CN-121979751-A - System performance simulation verification method based on information data analysis

CN121979751ACN 121979751 ACN121979751 ACN 121979751ACN-121979751-A

Abstract

The invention discloses a system performance simulation verification method based on information data analysis, and relates to the technical field of system performance test and simulation. The system performance simulation verification method based on the information data analysis comprises six steps of information data acquisition, preprocessing, simulation model construction, model calibration, performance simulation verification and result analysis, wherein a high-precision simulation model is constructed by combining real-time acquisition and historical data, performance verification is carried out under various load scenes, and finally a verification report containing performance bottleneck diagnosis and optimization suggestions is generated. The system performance simulation verification method based on the information data analysis can effectively improve the accuracy and the comprehensiveness of system performance simulation, reduce the actual test cost and provide scientific basis for system optimization.

Inventors

ZHANG YIJUN
JING FENG
ZHENG KOUQUAN
XIN DAN
SHI YOUWEI
ZHAO LE
HUANG XIAOTING

Assignees

中国人民解放军国防科技大学

Dates

Publication Date: 20260505
Application Date: 20251224

Claims (9)

1. The system performance simulation verification method based on information data analysis is characterized by comprising the following steps of: S1, information data acquisition, namely acquiring hardware resource data, business load data and system configuration data of a system to be verified, wherein the acquisition mode is combined with real-time acquisition and history data import; S2, preprocessing information data, namely sequentially performing data cleaning, data noise reduction, data normalization and key feature extraction on the data acquired in the step S1 to obtain key feature data; S3, constructing a simulation model, namely constructing a system performance simulation model comprising a hardware resource sub-model, a business load sub-model and a system interaction sub-model by adopting a simulation tool based on key characteristic data; s4, calibrating a simulation model, namely comparing actual operation data of the system to be verified with simulation indexes of the simulation model, and adjusting model parameters until the deviation rate of the simulation indexes and the actual indexes is less than or equal to 5%; s5, performance simulation verification, namely setting conventional, sudden load and service switching scenes based on the calibrated model, running simulation and collecting performance indexes, and taking the average value of multiple simulation results as a final simulation result; And S6, analyzing the simulation result, namely comparing the final simulation result with the system performance requirement index, diagnosing the performance bottleneck and generating a performance simulation verification report.
2. The method is characterized in that in the step S1, the hardware resource data comprises CPU utilization rate, memory occupancy rate, disk I/O rate and network bandwidth utilization rate, the service load data comprises service request quantity, request response time and request failure rate, the system configuration data comprises process quantity, thread scheduling strategy and cache size, real-time acquisition is realized through SNMP or Prometheus interfaces, sampling frequency is 1-10Hz, history data is extracted from an elastic search log database, the real-time acquired data is primarily filtered through edge computing nodes, only effective data is transmitted to a data processing center, data transmission quantity is reduced, and filtering rules are that data exceeding a reasonable threshold range of hardware resources, such as CPU utilization rate is more than 100%, and memory occupancy rate is less than 0%.
3. The system performance simulation verification method based on information data analysis according to claim 1, wherein in the step S2, abnormal values are removed by using a3 sigma principle or a box graph method in data cleaning, missing values are filled by using linear interpolation or average values, wavelet transformation or a moving average method is adopted in data noise reduction, a Min-Max algorithm is adopted in data normalization, and a principal component analysis or a random forest algorithm is adopted in key feature extraction.
4. The method for simulating and verifying the system performance based on the information data analysis according to claim 1, wherein in the step S3, a simulation tool comprises NS-3, MATLAB/Simulink or OPNET, a hardware resource sub-model simulates CPU scheduling, memory allocation, disk I/O and network transmission, a service load sub-model supports self-defined load intensity, and a system interaction sub-model defines the interaction rule of hardware and service.
5. The method according to claim 1, wherein in step S4, the actual running data is collected by performing performance test of the same configuration on the system to be verified, and the adjusted model parameters include CPU scheduling delay, memory access time consumption and request generation interval.
6. The method for verifying system performance simulation based on information data analysis according to claim 1, wherein in the step S5, the service request amount of the sudden load scene is increased by 10-20 times, the performance index includes average response time, maximum response time, throughput per unit time, average utilization rate of CPU/memory/disk/network and request success rate, and the simulation execution times are 3-5 times.
7. The method according to claim 1, wherein in step S6, the performance simulation verification report includes simulation scene setting, performance index details, demand contrast analysis, bottleneck diagnosis results, and system optimization suggestions.
8. The system performance simulation verification method based on information data analysis according to claim 1, wherein in step S1, data collected in real time is preliminarily filtered by an edge computing node, only effective data is transmitted to a data processing center, and data transmission quantity is reduced.
9. The system performance simulation verification method based on information data analysis according to claim 1, wherein the simulation model constructed in the step S3 supports dynamic adjustment, and sub-model parameters can be updated based on real-time supplementary data in the simulation process to adapt to dynamic changes of business scenes.

Description

System performance simulation verification method based on information data analysis Technical Field The invention relates to the technical field of system performance test and simulation, in particular to a system performance simulation verification method based on information data analysis. Background With the rapid development of information technology, the scale and complexity of various information systems are continuously improved, and the requirements on the system performance are also increasingly improved. The system performance verification is used as a key link for ensuring the stable operation of the system, the traditional method is mostly dependent on performance test in an actual environment, and the problems of high cost, long period, limited scene coverage and the like exist. In the prior art, the system performance simulation method often has the defects of incomplete data acquisition, insufficient precision of a simulation model, difficulty in accurately reflecting the actual running state of the system, single simulation scene, incapability of coping with complex and changeable business loads, and imperfect model calibration mechanism, so that the simulation result and the actual deviation are larger. The prior art has clear defect cases that, for example, patent CN109885678A only discloses a scheme of combining and collecting real-time data and historical data, but does not adopt edge calculation to perform preliminary filtration on the real-time data, so that the data transmission quantity is overlarge and the processing efficiency is low, and patent CN110245789B realizes multi-scene simulation, but the deviation rate after model calibration is controlled within 10 percent, and cannot meet the performance verification requirement of high-precision systems (such as financial transaction systems and industrial control systems). Therefore, a method capable of comprehensively utilizing multi-source data, constructing a high-precision simulation model and performing comprehensive performance verification under various scenes is needed, so that accuracy and reliability of system performance simulation are improved, and scientific basis is provided for system optimization. Disclosure of Invention The invention aims to provide a system performance simulation verification method based on information data analysis, which is used for solving the problems of incomplete data acquisition, insufficient model precision, single scene and imperfect calibration mechanism in the prior art. In order to achieve the above purpose, the invention provides a system performance simulation verification method based on information data analysis, which comprises the following steps: S1, information data acquisition, namely acquiring hardware resource data, business load data and system configuration data of a system to be verified, wherein the acquisition mode is combined with real-time acquisition and history data import; S2, preprocessing information data, namely sequentially performing data cleaning, data noise reduction, data normalization and key feature extraction on the data acquired in the step S1 to obtain key feature data; S3, constructing a simulation model, namely constructing a system performance simulation model comprising a hardware resource sub-model, a business load sub-model and a system interaction sub-model by adopting a simulation tool based on key characteristic data; s4, calibrating a simulation model, namely comparing actual operation data of the system to be verified with simulation indexes of the simulation model, and adjusting model parameters until the deviation rate of the simulation indexes and the actual indexes is less than or equal to 5%; s5, performance simulation verification, namely setting conventional, sudden load and service switching scenes based on the calibrated model, running simulation and collecting performance indexes, and taking the average value of multiple simulation results as a final simulation result; And S6, analyzing the simulation result, namely comparing the final simulation result with the system performance requirement index, diagnosing the performance bottleneck and generating a performance simulation verification report. Preferably, in the step S1, the hardware resource data includes CPU utilization, memory occupancy, disk I/O rate and network bandwidth utilization, the service load data includes service request amount, request response time and request failure rate, the system configuration data includes process number, thread scheduling policy and buffer size, real-time acquisition is implemented through SNMP or promethaus interface, sampling frequency is 1-10Hz, historical data is extracted from an elastic search log database, real-time acquired data is preliminarily filtered through an edge computing node, only effective data is transmitted to a data processing center, data transmission amount is reduced, and filtering rules include eliminating data ex