics;

[0149] S10403. Upload the generated data to a cloud storage system using a cloud storage system (such as AWS S3, Google Cloud Storage, or Azure BlobStorage);

[0150] S10404. Set up appropriate access control and backup strategies to ensure the security and high availability of data in the cloud storage system.

[0151] The generation of workload in step S2 of this embodiment is specifically as follows:

[0152] S201, define operation mode: define multiple operation modes such as read, write, update and delete as well as combined operation and high concurrency scenario according to actual application scenarios; specifically: first, through in-depth analysis of actual applications, define detailed descriptions of basic operations and combined operations; then determine the operation mode in high concurrency scenario and write detailed operation procedures;

[0153] S202, analyzing workload characteristics: using a large-scale pre-trained model to analyze workload characteristics in actual applications; wherein workload characteristics include operation frequency, operation sequence, and operation dependency; specifically as follows:

[0154] S20201. Collect and analyze the operation logs of existing systems to extract workload characteristics.

[0155] S20202. Use pre-trained models to perform feature analysis and identify frequency, sequence, and dependencies.

[0156] S20203. Write an analysis report describing the detailed characteristics of the workload;

[0157] S203, workload generation strategy: Based on the workload characteristic analysis results, a workload generation strategy is formulated to ensure that the generated workload is complex and representative; the details are as follows:

[0158] S20301. Develop workload generation strategies covering multiple operation modes;

[0159] S20302. Determine the parameters and rules for generating workloads to ensure their complexity and representativeness;

[0160] S20303. Adjust strategies to suit different testing needs and scenarios;

[0161] S204, generating workload: calling the pre-trained model to generate a complex workload, and applying the workload to the cloud storage system; the details are as follows:

[0162] S20401. Generate workload scripts and tasks using the pre-trained model; then verify the generated workload to ensure that it meets the expected characteristics;

[0163] S20402, deploying workload tasks to the cloud storage system and preparing for execution;

[0164] S20403. Configure the relevant parameters of the cloud storage system (such as AWS S3, Google Cloud Storage, or Azure BlobStorage) to ensure that the workload can be executed correctly and record operation logs.

[0165] The test execution in step S3 of this embodiment is specifically as follows:

[0166] S301, test environment preparation: configure the test environment of the cloud storage system, which includes hardware resources, software configuration and network settings; the details are as follows:

[0167] S30101. Configure corresponding cloud storage resources (such as computing nodes and storage devices) according to test requirements.

[0168] S30102. Install and configure necessary software, including operating system, database and other dependent components;

[0169] S30103. Set up the network environment of the cloud storage system to ensure the stability and scalability of the test environment;

[0170] S30104. Use the cloud service provider's management console or CLI tool to configure and manage resources;

[0171] S302, test task scheduling: Schedule test tasks according to the predefined test plan to ensure that the test tasks are executed in the set order and time; the details are as follows:

[0172] S30201. Write a test plan that describes in detail the sequence and timing of test tasks;

[0173] S30202. Use automation tools (such as AWS Batch, Google Cloud Scheduler, or Azure LogicApps) to schedule test tasks and ensure that tasks are executed as planned.

[0174] S30203. Monitor task progress in real time and adjust task scheduling in a timely manner;

[0175] S303, test data loading: Load the generated test data into the cloud storage system and ensure the integrity and consistency of the data; the details are as follows:

[0176] S30301. Migrate the generated data from local or other storage systems to cloud storage;

[0177] S30302. Verify the integrity and consistency of data loading to ensure that the data is correct;

[0178] S30303. Use scripts or automation tools (such as AWS Data Sync, Google Transfer Appliance, or Azure Data Box) to automate the data loading process and improve efficiency.

[0179] S304, workload execution: Execute the generated workload on the cloud storage system and record the operation logs and performance indicators in real time; the details are as follows:

[0180] S30401, start the workload task and monitor its execution;

[0181] S30402. Use cloud monitoring tools such as AWS CloudWatch, Google Stackdriver, or AzureMonitor to log detailed information about each operation.

[0182] S30403. Collect performance indicators, including response time, throughput, and resource utilization;

[0183] S305, exception handling: During the test process, if any abnormal situation of the monitoring system status is found, it will be handled in time to ensure the stability and continuity of the test process; the details are as follows:

[0184] S30501. Monitor system status in real time and use the alarm system to detect abnormal situations;

[0185] S30502. Develop detailed exception handling procedures to ensure rapid response and repair;

[0186] S30503. Record abnormal situations and handling procedures for subsequent analysis and improvement.

[0187] The performance analysis in step S4 of this embodiment is as follows:

[0188] S401, data collection: obtain operation logs and performance indicators from the test execution process, the operation logs and performance indicators include response time, throughput and resource utilization; the details are as follows:

[0189] S40101. Collect operational logs and performance metrics using data collection tools such as AWS CloudWatch Logs, Google Cloud Logging, or Azure Log Analytics.

[0190] S40102. Ensure the integrity and accuracy of data and prevent data loss or tampering;

[0191] S40103. Collect data regularly for continuous monitoring and analysis;

[0192] S402, data processing: pre-processing the collected data; the details are as follows:

[0193] S40201. Clean data and remove invalid or erroneous data;

[0194] S40202. Convert data formats to suit the requirements of analysis tools;

[0195] S40203, aggregate data and generate necessary statistical indicators;

[0196] S403, performance analysis: Based on the preprocessed data, performance analysis is performed to evaluate the performance of the cloud storage system under different workloads and identify performance bottlenecks and optimization points; the details are as follows:

[0197] S40301. Use analytical tools and models to evaluate system performance (such as AWS QuickSight, Google DataStudio, or Azure Power BI).

[0198] S40302. Identify performance bottlenecks, which include long response times and low throughput.

[0199] S40303. Propose optimization suggestions to help improve the system;

[0200] S404, report generation: based on the analysis results, write a performance report, generate a detailed description of the test background, and then generate a detailed performance report; wherein the detailed performance report includes the test background, test methods, test results, performance evaluation and optimization suggestions.

[0201] Embodiment 2:

[0202] As attached Figure 1 As shown, this embodiment provides a cloud storage system performance testing system based on a large model, the system comprising:

[0203] The data generation module is used to generate diversified test data using a large model pre-training model (such as the Hairuo large model);

[0204] The workload generation module is used to define the operation mode and generate corresponding and complex workloads to simulate the operation mode in the actual application scenario;

[0205] A test execution module, used to execute the generated test data and workload on the cloud storage system and record performance indicators;

[0206] The performance analysis module is used to analyze the recorded performance indicators and generate detailed performance reports.

[0207] The data generation module in this embodiment includes:

[0208] The data type definition submodule is used to define the text, image, video and audio data types that need to be generated according to the actual business scenarios; the working process of the data type definition submodule is as follows:

[0209] ① Identify and classify data requirements in actual application scenarios;

[0210] ② List all possible data types and create detailed description documents for each type;

[0211] ③ Determine the characteristics of each data type: text data requires different lengths and languages, image data requires different resolutions and color depths;

[0212] The data characteristic analysis submodule is used to analyze the data characteristics of data distribution, size and format in actual application scenarios using a large-scale pre-trained model; the working process of the data characteristic analysis submodule is as follows:

[0213] ① Analyze existing data sets through pre-trained models to extract feature distribution, data format and size information;

[0214] ② Use statistical analysis methods to quantify and classify data characteristics;

[0215] ③Write an analysis report, describing in detail the characteristics of each data type;

[0216] The data generation strategy formulation submodule is used to formulate data generation strategies based on the data characteristic analysis results to ensure that the generated data is diverse and representative. The working process of the data generation strategy formulation submodule is as follows:

[0217] ① Develop diversified data generation plans based on the analysis results to ensure extensive data coverage;

[0218] ② Formulate rules and parameters for data generation to ensure that the generated data conforms to the expected distribution and characteristics;

[0219] ③ Use pre-trained models to configure the generation strategy, which includes parameter adjustment and data enhancement techniques;

[0220] The data generation submodule is used to call the pre-trained model to generate diversified test data and store the data in the cloud storage system. The working process of the data generation submodule is as follows:

[0221] ①Use pre-trained models to generate data according to the strategy;

[0222] ②Verify the quality of the generated data to ensure that the generated data meets the defined characteristics;

[0223] ③Use cloud storage systems (such as AWS S3, Google Cloud Storage, or Azure Blob Storage) to upload the generated data to a cloud storage system;

[0224] ④ Set up appropriate access control and backup strategies to ensure the security and high availability of data in the cloud storage system.

[0225] The workload generation module in this embodiment includes:

[0226] The operation mode definition submodule is used to define multiple operation modes of reading, writing, updating and deleting as well as combined operations and high concurrency scenarios according to actual application scenarios; the working process of the operation mode definition submodule is as follows:

[0227] ① Through in-depth analysis of actual applications, define detailed descriptions of basic operations and combined operations;

[0228] ②Determine the operation mode in high-concurrency scenarios and write detailed operation procedures;

[0229] The workload characteristics analysis submodule is used to analyze the workload characteristics in actual applications using a large-scale pre-trained model; the workload characteristics include operation frequency, operation sequence, and operation dependency; the specific working process of the workload characteristics analysis submodule is as follows:

[0230] ① Collect and analyze the operation logs of existing systems to extract workload characteristics;

[0231] ② Use pre-trained models for feature analysis to identify frequency, sequence, and dependencies;

[0232] ③ Write an analysis report describing the detailed characteristics of the workload;

[0233] The workload generation strategy submodule is used to formulate a workload generation strategy based on the workload characteristic analysis results to ensure that the generated workload is complex and representative. The working process of the workload generation strategy submodule is as follows:

[0234] ① Develop workload generation strategies that cover multiple operating modes;

[0235] ② Determine the parameters and rules for generating workloads to ensure their complexity and representativeness;

[0236] ③Adjust strategies to suit different testing needs and scenarios;

[0237] The workload generation submodule is used to call the pre-trained model to generate complex workloads and apply the workloads to the cloud storage system. The working process of the workload generation submodule is as follows:

[0238] ① Generate workload scripts and tasks using pre-trained models;

[0239] ② Verify the generated workload to ensure it meets the expected characteristics;

[0240] ③Deploy the workload tasks to the cloud storage system and prepare for execution;

[0241] ④Configure relevant parameters of the cloud storage system (such as AWS S3, Google Cloud Storage, or Azure Blob Storage) to ensure that the workload can be executed correctly and record operation logs.

[0242] The test execution module in this embodiment includes:

[0243] The test environment preparation submodule is used to configure the test environment of the cloud storage system. The test environment of the cloud storage system includes hardware resources, software configuration and network settings. The working process of the test environment preparation submodule is as follows:

[0244] ①Configure corresponding cloud storage resources (such as computing nodes and storage devices) according to test requirements;

[0245] ②Install and configure necessary software, including operating system, database and other dependent components;

[0246] ③ Set up the network environment of the cloud storage system to ensure the stability and scalability of the test environment;

[0247] ④ Use the cloud service provider’s management console or CLI tools to configure and manage resources;

[0248] The test task scheduling submodule is used to schedule test tasks according to the predefined test plan to ensure that the test tasks are executed in the set order and time; the working process of the test task scheduling submodule is as follows:

[0249] ① Write a test plan, describing in detail the sequence and timing of test tasks;

[0250] ② Use automation tools (such as AWS Batch, Google Cloud Scheduler, or Azure LogicApps) to schedule test tasks and ensure that tasks are executed as planned;

[0251] ③ Monitor task progress in real time and adjust task scheduling in a timely manner;

[0252] The test data loading submodule is used to load the generated test data into the cloud storage system and ensure the integrity and consistency of the data. The working process of the test data loading submodule is as follows:

[0253] ① Migrate the generated data from local or other storage systems to cloud storage;

[0254] ②Verify the integrity and consistency of data loading to ensure that the data is correct;

[0255] ③ Use scripts or automation tools (such as AWS Data Sync, Google Transfer Appliance, or Azure Data Box) to automate the data loading process and improve efficiency;

[0256] The workload execution submodule is used to execute the generated workload on the cloud storage system and record the operation logs and performance indicators in real time. The working process of the workload execution submodule is as follows:

[0257] ① Start workload tasks and monitor their execution;

[0258] ② Use cloud monitoring tools (such as AWS CloudWatch, Google Stackdriver, or Azure Monitor) to record detailed information for each operation;

[0259] ③ Collect performance indicators, including response time, throughput and resource utilization;

[0260] The exception handling submodule is used to handle abnormal situations in the monitoring system status in a timely manner during the test process to ensure the stability and continuity of the test process; the working process of the exception handling submodule is as follows:

[0261] ① Monitor system status in real time and use the alarm system to detect abnormal situations;

[0262] ② Develop a detailed exception handling process to ensure rapid response and repair;

[0263] ③Record abnormal situations and handling processes for subsequent analysis and improvement.

[0264] The performance analysis module in this embodiment includes:

[0265] The data acquisition submodule is used to obtain operation logs and performance indicators from the test execution process. The operation logs and performance indicators include response time, throughput and resource utilization. The working process of the data acquisition submodule is as follows:

[0266] ① Use data collection tools (such as AWS CloudWatch Logs, Google Cloud Logging, or AzureLog Analytics) to collect operation logs and performance metrics;

[0267] ② Ensure the integrity and accuracy of data and prevent data loss or tampering;

[0268] ③ Collect data regularly for continuous monitoring and analysis;

[0269] The data processing submodule is used to pre-process the collected data; the working process of the data processing submodule is as follows:

[0270] ① Clean the data and remove invalid or erroneous data;

[0271] ②Convert the data format to adapt it to the requirements of the analysis tool;

[0272] ③ Aggregate data and generate necessary statistical indicators;

[0273] The performance analysis submodule is used to perform performance analysis based on preprocessed data, evaluate the performance of the cloud storage system under different workloads, and identify performance bottlenecks and optimization points. The working process of the performance analysis submodule is as follows:

[0274] ① Use analytical tools and models to evaluate system performance (such as AWS QuickSight, Google Data Studio, or Azure Power BI);

[0275] ②Identify performance bottlenecks, including long response time and low throughput;

[0276] ③Propose optimization suggestions to help improve the system;

[0277] The report generation submodule is used to write a performance report based on the analysis results, generate a detailed description of the test background, and then generate a detailed performance report; wherein the detailed performance report includes the test background, test methods, test results, performance evaluation and optimization suggestions.

[0278] Embodiment 3:

[0279] This embodiment also provides an electronic device, including: a memory and at least one processor;

[0280] Wherein, the memory stores computer-executable instructions;

[0281] The at least one processor executes the computer-executable instructions stored in the memory, so that the at least one processor executes the cloud storage system performance testing method based on a large model as described in any one of the present inventions.

[0282] The processor may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The processor may be a microprocessor or any conventional processor, etc.

[0283] The memory can be used to store computer programs and / or modules. The processor realizes various functions of the electronic device by running or executing the computer programs and / or modules stored in the memory, and calling the data stored in the memory. The memory can mainly include a program storage area and a data storage area, wherein the program storage area can store an operating system, at least one application required for a function, etc.; the data storage area can store data created according to the use of the terminal, etc. In addition, the memory can also include a high-speed random access memory, and can also include a non-volatile memory, such as a hard disk, a memory, a plug-in hard disk, a smart memory card (SMC), a secure digital (SD) card, a flash memory card, at least one disk storage period, a flash memory device, or other volatile solid-state storage devices.

[0284] Embodiment 4:

[0285] This embodiment also provides a computer-readable storage medium, which stores a plurality of instructions, which are loaded by a processor, so that the processor executes the cloud storage system performance test method based on a large model in any embodiment of the present invention. Specifically, a system or device equipped with a storage medium can be provided, on which a software program code that implements the functions of any of the above embodiments is stored, and a computer (or CPU or MPU) of the system or device reads and executes the program code stored in the storage medium.

[0286] In this case, the program code itself read from the storage medium can realize the function of any one of the above-mentioned embodiments, and thus the program code and the storage medium storing the program code constitute a part of the present invention.

[0287] The storage medium embodiments for providing the program code include a floppy disk, a hard disk, a magneto-optical disk, an optical disk (such as CD-ROM, CD-R, CD-RW, DVD-ROM, DVD-RAM, DVD-RW, DVD+RW), a magnetic tape, a non-volatile memory card, and a ROM. Alternatively, the program code can be downloaded from a server computer by a communication network.

[0288] In addition, it should be clear that the functions of any of the above embodiments can be implemented not only by executing the program code read by the computer, but also by enabling an operating system operating on the computer to complete part or all of the actual operations based on instructions from the program code.

[0289] In addition, it can be understood that the program code read from the storage medium is written to a memory provided in an expansion board inserted into the computer or written to a memory provided in an expansion unit connected to the computer, and then based on the instructions of the program code, a CPU installed on the expansion board or the expansion unit is enabled to perform part or all of the actual operations, thereby realizing the functions of any of the above-mentioned embodiments.

[0290] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A cloud storage system performance testing method based on a large model, characterized in that: The method is as follows: Generate data: Use large model pre-training models to generate diverse test data; Generate workload: define the operation mode and generate corresponding workload to simulate the operation mode in the actual application scenario; Test execution: Execute the generated test data and workload on the cloud storage system and record performance metrics; Performance analysis: Analyze the recorded performance indicators and generate detailed performance reports.

2. The cloud storage system performance testing method based on a large model according to claim 1 is characterized in that: The generated data is as follows: Define data types: Based on actual business scenarios, define the text, image, video, and audio data types that need to be generated; specifically: first identify and classify the data requirements in the actual application scenario; then list all possible data types and create a detailed description document for each type; finally determine the characteristics of each data type: text data requires different lengths and languages, and image data requires different resolutions and color depths; Analyze data characteristics: Use large-scale pre-trained models to analyze the data characteristics of data distribution, size, and format in actual application scenarios; specifically: first analyze the existing data set through the pre-trained model to extract feature distribution, data format, and size information; then use statistical analysis methods to quantify and classify data characteristics; finally, write an analysis report that describes the characteristics of each data type in detail; Formulate data generation strategy: Based on the results of data characteristic analysis, formulate data generation strategy to ensure that the generated data is diverse and representative; Specifically, we first formulate diversified data generation plans based on the analysis results to ensure wide coverage of data; then formulate data generation rules and parameters to ensure that the generated data meets the expected distribution and characteristics; finally, we use the pre-trained model to configure the generation strategy, which includes parameter adjustment and data enhancement technology; Generate data: call the pre-trained model to generate diversified test data and store the data in the cloud storage system; specifically: first use the pre-trained model to generate data according to the strategy; then verify the quality of the generated data to ensure that the generated data meets the defined characteristics; then use the cloud storage system to upload the generated data to the cloud storage system; finally, set appropriate access control and backup policies to ensure the security and high availability of data in the cloud storage system.

3. The cloud storage system performance testing method based on a large model according to claim 1 is characterized in that: The generated workload is as follows: Define operation modes: Define multiple operation modes such as read, write, update, and delete, as well as combined operations and high-concurrency scenarios according to actual application scenarios; specifically: first, through in-depth analysis of actual applications, define detailed descriptions of basic operations and combined operations; then determine the operation mode in high-concurrency scenarios and write detailed operation procedures; Analyze workload characteristics: Use large-scale pre-trained models to analyze workload characteristics in actual applications. Workload characteristics include operation frequency, operation sequence, and operation dependency. Specifically, first collect and analyze the operation logs of existing systems to extract workload characteristics. Then use pre-trained models to perform feature analysis to identify frequency, sequence, and dependency. Finally, write an analysis report to describe the detailed characteristics of the workload. Workload generation strategy: Based on the results of workload characteristic analysis, a workload generation strategy is formulated to ensure that the generated workload is complex and representative. Specifically, a workload generation strategy covering multiple operation modes is first formulated; then the parameters and rules for generating workloads are determined to ensure their complexity and representativeness; finally, the strategy is adjusted to suit different test requirements and scenarios. Generate workload: call the pre-trained model to generate complex workloads, and apply the workloads to the cloud storage system; specifically: first use the pre-trained model to generate workload scripts and tasks; then verify the generated workload to ensure that it meets the expected characteristics; then deploy the workload tasks to the cloud storage system and prepare for execution; finally, configure the relevant parameters of the cloud storage system to ensure that the workload can be executed correctly and record operation logs.

4. The cloud storage system performance testing method based on a large model according to claim 1 is characterized in that: The test execution is as follows: Test environment preparation: Configure the test environment of the cloud storage system, which includes hardware resources, software configuration and network settings. Specifically, first configure the corresponding cloud storage resources according to the test requirements; then install and configure the necessary software, including the operating system, database and other dependent components; then set up the network environment of the cloud storage system to ensure the stability and scalability of the test environment; finally, use the cloud service provider's management console or CLI tool to configure and manage resources; Test task scheduling: Schedule test tasks according to the predefined test plan to ensure that the test tasks are executed in the set order and time; specifically: first write a test plan to describe the order and time arrangement of the test tasks in detail; then use automated tools to schedule test tasks to ensure that the tasks are executed as planned; finally, monitor the task progress in real time and adjust the task scheduling in a timely manner; Test data loading: Load the generated test data into the cloud storage system and ensure the integrity and consistency of the data; Specifically, first migrate the generated data from local or other storage systems to cloud storage; then verify the integrity and consistency of the data loading to ensure that the data is correct; finally, use scripts or automation tools to automate the data loading process to improve efficiency; Workload execution: Execute the generated workload on the cloud storage system and record the operation logs and performance indicators in real time. Specifically, start the workload task first and monitor its execution. Then use the cloud monitoring tool to record the details of each operation. Finally, collect performance indicators, including response time, throughput and resource utilization. Exception handling: During the test process, any abnormalities in the monitoring system status are promptly handled to ensure the stability and continuity of the test process; specifically: first monitor the system status in real time and use the alarm system to detect abnormalities; then develop a detailed exception handling process to ensure rapid response and repair; finally, record the abnormal situation and handling process for subsequent analysis and improvement.

5. The cloud storage system performance testing method based on a large model according to any one of claims 1 to 4, characterized in that: The performance analysis is as follows: Data collection: Obtain operation logs and performance indicators from the test execution process, including response time, throughput, and resource utilization. Specifically, first use data collection tools to collect operation logs and performance indicators; then ensure the integrity and accuracy of the data to prevent data loss or tampering; finally, collect data regularly for continuous monitoring and analysis; Data processing: pre-process the collected data; Specifically, first clean the data to remove invalid or erroneous data; then convert the data format to adapt the data format to the requirements of the analysis tool; finally aggregate the data to generate necessary statistical indicators; Performance analysis: Based on the preprocessed data, we conduct performance analysis to evaluate the performance of the cloud storage system under different workloads and identify performance bottlenecks and optimization points. Specifically, we first use analytical tools and models to evaluate system performance; then identify performance bottlenecks, such as long response time and low throughput; and finally make optimization suggestions to help improve the system. Report generation: Based on the analysis results, a performance report is written to describe the test background in detail, and then a detailed performance report is generated; the detailed performance report includes the test background, test methods, test results, performance evaluation and optimization suggestions.

6. A cloud storage system performance testing system based on a large model, characterized in that: The system includes: Data generation module, used to generate diverse test data using the large model pre-training model; A workload generation module is used to define the operation mode and generate corresponding workloads to simulate the operation mode in the actual application scenario; A test execution module, used to execute the generated test data and workload on the cloud storage system and record performance indicators; The performance analysis module is used to analyze the recorded performance indicators and generate detailed performance reports.

7. The cloud storage system performance testing system based on a large model according to claim 6, characterized in that: The data generation module comprises: The data type definition submodule is used to define the text, image, video and audio data types that need to be generated according to the actual business scenarios; the working process of the data type definition submodule is as follows: ① Identify and classify data requirements in actual application scenarios; ② List all possible data types and create detailed description documents for each type; ③ Determine the characteristics of each data type: text data requires different lengths and languages, image data requires different resolutions and color depths; The data characteristic analysis submodule is used to analyze the data characteristics of data distribution, size and format in actual application scenarios using a large-scale pre-trained model; the working process of the data characteristic analysis submodule is as follows: ① Analyze existing data sets through pre-trained models to extract feature distribution, data format and size information; ② Use statistical analysis methods to quantify and classify data characteristics; ③Write an analysis report, describing in detail the characteristics of each data type; The data generation strategy formulation submodule is used to formulate data generation strategies based on the data characteristic analysis results to ensure that the generated data is diverse and representative. The working process of the data generation strategy formulation submodule is as follows: ① Develop diversified data generation plans based on the analysis results to ensure extensive data coverage; ② Formulate rules and parameters for data generation to ensure that the generated data conforms to the expected distribution and characteristics; ③ Use pre-trained models to configure the generation strategy, which includes parameter adjustment and data enhancement techniques; The data generation submodule is used to call the pre-trained model to generate diversified test data and store the data in the cloud storage system. The working process of the data generation submodule is as follows: ①Use pre-trained models to generate data according to the strategy; ②Verify the quality of the generated data to ensure that the generated data meets the defined characteristics; ③Use the cloud storage system to upload the generated data to the cloud storage system; ④ Set up appropriate access control and backup strategies to ensure the security and high availability of data in the cloud storage system; The workload generation module comprises: The operation mode definition submodule is used to define multiple operation modes of reading, writing, updating and deleting as well as combined operations and high concurrency scenarios according to actual application scenarios; the working process of the operation mode definition submodule is as follows: ① Through in-depth analysis of actual applications, define detailed descriptions of basic operations and combined operations; ②Determine the operation mode in high-concurrency scenarios and write detailed operation procedures; The workload characteristics analysis submodule is used to analyze the workload characteristics in actual applications using a large-scale pre-trained model; the workload characteristics include operation frequency, operation sequence, and operation dependency; the specific working process of the workload characteristics analysis submodule is as follows: ① Collect and analyze the operation logs of existing systems to extract workload characteristics; ② Use pre-trained models for feature analysis to identify frequency, sequence, and dependencies; ③ Write an analysis report describing the detailed characteristics of the workload; The workload generation strategy submodule is used to formulate a workload generation strategy based on the workload characteristic analysis results to ensure that the generated workload is complex and representative. The working process of the workload generation strategy submodule is as follows: ① Develop workload generation strategies that cover multiple operating modes; ② Determine the parameters and rules for generating workloads to ensure their complexity and representativeness; ③Adjust strategies to suit different testing needs and scenarios; The workload generation submodule is used to call the pre-trained model to generate complex workloads and apply the workloads to the cloud storage system. The working process of the workload generation submodule is as follows: ① Generate workload scripts and tasks using pre-trained models; ② Verify the generated workload to ensure it meets the expected characteristics; ③Deploy the workload tasks to the cloud storage system and prepare for execution; ④Configure relevant parameters of the cloud storage system to ensure that the workload can be executed correctly and record operation logs.

8. The cloud storage system performance testing system based on a large model according to claim 6 or 7, characterized in that: The test execution module includes: The test environment preparation submodule is used to configure the test environment of the cloud storage system. The test environment of the cloud storage system includes hardware resources, software configuration and network settings. The working process of the test environment preparation submodule is as follows: ①Configure corresponding cloud storage resources according to test requirements; ②Install and configure necessary software, including operating system, database and other dependent components; ③ Set up the network environment of the cloud storage system to ensure the stability and scalability of the test environment; ④ Use the cloud service provider’s management console or CLI tools to configure and manage resources; The test task scheduling submodule is used to schedule test tasks according to the predefined test plan to ensure that the test tasks are executed in the set order and time; the working process of the test task scheduling submodule is as follows: ① Write a test plan, describing in detail the sequence and timing of test tasks; ② Use automated tools to schedule test tasks to ensure that tasks are executed as planned; ③ Monitor task progress in real time and adjust task scheduling in a timely manner; The test data loading submodule is used to load the generated test data into the cloud storage system and ensure the integrity and consistency of the data. The working process of the test data loading submodule is as follows: ① Migrate the generated data from local or other storage systems to cloud storage; ②Verify the integrity and consistency of data loading to ensure that the data is correct; ③ Use scripts or automation tools to automate the data loading process and improve efficiency; The workload execution submodule is used to execute the generated workload on the cloud storage system and record the operation logs and performance indicators in real time. The working process of the workload execution submodule is as follows: ① Start workload tasks and monitor their execution; ②Use cloud monitoring tools to record detailed information of each operation; ③ Collect performance indicators, including response time, throughput and resource utilization; The exception handling submodule is used to handle abnormal situations in the monitoring system status in a timely manner during the test process to ensure the stability and continuity of the test process; the working process of the exception handling submodule is as follows: ① Monitor system status in real time and use the alarm system to detect abnormal situations; ② Develop a detailed exception handling process to ensure rapid response and repair; ③Record abnormal situations and handling processes for subsequent analysis and improvement; The performance analysis module includes: The data acquisition submodule is used to obtain operation logs and performance indicators from the test execution process. The operation logs and performance indicators include response time, throughput and resource utilization. The working process of the data acquisition submodule is as follows: ①Use data collection tools to collect operation logs and performance indicators; ② Ensure the integrity and accuracy of data and prevent data loss or tampering; ③ Collect data regularly for continuous monitoring and analysis; The data processing submodule is used to pre-process the collected data; the working process of the data processing submodule is as follows: ① Clean the data and remove invalid or erroneous data; ②Convert the data format to adapt it to the requirements of the analysis tool; ③ Aggregate data and generate necessary statistical indicators; The performance analysis submodule is used to perform performance analysis based on preprocessed data, evaluate the performance of the cloud storage system under different workloads, and identify performance bottlenecks and optimization points. The working process of the performance analysis submodule is as follows: ① Use analytical tools and models to evaluate system performance; ②Identify performance bottlenecks, including long response time and low throughput; ③Propose optimization suggestions to help improve the system; The report generation submodule is used to write a performance report based on the analysis results, generate a detailed description of the test background, and then generate a detailed performance report; wherein the detailed performance report includes the test background, test methods, test results, performance evaluation and optimization suggestions.

9. An electronic device, characterized in that: include: memory and at least one processor; Wherein, the memory stores computer-executable instructions; The at least one processor executes the computer-executable instructions stored in the memory, so that the at least one processor executes the cloud storage system performance testing method based on a large model as described in any one of claims 1 to 5.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer-executable instructions, and when the processor executes the computer-executable instructions, the cloud storage system performance testing method based on a large model as described in any one of claims 1 to 5 is implemented.

Citation Information

Cited By

  • Load generation method, computer program product and network attached storage system

    CN120029870A

  • Load generation method, computer program product, and network attached storage system

    CN120029870B