Network Fault Management System and Method Based on Multi-Source Data

Abstract

This invention discloses a network fault management system and method based on multi-source data, relating to the field of server hard drive fault monitoring technology. Specifically, it includes the following steps: collecting hardware health data, service performance data, network traffic data, and historical fault records, and determining the collection frequency; acquiring key features of server hard drive faults and performing data quantification processing on these key features; using the pre-processed data as a benchmark, selecting several key feature data values ​​of the server hard drive, and constructing a server hard drive fault probability prediction model through logistic regression; acquiring real-time key feature data, calculating the probability of server hard drive fault occurrence through the server hard drive fault probability prediction model, and comparing it with a preset classification range to obtain the fault level; and triggering an automated response action based on the fault level. This invention effectively reduces fault diagnosis time and service interruption risk, achieving more intelligent and efficient network fault management.

Description

Technical Field

[0001] This invention relates to the field of server hard drive fault monitoring technology, specifically a network fault management system and method based on multi-source data. Background Technology

[0002] In modern data centers and high-concurrency business environments, server hard drives, as the core of storage, endure high-intensity read / write loads for extended periods, making them susceptible to physical damage, environmental factors, and software anomalies. This can lead to performance degradation or even failure, consequently impacting business continuity. While the widely used SMART technology provides basic health monitoring data such as hard drive temperature, bad sector count, and I / O speed, and AIOps technology uses machine learning models for fault prediction, existing methods still have several shortcomings. On one hand, traditional data collection typically uses a fixed frequency, failing to dynamically adjust based on network traffic. This results in over-sampling at low loads, wasting resources, and under-sampling at high loads, missing critical fault signals. On the other hand, the selection of key features often relies on empirical judgment, failing to fully utilize business performance data and network traffic data, thus affecting prediction accuracy. Furthermore, existing fault probability calculation methods do not fully consider the non-linear changes in hard drive lifespan, simply using linear regression or statistical thresholds, ignoring the different characteristics of early failures, stable periods, and wear-out periods, leading to significant prediction errors. Simultaneously, traditional operations and maintenance mainly rely on manual monitoring and threshold alarms, lacking intelligent decision-making mechanisms, resulting in delayed fault response and increased risk of business interruption. Summary of the Invention

[0003] The purpose of this invention is to provide a network fault management system and method based on multi-source data to solve the problems raised in the prior art.

[0004] To achieve the above objectives, the present invention provides the following technical solution: a network fault management method based on multi-source data, wherein the network fault management method specifically includes the following steps:

[0005] Step S100: Collect hardware health data, service performance data, network traffic data and historical fault records, construct an adaptive sampling frequency control function, determine the sampling frequency, and update the historical fault records;

[0006] Step S200: Obtain key features of server hard disk failure and perform data quantification processing on the key features;

[0007] Step S300: Based on the preprocessed data, select key feature data values ​​of several server hard drives and construct a server hard drive failure probability prediction model through logistic regression.

[0008] Step S400: Obtain real-time key feature data, calculate the probability of server hard drive failure using the server hard drive failure probability prediction model, and compare it with the preset classification range to obtain the failure level.

[0009] Step S500: Trigger an automated response action based on the fault level.

[0010] In step S100, hardware health data, service performance data, network traffic data, and historical fault records are collected. An adaptive sampling frequency control function is constructed to determine the collection frequency for different data, and the historical fault records are updated. Specifically, this includes the following steps:

[0011] Step S101: Collect hardware health data, service performance data, network traffic data, and historical fault records; the hardware health data includes server hard drive temperature, response time, read error rate, and sector anomaly ratio; the service performance data includes database query latency, page load time, and request timeout rate; the network traffic data includes disk IOPS and switch port bandwidth utilization; the historical fault records include server hard drive model, fault time, fault type, and maintenance records;

[0012] Step S102: Construct an adaptive sampling frequency control function to determine the acquisition frequency for different data.

[0013] Step S103: Update the historical fault records periodically.

[0014] In step S102, an adaptive sampling frequency control function is constructed to determine the sampling frequency, specifically as follows:

[0015] Construct an adaptive sampling frequency control function based on changes in network traffic;

[0016] The adaptive sampling frequency control function includes a basic sampling frequency term, a traffic load sensing term, and a traffic mutation response term.

[0017] ;

[0018] Where f represents the adaptive sampling frequency; Indicates the basic sampling frequency; Represents real-time network traffic; and These represent the historical peak and trough values ​​of network traffic, respectively. This represents the rate of change in real-time network traffic; This indicates the flow load control parameters.

[0019] In step S200, the key characteristics of the server hard drive failure are obtained, specifically:

[0020] Step S201: Arrange the hardware health data, service performance data, and network traffic data in descending order along with the cumulative working time of the server hard drive, and denote them as sequences r1 and r2.

[0021] Step S202: Calculate the rank of each data point in its respective sequence, that is, find the position of the data point in the sequence. For repeated data points, the average rank method is adopted.

[0022] Step S203: Calculate the rank difference d, specifically: subtract the rank of the corresponding data point in r2 from the rank of the corresponding data point in r1;

[0023] Step S204: Calculate the correlation coefficient between different data and the cumulative working time of the server hard drive by using the change relationship of rank difference;

[0024] As a preferred option, the specific calculation formula is as follows:

[0025] ;

[0026] Where ρ represents the correlation coefficient between the data and the cumulative working time of the server hard drive; N represents the amount of data, where N is a positive integer; d i This represents the rank difference of the i-th data.

[0027] Step S205: Identify data points with correlation coefficients greater than the average as key features, denoted as w1, w2, ..., w... n Among them, w1, w2, ..., w n This represents the 1st, 2nd, ..., nth key features, where n is a positive integer.

[0028] The key features are subjected to data quantization processing, specifically as follows:

[0029] Step S211: Based on the obtained key features, obtain the historical maximum value w of the key features. max Compared with the historical minimum value w min ;

[0030] Step S212: Perform data quantization processing on the acquired key feature data, specifically as follows:

[0031] δ(w)=(ww min ) / (w max -w min ); where δ(w) represents the standardized function of the key feature w.

[0032] In step S300, based on the preprocessed data, several key feature data values ​​of the server hard drives are selected, and a server hard drive failure probability prediction model is constructed through logistic regression, specifically as follows:

[0033] Step S301: Based on historical fault records and using the pre-processed data as a benchmark, select key feature data values ​​of several server hard drives as training data.

[0034] Step S302: Perform data quantization processing on the training data;

[0035] Step S303: Construct a decay function for the cumulative runtime of the server hard drive, and combine it with the training data after data quantization processing to construct a server hard drive failure probability prediction model through logistic regression.

[0036] Preferably, the characterization formula of the server hard drive failure probability prediction model is as follows:

[0037] ;

[0038] Where p represents the probability of server hard drive failure; T represents the cumulative operating time of the server hard drive; w j ρ represents the j-th key feature, where j is the key feature label, j is a positive integer, and j∈[1,n]; j The weight coefficient of the j-th key feature is represented by the correlation coefficient between the data and the cumulative working time of the server hard drive; e represents the natural constant; λ(t) represents the decay function of the cumulative running time of the server hard drive. Represents the j-th key feature The standardized function.

[0039] Specifically, in step S303, the decay function for the cumulative runtime of the server hard drive is defined as follows:

[0040] A three-stage hybrid model is used to obtain the decay function of the cumulative runtime of the server hard drive. The three-stage hybrid model includes an early failure period, a stable period, and a wear-out period.

[0041] Data simulations were conducted for three different periods to obtain the decay function of the cumulative runtime of the server's hard drive.

[0042] Preferably, the expression for the decay function of the server hard disk's cumulative runtime is as follows:

[0043] ;in, The decay function value represents the cumulative runtime of the server hard drive; t represents the cumulative runtime of the server hard drive.

[0044] The fitted expression for the early failure period is characterized by exponential decay. ;

[0045] α and β represent the initial risk and decay rate during the early failure period, respectively; β is the improvement rate of the early failure; the fitting expression for the stable period is represented as a constant. The The random failure rate (environmental factors) during the stable period; the fitted expression for the wear-out period is characterized by power-law growth. The The intensity coefficient representing the failure growth over the wear-out period is typically obtained from the manufacturer's reliability report; The acceleration index representing the risk of wear and tear; This is the threshold time at which the wear-out period begins.

[0046] In step S400, real-time key feature data is acquired, the probability of server hard drive failure is calculated using the server hard drive failure probability prediction model, and the failure level is obtained by comparing it with a preset classification range. Specifically:

[0047] Step S401: Obtain real-time key feature data;

[0048] Step S402: Perform data quantization processing using the method described in step S212;

[0049] Step S403: Calculate the probability of server hard drive failure using the server hard drive failure probability prediction model;

[0050] Step S404: Compare with the preset grading range to obtain the fault level; the grading range is set according to the rules of normal distribution, and the grading range is preset based on the sum of the average value and an integer multiple of the standard deviation of the key features.

[0051] In step S500, an automated response action is triggered based on the fault level, specifically as follows:

[0052] Based on historical fault records, a fuzzy control table is preset;

[0053] The response strategy is executed automatically based on the fuzzy control table.

[0054] A network fault management system based on multi-source data, the network fault management system includes a data acquisition module, a data processing module, a fault prediction module, an automated operation and maintenance module, and an early warning and visualization module;

[0055] The data acquisition module is used to collect hardware health data, service performance data, and network traffic data; and to construct an adaptive sampling frequency control function to adjust the sampling frequency.

[0056] The data processing module is used to determine key features by analyzing the changes in rank differences; and to perform data quantization based on the historical maximum and minimum values ​​of the key features.

[0057] The fault prediction module is used to build a server hard drive fault probability prediction model through logistic regression and predict the probability of server hard drive failure based on real-time key feature data.

[0058] The automated operation and maintenance module is used to compare the probability of failure occurrence with a preset classification range to obtain the failure level; and to automatically execute response strategies according to a preset fuzzy control table.

[0059] The early warning and visualization module is used to issue early warnings based on the fault level and provide a visual display of network traffic.

[0060] The output of the data acquisition module is connected to the input of the data processing module; the output of the data processing module is connected to the input of the fault prediction module; the output of the fault prediction module is connected to the input of the automated operation and maintenance module; and the output of the automated operation and maintenance module is connected to the input of the visualization module.

[0061] Compared with existing technologies, the beneficial effects of this invention are as follows: This invention integrates hardware health data, service performance data, network traffic data, and historical fault records to optimize data collection strategies and improves the accuracy of server hard drive failure prediction based on logistic regression and hard drive lifespan decay models. This method employs an adaptive sampling frequency control function to dynamically adjust the data collection frequency according to network traffic, ensuring that key fault signals are not missed while avoiding resource waste. Furthermore, it utilizes rank correlation analysis to screen key features and combines standardization processing to improve the model's robustness and generalization ability. The fault prediction model incorporates the three-stage decay characteristics of hard drive lifespan, making the prediction results more consistent with the actual operating patterns of hard drives, thereby improving the reliability of early warnings. Based on the prediction results, this method can also provide automated response strategies, triggering corresponding maintenance measures according to preset fault levels, reducing fault investigation time and service interruption risks, and achieving more intelligent and efficient network fault management. Attached Figure Description

[0062] Figure 1 This is a schematic diagram illustrating the steps of a network fault management method based on multi-source data according to the present invention;

[0063] Figure 2 This is a schematic diagram of the structure of a network fault management system based on multi-source data according to the present invention. Detailed Implementation

[0064] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0065] Example: Figures 1-2 As shown, the present invention provides a technical solution, a network fault management method based on multi-source data, the network fault management method specifically including the following steps:

[0066] Step S100: Collect hardware health data, service performance data, network traffic data and historical fault records, construct an adaptive sampling frequency control function, determine the sampling frequency, and update the historical fault records;

[0067] Step S200: Obtain key features of server hard disk failure and perform data quantification processing on the key features;

[0068] Step S300: Based on the preprocessed data, select key feature data values ​​of several server hard drives and construct a server hard drive failure probability prediction model through logistic regression.

[0069] Step S400: Obtain real-time key feature data, calculate the probability of server hard drive failure using the server hard drive failure probability prediction model, and compare it with the preset classification range to obtain the failure level.

[0070] Step S500: Trigger an automated response action based on the fault level.

[0071] In step S100, hardware health data, service performance data, network traffic data, and historical fault records are collected. An adaptive sampling frequency control function is constructed to determine the collection frequency for different data, and the historical fault records are updated. Specifically, this includes the following steps:

[0072] Step S101: Collect hardware health data, service performance data, network traffic data, and historical fault records; the hardware health data includes server hard drive temperature, response time, read error rate, and sector anomaly ratio; the service performance data includes database query latency, page load time, and request timeout rate; the network traffic data includes disk IOPS and switch port bandwidth utilization; the historical fault records include server hard drive model, fault time, fault type, and maintenance records;

[0073] Step S102: Construct an adaptive sampling frequency control function to determine the acquisition frequency for different data.

[0074] Step S103: Update the historical fault records periodically.

[0075] In step S102, an adaptive sampling frequency control function is constructed to determine the sampling frequency, specifically as follows:

[0076] Construct an adaptive sampling frequency control function based on changes in network traffic;

[0077] The adaptive sampling frequency control function includes a basic sampling frequency term, a traffic load sensing term, and a traffic mutation response term.

[0078] ;

[0079] Where f represents the adaptive sampling frequency; Indicates the basic sampling frequency; Represents real-time network traffic; and These represent the historical peak and trough values ​​of network traffic, respectively. This represents the rate of change in real-time network traffic; This indicates the flow load control parameters.

[0080] In step S200, the key characteristics of the server hard drive failure are obtained, specifically:

[0081] Step S201: Arrange the hardware health data, service performance data, and network traffic data in descending order along with the cumulative working time of the server hard drive, and denote them as sequences r1 and r2.

[0082] Step S202: Calculate the rank of each data point in its respective sequence, that is, find the position of the data point in the sequence. For repeated data points, the average rank method is adopted.

[0083] Step S203: Calculate the rank difference d, specifically: subtract the rank of the corresponding data point in r2 from the rank of the corresponding data point in r1;

[0084] Step S204: Calculate the correlation coefficient between different data and the cumulative working time of the server hard drive by using the change relationship of rank difference;

[0085] As a preferred option, the specific calculation formula is as follows:

[0086] ;

[0087] Where ρ represents the correlation coefficient between the data and the cumulative working time of the server hard drive; N represents the amount of data, where N is a positive integer; d i This represents the rank difference of the i-th data.

[0088] Step S205: Identify data points with correlation coefficients greater than the average as key features, denoted as w1, w2, ..., w... n Among them, w1, w2, ..., w n This represents the 1st, 2nd, ..., nth key features, where n is a positive integer.

[0089] The key features are subjected to data quantization processing, specifically as follows:

[0090] Step S211: Based on the obtained key features, obtain the historical maximum value w of the key features. max Compared with the historical minimum value w min ;

[0091] Step S212: Perform data quantization processing on the acquired key feature data, specifically as follows:

[0092] δ(w)=(ww min ) / (w max -w min ); where δ(w) represents the standardized function of the key feature w.

[0093] In step S300, based on the preprocessed data, several key feature data values ​​of the server hard drives are selected, and a server hard drive failure probability prediction model is constructed through logistic regression, specifically as follows:

[0094] Step S301: Based on historical fault records and using the pre-processed data as a benchmark, select key feature data values ​​of several server hard drives as training data.

[0095] Step S302: Perform data quantization processing on the training data;

[0096] Step S303: Construct a decay function for the cumulative runtime of the server hard drive, and combine it with the training data after data quantization processing to construct a server hard drive failure probability prediction model through logistic regression.

[0097] Preferably, the characterization formula of the server hard drive failure probability prediction model is as follows:

[0098] ;

[0099] Where p represents the probability of server hard drive failure; T represents the cumulative operating time of the server hard drive; w j ρ represents the j-th key feature, where j is the key feature label, j is a positive integer, and j∈[1,n]; jThe weight coefficient of the j-th key feature is represented by the correlation coefficient between the data and the cumulative working time of the server hard drive; e represents the natural constant; λ(t) represents the decay function of the cumulative running time of the server hard drive. Represents the j-th key feature The standardized function.

[0100] Specifically, in step S303, the decay function for the cumulative runtime of the server hard drive is defined as follows:

[0101] A three-stage hybrid model is used to obtain the decay function of the cumulative runtime of the server hard drive. The three-stage hybrid model includes an early failure period, a stable period, and a wear-out period.

[0102] Data simulations were conducted for three different periods to obtain the decay function of the cumulative runtime of the server's hard drive.

[0103] Preferably, the expression for the decay function of the server hard disk's cumulative runtime is as follows:

[0104] ;in, The decay function value represents the cumulative runtime of the server hard drive; t represents the cumulative runtime of the server hard drive.

[0105] The fitted expression for the early failure period is characterized by exponential decay. ;

[0106] α and β represent the initial risk and decay rate during the early failure period, respectively;

[0107] α is the initial failure rate caused by manufacturing defects at the beginning of the manufacturing process, which is obtained by fitting the manufacturer's MTBF data.

[0108] β is the improvement rate of early failures, preferably obtained by regression training using the first 1000 hours of failure data;

[0109] The fitting expression for the stable period is characterized as a constant. ;

[0110] The The random failure rate (environmental factors) during the stable period is represented; preferably, the long-term average of the historical failure rate is used as the reference. The value;

[0111] The fitted expression for the attenuation period is characterized by power-law growth. ;

[0112] The The intensity factor representing the failure growth during the wear-out period is usually obtained from the manufacturer's reliability report;

[0113] The The acceleration index representing the risk during the attrition period; obtained by fitting the shape parameters of the Weibull distribution;

[0114] The This is the threshold time at which the wear-out period begins, typically 0.7 times the manufacturer's nominal lifespan.

[0115] In step S400, real-time key feature data is acquired, the probability of server hard drive failure is calculated using the server hard drive failure probability prediction model, and the failure level is obtained by comparing it with a preset classification range. Specifically:

[0116] Step S401: Obtain real-time key feature data;

[0117] Step S402: Perform data quantization processing using the method described in step S212;

[0118] Step S403: Calculate the probability of server hard drive failure using the server hard drive failure probability prediction model;

[0119] Step S404: Compare with the preset grading range to obtain the fault level; the grading range is set according to the rules of normal distribution, and the grading range is preset based on the sum of the average value and an integer multiple of the standard deviation of the key features.

[0120] In step S500, an automated response action is triggered based on the fault level, specifically as follows:

[0121] Based on historical fault records, a fuzzy control table is preset;

[0122] The response strategy is executed automatically based on the fuzzy control table.

[0123] Example 1,

[0124] By integrating SMART into the hard drive controller, data monitoring functions are realized, continuously tracking various key operating parameters and status information of the hard drive; using the sensors and logic circuits inside the hard drive, dozens of different data indicators such as head seek time, motor speed, temperature, data transfer rate, and sector error rate are collected in real time.

[0125] Adaptive data acquisition:

[0126] Hardware parameters:

[0127] Hard drive model: HDD-X***;

[0128] Total operating time: 20,000 hours;

[0129] Network bandwidth: 1Gbps;

[0130] Real-time data collection: (Data type - Indicator name - Real-time value):

[0131] Hardware health data - Temperature -50℃;

[0132] Response time -8ms;

[0133] Read error rate -0.15%;

[0134] Sector anomaly ratio -0.02%;

[0135] Business performance data - Database query latency - 120ms;

[0136] Page load time -2.1s;

[0137] Request timeout rate -1.8%;

[0138] Network traffic data - disk;

[0139] IOPS-8500;

[0140] Port bandwidth utilization -85%;

[0141] Basic sampling frequency: f base =10Hz;

[0142] Network traffic parameters:

[0143] G max =900Mbps (historical peak).

[0144] G min =200Mbp (historical low value);

[0145] Real-time traffic G = 850Mbps;

[0146] Based on the continuously collected traffic flow, the traffic flow change rate ΔG = 50 Mbps / s was obtained;

[0147] Flow load control parameters Take 0.5;

[0148] The adaptive sampling frequency f is approximately 14Hz. (The sampling frequency is adjusted according to real-time network traffic.)

[0149] Example 2,

[0150] By analyzing the changes in rank differences, the correlation coefficient between different data and the cumulative working time of the server hard drive was calculated. Some results are shown in Table 1:

[0151]

[0152] Table 1

[0153] Example 3,

[0154] The grading range is set according to the rules of normal distribution. The grading range is preset based on the sum of the average value and an integer multiple of the standard deviation of the key features.

[0155] Taking response time as an example (approximate value), by observing the approximately linear distribution of data points, it conforms to normality;

[0156] Mean μ = 50 ms; Standard deviation σ' = 10 ms; (L represents the range identifier of response time):

[0157]

[0158] Table 2

[0159] The same analysis was performed on different key features, and the results were substituted into the server hard drive failure probability model to calculate the different ranges of runtime.

[0160] A network fault management system based on multi-source data, the network fault management system includes a data acquisition module, a data processing module, a fault prediction module, an automated operation and maintenance module, and an early warning and visualization module;

[0161] The data acquisition module is used to collect hardware health data, service performance data, and network traffic data; and to construct an adaptive sampling frequency control function to adjust the sampling frequency.

[0162] The data processing module is used to determine key features by analyzing the changes in rank differences; and to perform data quantization based on the historical maximum and minimum values ​​of the key features.

[0163] The fault prediction module is used to build a server hard drive fault probability prediction model through logistic regression and predict the probability of server hard drive failure based on real-time key feature data.

[0164] The automated operation and maintenance module is used to compare the probability of failure occurrence with a preset classification range to obtain the failure level; and to automatically execute response strategies according to a preset fuzzy control table.

[0165] The early warning and visualization module is used to issue early warnings based on the fault level and provide a visual display of network traffic.

[0166] The output of the data acquisition module is connected to the input of the data processing module; the output of the data processing module is connected to the input of the fault prediction module; the output of the fault prediction module is connected to the input of the automated operation and maintenance module; and the output of the automated operation and maintenance module is connected to the input of the visualization module.

[0167] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the invention can be implemented in other specific forms without departing from its spirit or essential characteristics. Therefore, the embodiments should be considered in all respects as exemplary and non-limiting, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of equivalents of the claims are intended to be included within the present invention. No reference numerals in the claims should be construed as limiting the scope of the claims.

Claims

1. A network fault management method based on multi-source data, characterized in that: The network fault management method specifically includes the following steps: Step S100: Collect hardware health data, service performance data, network traffic data and historical fault records, construct an adaptive sampling frequency control function, determine the sampling frequency, and update the historical fault records; Step S200: Obtain key features of server hard disk failure and perform data quantification processing on the key features; Specifically: Step S201: Arrange the hardware health data, service performance data, and network traffic data in descending order along with the cumulative working time of the server hard drive, and denote them as sequences r1 and r2. Step S202: Calculate the rank of each data point in its respective sequence, that is, find the position of the data point in the sequence. For repeated data points, the average rank method is adopted. Step S203: Calculate the rank difference d, specifically: subtract the rank of the corresponding data point in r2 from the rank of the corresponding data point in r1; Step S204: Calculate the correlation coefficient between different data and the cumulative working time of the server hard drive based on the changes in rank differences; the specific calculation formula is as follows: ； Where ρ represents the correlation coefficient between the data and the cumulative working time of the server hard drive; N represents the amount of data, where N is a positive integer; d i This represents the rank difference of the i-th data. Step S205: Identify data points with correlation coefficients greater than the average as key features, denoted as w1, w2, ..., w... n Among them, w1, w2, ..., w n These represent the 1st, 2nd, ..., nth key features, where n is a positive integer; Step S300: Based on the preprocessed data, select key feature data values ​​of several server hard drives and construct a server hard drive failure probability prediction model through logistic regression. Step S400: Obtain real-time key feature data, calculate the probability of server hard drive failure using the server hard drive failure probability prediction model, and compare it with the preset classification range to obtain the failure level. Step S500: Trigger an automated response action based on the fault level.

2. The network fault management method based on multi-source data according to claim 1, characterized in that: In step S100, hardware health data, service performance data, network traffic data, and historical fault records are collected. An adaptive sampling frequency control function is constructed to determine the collection frequency for different data, and the historical fault records are updated. Specifically, this includes the following steps: Step S101: Collect hardware health data, service performance data, network traffic data, and historical fault records; the hardware health data includes server hard drive temperature, response time, read error rate, and sector anomaly ratio; the service performance data includes database query latency, page load time, and request timeout rate; the network traffic data includes disk IOPS and switch port bandwidth utilization; the historical fault records include server hard drive model, fault time, fault type, and maintenance records; Step S102: Construct an adaptive sampling frequency control function to determine the acquisition frequency for different data. Step S103: Update the historical fault records periodically.

3. The network fault management method based on multi-source data according to claim 2, characterized in that: In step S102, an adaptive sampling frequency control function is constructed to determine the sampling frequency, specifically as follows: Construct an adaptive sampling frequency control function based on changes in network traffic; The adaptive sampling frequency control function includes a basic sampling frequency term, a traffic load sensing term, and a traffic mutation response term.

4. The network fault management method based on multi-source data according to claim 3, characterized in that: The key features are subjected to data quantization processing, specifically as follows: Step S211: Based on the obtained key features, obtain the historical maximum value w of the key features. max Compared with the historical minimum value w min ; Step S212: Perform data quantization processing on the acquired key feature data, specifically as follows: δ(w)=(ww min ) / (w max -w min ); where δ(w) represents the standardized function of the key feature w.

5. The network fault management method based on multi-source data according to claim 4, characterized in that: In step S300, based on the preprocessed data, several key feature data values ​​of the server hard drives are selected, and a server hard drive failure probability prediction model is constructed through logistic regression, specifically as follows: Step S301: Based on historical fault records and using the pre-processed data as a benchmark, select key feature data values ​​of several server hard drives as training data. Step S302: Perform data quantization processing on the training data; Step S303: Construct a decay function for the cumulative runtime of the server hard drive, and combine it with the training data after data quantization processing to construct a server hard drive failure probability prediction model through logistic regression.

6. The network fault management method based on multi-source data according to claim 5, characterized in that: Specifically, in step S303, the decay function for the cumulative runtime of the server hard drive is defined as follows: A three-stage hybrid model is used to obtain the decay function of the cumulative runtime of the server hard drive. The three-stage hybrid model includes an early failure period, a stable period, and a wear-out period. Data simulations were conducted for three different periods to obtain the decay function of the server's cumulative operating time.

7. A network fault management method based on multi-source data according to claim 6, characterized in that: In step S400, real-time key feature data is acquired, the probability of server hard drive failure is calculated using the server hard drive failure probability prediction model, and the failure level is obtained by comparing it with a preset classification range. Specifically: Step S401: Obtain real-time key feature data; Step S402: Perform data quantization processing using the method described in step S212; Step S403: Calculate the probability of server hard drive failure using the server hard drive failure probability prediction model; Step S404: Compare with the preset grading range to obtain the fault level; the grading range is set according to the rules of normal distribution, and the grading range is preset based on the sum of the average value and an integer multiple of the standard deviation of the key features.

8. The network fault management method based on multi-source data according to claim 7, characterized in that: In step S500, an automated response action is triggered based on the fault level, specifically as follows: Based on historical fault records, a fuzzy control table is preset; The response strategy is executed automatically based on the fuzzy control table.

9. A network fault management system based on multi-source data, employing the network fault management method based on multi-source data as described in any one of claims 1-8, characterized in that: The network fault management system includes a data acquisition module, a data processing module, a fault prediction module, an automated operation and maintenance module, and an early warning and visualization module. The data acquisition module is used to collect hardware health data, service performance data, and network traffic data; and to construct an adaptive sampling frequency control function to adjust the sampling frequency. The data processing module is used to sort hardware health data, business performance data, and network traffic data with the cumulative working time of the server hard drive in descending order, calculate the rank of each data point in its respective sequence, that is, find the position of the data point in the sequence. For duplicate data points, the average rank method is adopted, and the key features are determined by the change relationship of the rank difference. The data is then quantified based on the historical maximum and historical minimum values ​​of the key features. The fault prediction module is used to build a server hard drive fault probability prediction model through logistic regression and predict the probability of server hard drive failure based on real-time key feature data. The automated operation and maintenance module is used to compare the probability of failure occurrence with a preset classification range to obtain the failure level; and to automatically execute response strategies according to a preset fuzzy control table. The early warning and visualization module is used to issue early warnings based on the fault level and to provide a visual display of network traffic. The output of the data acquisition module is connected to the input of the data processing module; the output of the data processing module is connected to the input of the fault prediction module; the output of the fault prediction module is connected to the input of the automated operation and maintenance module; and the output of the automated operation and maintenance module is connected to the input of the visualization module.

Images