Big data-based information system comprehensive operation and maintenance management platform method and system

Abstract

The application relates to the technical field of operation and maintenance management, in particular to a big data-based information system comprehensive operation and maintenance management platform method and system, which comprises the following steps: collecting system operation data of a specified enterprise information system in real time, storing and preprocessing the system operation data, and outputting a system operation data set; performing time axis calibration and fusion processing on the system operation data set, and outputting a multidimensional correlation data set; establishing a dynamic correlation model based on the multidimensional correlation data set, monitoring abnormal conditions in real time, and outputting an early warning level; matching a candidate treatment scheme set based on the early warning level; selecting an optimal treatment scheme from the candidate treatment scheme set based on a preset screening rule and executing the optimal treatment scheme; and outputting an execution effect evaluation result of the optimal treatment scheme and feeding back the execution effect evaluation result to a preset big data platform. The application has the advantages of real-time discovery of equipment hidden dangers and automatic selection of optimal solution effects.

Description

Technical Field

[0001] This application relates to the field of operation and maintenance management technology, and in particular to a method and system for a comprehensive operation and maintenance management platform for information systems based on big data. Background Technology

[0002] Traditional operation and maintenance (O&M) models suffer from the following core technical deficiencies, making it difficult to meet the stability and agility requirements of enterprise IT systems: 1. Real-time detection challenges of progressive anomalies: Traditional threshold monitoring (such as CPU > 80% alarms) cannot capture progressive faults such as memory leaks and service response degradation, leading to delayed fault detection and a wider impact. 2. Lack of cross-domain data correlation analysis capabilities: Server, network, and application system data are isolated, lacking automated cross-domain correlation analysis methods. Manual root cause diagnosis is time-consuming, resulting in a high MTTR (Mean Time To Repair). 3. Bottlenecks in quantitative assessment of handling decisions: Solution formulation relies on human experience, lacking quantitative assessment of business impact, cost, complexity, and other dimensions, leading to "dilemmas," such as conflicts between service restart and business continuity. 4. Disruptions in the closed-loop mechanism of O&M optimization: The handling effect does not form a feedback loop, similar faults recur, model iteration relies on manual annotation, the system self-optimization cycle is long, and it is difficult to adapt to dynamic business needs.

[0003] Therefore, improvements are needed. Summary of the Invention

[0004] To address the above technical issues, this application provides a method and system for a comprehensive information system operation and maintenance management platform based on big data.

[0005] The first objective of this invention is achieved through the following technical solution:

[0006] A method for a comprehensive operation and maintenance management platform for information systems based on big data includes the following steps:

[0007] The system collects system operation data of a specified enterprise information system in real time, stores and preprocesses the system operation data, and outputs a system operation dataset.

[0008] The system runtime dataset is subjected to time axis calibration and fusion processing to output a multi-dimensional correlated dataset;

[0009] A dynamic association model is established based on the multi-dimensional association dataset to monitor abnormal situations in real time and output early warning levels.

[0010] Based on the aforementioned warning level, a set of candidate response plans will be matched;

[0011] Based on preset screening rules, the optimal treatment plan is selected from the set of candidate treatment plans and then executed.

[0012] Output the evaluation results of the optimal handling solution and feed them back to the preset big data platform.

[0013] In a preferred embodiment, the step of real-time collection of system operation data of a specified enterprise information system, storage and preprocessing of the system operation data, and outputting a system operation dataset includes:

[0014] The system operation data of the specified enterprise information system is collected in real time through sensor acquisition, log file extraction, and API interface reception.

[0015] The system operation data includes server performance indicators and network device status data;

[0016] The service performance metrics include CPU utilization, memory usage, and disk space.

[0017] The network device status data includes network traffic, number of network connections, packet loss rate, and application runtime logs.

[0018] Based on a pre-set big data platform, the system operation data is stored and preprocessed to output a system operation dataset.

[0019] In a preferred embodiment, the step of performing time-axis calibration and fusion processing on the system runtime dataset to output a multi-dimensional correlated dataset includes:

[0020] For the system operation data of two adjacent related nodes in the system operation dataset, the system operation data is aligned by eliminating the acquisition time deviation based on the preset data interpolation algorithm;

[0021] Extract temporal and spatial features from the aligned system operation data;

[0022] The time characteristics include the collection time, trend of change, and periodicity of system operation data;

[0023] The spatial characteristics include the source device, device location, device type, and data association path of the system operation data;

[0024] The extracted temporal and spatial features are input into a pre-trained weighted spatiotemporal correlation matrix for fusion.

[0025] Based on the fusion results, multi-dimensional data fusion calculations are performed, including but not limited to device operation status fusion, service response latency fusion, and business health indicator fusion.

[0026] The results of the fusion computing are used to generate a multi-dimensional related dataset according to a predefined data structure and format. The multi-dimensional dataset includes device dimension, service dimension, business dimension, and time dimension.

[0027] In a preferred embodiment, the step of establishing a dynamic association model based on the multi-dimensional association dataset, monitoring anomalies in real time, and outputting early warning levels includes:

[0028] For the aforementioned multi-dimensional associated dataset, a dynamic correlation model between hardware performance indicators and service response quality is established;

[0029] Based on the aforementioned dynamic correlation model, abnormal hardware parameters are monitored in real time, and corresponding service response time degradation is detected simultaneously.

[0030] Once both abnormal hardware parameters and deteriorating service response time are detected, the corresponding warning level will be output.

[0031] In a preferred embodiment, the step of matching and outputting a corresponding warning level once both hardware parameter anomalies and service response time degradation are detected simultaneously includes:

[0032] For abnormal values ​​of hardware parameters, calculate the deviation L1;

[0033] Calculate the degradation rate L2 for service response time degradation;

[0034] Based on the preset formula S=α·L1+β·L2, where α+β=1, a dynamic score S is generated;

[0035] When S≥δ1 and δ1 is the first warning threshold, a first-level warning is matched;

[0036] When δ2≤S<δ1 and δ2 is the second warning threshold, a level-two warning is matched;

[0037] When δ3≤S<δ2 and δ3 is the third warning threshold, a level 3 warning is matched.

[0038] In a preferred embodiment, the step of matching a set of candidate treatment plans based on the warning level includes:

[0039] Based on the handling priority corresponding to the warning level, candidate solutions that meet the following conditions are selected from a preset historical case database:

[0040] The candidate solution type must match the current warning level with a preset similarity threshold, and the historical execution success rate of the candidate solution must not be lower than a preset success rate threshold.

[0041] After the filtering is complete, output the Top-N candidate solutions that meet the criteria;

[0042] Within the Top-N candidate solution set, for different warning levels, a pre-defined differentiated handling strategy is executed, and a candidate handling solution set is output.

[0043] The preset differentiated handling strategy includes:

[0044] When it is a Level 1 warning, match the set of emergency plans that need to be implemented immediately;

[0045] When it is a Level II warning, match the optimized solution set that requires a time limit for handling;

[0046] When it is a Level 3 warning, a corresponding observation period response plan will be implemented.

[0047] In a preferred embodiment, the step of selecting the optimal treatment plan from the candidate treatment plan set based on preset screening rules and executing it includes:

[0048] Set up business impact dimension, resource cost dimension, operational complexity dimension and historical success dimension, and assign preset weights W1, W2, W3 and W4 to each dimension;

[0049] After normalizing the scores of each candidate solution across all dimensions, the results are then processed based on a pre-defined formula. (Dimensional Scoring) k ×weight W k ), calculate the overall score;

[0050] The comprehensive scores are sorted from high to low, and the candidate solution with the highest comprehensive score is selected as the optimal solution.

[0051] The second objective of this invention is achieved through the following technical solution:

[0052] A comprehensive information system operation and maintenance management platform based on big data includes:

[0053] Data Acquisition Module: Collects system operation data of a specified enterprise information system in real time, stores and preprocesses the system operation data, and outputs a system operation dataset;

[0054] Fusion module: Performs time axis calibration and fusion processing on the system operation dataset, and outputs a multi-dimensional correlated dataset;

[0055] Output module: Based on the multi-dimensional associated dataset, a dynamic association model is established to monitor abnormal situations in real time and output the early warning level;

[0056] Matching module: Based on the warning level, matches a set of candidate treatment plans;

[0057] Filtering module: Based on preset filtering rules, it filters the optimal treatment plan from the candidate treatment plan set and executes it;

[0058] Feedback module: Outputs the evaluation results of the execution effect of the optimal solution and feeds the evaluation results back to the preset big data platform.

[0059] The third objective of this invention is achieved through the following technical solution:

[0060] A computer device includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of the above-described method for a comprehensive operation and maintenance management platform for an information system based on big data.

[0061] The fourth objective of this invention is achieved through the following technical solution:

[0062] A computer-readable storage medium storing a computer program, which, when executed by a processor, implements the steps of the above-described method for a comprehensive operation and maintenance management platform for an information system based on big data.

[0063] In summary, this application includes at least one of the following beneficial technical effects:

[0064] The system achieves early warning and rapid handling of faults through a fully automated end-to-end process. The technical solution can be broken down into the following steps: The system acquires real-time operational data (such as CPU load, network traffic, and application logs) from servers, network devices, and business systems via distributed acquisition nodes. Streaming processing technology is used to clean, standardize, and initially screen for anomalies in the data, ensuring the quality and integrity of the input data and laying a reliable foundation for subsequent analysis. Addressing the time-base differences between different devices, dynamic time alignment technology unifies data time series. A spatiotemporal correlation model integrates multi-dimensional information such as device status, service performance, and business logic to generate a correlated dataset reflecting the overall system operation, significantly improving the efficiency of cross-domain fault tracing. Based on the fused data, a dynamic correlation model is built. Machine learning algorithms identify potential correlations between hardware indicators and service quality. A two-factor anomaly scoring mechanism (hardware parameter deviation + service response degradation rate) enables quantitative early warning, and graded warning signals are output according to preset rules, transforming traditional passive response into proactive prediction. When an alert is triggered, the system automatically matches candidate solutions from the historical case library. It then uses a multi-dimensional evaluation model (considering factors such as comprehensive business impact, execution costs, and operational risks) to select the optimal strategy and execute it. Simultaneously, the results are fed back to the big data platform, driving iterative model optimization and forming a continuous improvement cycle of "monitoring-early warning-response-optimization." This solution, through data-driven intelligent transformation, achieves a leap from single-point monitoring to global control, significantly improving the operational agility and self-healing capabilities of enterprise information systems, effectively reducing failure rates and handling costs, and providing technical support for the stable operation of complex IT environments. Attached Figure Description

[0065] Figure 1This is a flowchart illustrating an embodiment of a big data-based information system integrated operation and maintenance management platform method according to this application.

[0066] Figure 2 This is a flowchart of step S20 in an embodiment of the method for a comprehensive operation and maintenance management platform for an information system based on big data in this application;

[0067] Figure 3 This is a flowchart of step S30 in an embodiment of the method for a comprehensive operation and maintenance management platform for an information system based on big data in this application;

[0068] Figure 4 This is a flowchart of step S303 in an embodiment of the method for a comprehensive operation and maintenance management platform for an information system based on big data in this application;

[0069] Figure 5 This is a schematic block diagram of a computer device according to this application. Detailed Implementation

[0070] The following is in conjunction with the appendix Figure 1-5 This application will be described in further detail. In one embodiment, as... Figure 1 As shown, this application discloses a method for a comprehensive operation and maintenance management platform for information systems based on big data, which specifically includes the following steps:

[0071] S10: Collect system operation data of the designated enterprise information system in real time, store and preprocess the system operation data, and output the system operation dataset;

[0072] S20: Perform time axis calibration and fusion processing on the system operation dataset to output a multi-dimensional correlated dataset;

[0073] S30: Establish a dynamic association model based on the multi-dimensional association dataset, monitor abnormal situations in real time, and output early warning levels;

[0074] S40: Based on the warning level, match a set of candidate handling solutions;

[0075] S50: Based on preset filtering rules, select the optimal treatment plan from the candidate treatment plan set and execute it;

[0076] S60: Output the evaluation result of the execution effect of the optimal solution, and feed the evaluation result back to the preset big data platform.

[0077] In this embodiment, early warning and rapid handling of faults are achieved through a fully automated end-to-end process. The technical solution can be broken down into the following steps: The system acquires real-time operational data (such as CPU load, network traffic, and application logs) from servers, network devices, and business systems via distributed acquisition nodes. Streaming processing technology is used to clean, standardize, and initially screen for anomalies in the data, ensuring the quality and integrity of the input data and laying a reliable foundation for subsequent analysis. Addressing the time-base differences between different devices, dynamic time alignment technology unifies the data time sequence. A spatiotemporal correlation model is then used to integrate multi-dimensional information such as device status, service performance, and business logic, generating a correlated dataset that reflects the overall operational status of the system, significantly improving the efficiency of tracing cross-domain faults. Based on the fused data, a dynamic correlation model is constructed. Machine learning algorithms identify potential correlations between hardware indicators and service quality. A two-factor anomaly scoring mechanism (hardware parameter deviation + service response degradation rate) is used to achieve quantitative early warning. Graded early warning signals are output according to preset rules, transforming traditional passive response into proactive prediction. When an alert is triggered, the system automatically matches candidate solutions from the historical case library. It then uses a multi-dimensional evaluation model (considering factors such as comprehensive business impact, execution costs, and operational risks) to select the optimal strategy and execute it. Simultaneously, the results are fed back to the big data platform, driving iterative model optimization and forming a continuous improvement cycle of "monitoring-early warning-response-optimization." This solution, through data-driven intelligent transformation, achieves a leap from single-point monitoring to global control, significantly improving the operational agility and self-healing capabilities of enterprise information systems, effectively reducing failure rates and handling costs, and providing technical support for the stable operation of complex IT environments.

[0078] Step S10 includes:

[0079] S101: Collects system operation data of a specified enterprise information system in real time through sensor acquisition, log file extraction, and API interface reception;

[0080] S102: The system operation data includes server performance indicators and network device status data;

[0081] S103: The service performance indicators include CPU utilization, memory usage, and disk space;

[0082] S104: The network device status data includes network traffic, number of network connections, packet loss rate, and application running logs;

[0083] S105: Based on a preset big data platform, store and preprocess the system operation data, and output the system operation dataset.

[0084] In this embodiment, server performance (CPU / memory / disk) and network device status (traffic / packet loss rate, etc.) data are collected in real time through multiple heterogeneous sources such as sensors, logs, and APIs, and distributed storage and preprocessing are performed using a big data platform. The core principle lies in constructing a full-data acquisition pipeline, combining it with distributed storage technologies (such as Hadoop / Spark) to achieve high-throughput data processing, and eliminating noise through preprocessing operations such as data cleaning and format standardization. This enables comprehensive monitoring of enterprise IT resources, ensuring data integrity; preprocessing improves subsequent analysis efficiency, reduces false alarm rates in anomaly detection, and provides a reliable data foundation for operational decisions.

[0085] Figure 2 Step S20 includes:

[0086] S201: For the system operation data of two adjacent related nodes in the system operation dataset, based on a preset data interpolation algorithm, eliminate the acquisition time deviation and align the system operation data;

[0087] S202: Extract temporal and spatial features from the aligned system operation data;

[0088] S203: The time characteristics include the collection time, trend of change, and periodicity of system operation data;

[0089] S204: The spatial characteristics include the source device, device location, device type, and data association path of the system operation data;

[0090] S205: Input the extracted temporal and spatial features into the pre-trained weighted spatiotemporal correlation matrix for fusion;

[0091] S206: Based on the fusion results, perform multi-dimensional data fusion calculations, including but not limited to device operation status fusion, service response latency fusion, and business health indicator fusion;

[0092] S207: The results of the fusion calculation are used to generate a multi-dimensional related dataset according to a predefined data structure and format. The multi-dimensional dataset includes device dimension, service dimension, business dimension, and time dimension.

[0093] In this embodiment, a preset data interpolation algorithm (such as linear interpolation and spline interpolation) is used to align the time axis of multi-source data, eliminating acquisition bias. Based on spatiotemporal feature fusion technology (such as LSTM time series model and graph neural network), time features (periodicity / trend) and spatial features (device topology / association path) are mapped to a weighted matrix to generate a four-dimensional correlation dataset of device-service-business-time. This breaks through the limitations of traditional single-dimensional monitoring, enabling cross-device and cross-service correlation analysis. For example, it can trace disk I / O anomalies to database response latency, providing structured data support for root cause localization.

[0094] Figure 3 Step S30 includes:

[0095] S301: For the aforementioned multi-dimensional associated dataset, establish a dynamic correlation model between hardware performance indicators and service response quality;

[0096] S302: Based on the dynamic correlation model, monitor hardware parameter anomalies in real time and simultaneously detect the corresponding service response time degradation.

[0097] S303: Once abnormal hardware parameters and deteriorated service response time are detected simultaneously, the corresponding warning level will be output.

[0098] In this embodiment, a hardware-service dynamic correlation model (such as a causal inference model based on random forest) is constructed to monitor the dynamic coupling relationship between hardware parameters (such as CPU utilization) and service response time (such as API call latency) in real time, laying the foundation for matching early warning levels later.

[0099] Figure 4 Step S303 includes:

[0100] SB1: Calculate the deviation L1 for abnormal hardware parameters;

[0101] SB2: Calculate the degradation rate L2 for service response time degradation;

[0102] SB3: Based on the preset formula S=α·L1+β·L2, where α+β=1, a dynamic score S is generated;

[0103] SB4: When S≥δ1 and δ1 is the first warning threshold, a level one warning is matched;

[0104] SB5: When δ2≤S<δ1 and δ2 is the second warning threshold, a level 2 warning is matched;

[0105] SB6: When δ3≤S<δ2 and δ3 is the third warning threshold, a level 3 warning is matched.

[0106] In this embodiment, a dynamic score S is calculated using deviation (Z-Score normalization) and degradation rate (sliding window comparison), and a three-level early warning is output by combining it with a threshold rule engine. This achieves a shift from "passive alerting" to "proactive prediction," such as issuing an early warning 30 minutes before a memory leak causes service lag, shortening the MTTD (Mean Time To Detection) to the minute level.

[0107] Step S40 includes:

[0108] S401: Based on the handling priority corresponding to the warning level, select candidate solutions that meet the following conditions from the preset historical case database:

[0109] S402: The candidate solution type must match the current warning level to a degree that reaches or exceeds a preset similarity threshold, and the historical execution success rate of the candidate solution must not be lower than a preset success rate threshold.

[0110] S403: After the filtering is completed, output the Top-N candidate solution set that meets the conditions;

[0111] S404: In the Top-N candidate solution set, for different warning levels, execute the preset differentiated handling strategy and output the candidate handling solution set;

[0112] S405: The preset differentiated processing strategy includes:

[0113] S406: When it is a Level 1 warning, match the set of emergency plans that need to be executed immediately;

[0114] S407: When it is a Level II warning, match the set of optimized solutions that require time-limited handling;

[0115] S408: When it is a Level 3 warning, match the observation period response plan.

[0116] In this embodiment, the matching degree between the candidate solution type and the current warning level, and the historical execution success rate of the candidate solution are calculated based on a preset historical case library. Only when the matching degree between the candidate solution type and the current warning level reaches or exceeds a preset similarity threshold, and the historical execution success rate of the candidate solution is not lower than a preset success rate threshold, can the condition be met and the candidate solution become a Top-N candidate solution. Combined with the warning level, a preset differentiated handling strategy is triggered: when it is a Level 1 warning, an emergency solution set that needs to be executed immediately is matched; when it is a Level 2 warning, an optimized solution set that needs to be handled within a time limit is matched; when it is a Level 3 warning, an observation period handling plan is matched.

[0117] S50 steps include:

[0118] S501: Set the business impact dimension, resource cost dimension, operational complexity dimension, and historical success dimension, and assign preset weights W1, W2, W3, and W4 to each dimension;

[0119] S502: After normalizing the scores of each candidate solution across all dimensions, based on the preset formula... (Dimensional Scoring) k ×weight W k ), calculate the overall score;

[0120] S50: Sort the comprehensive scores from high to low, and select the candidate solution with the highest comprehensive score as the optimal solution.

[0121] In this embodiment, a "multi-dimensional quantitative scoring + intelligent sorting" mechanism is used to solve the problems of inefficiency and guesswork in traditional manual decision-making. The core logic consists of three steps: Like a "balance scale," four decision-making weights are set—business impact (e.g., the number of users and order volume affected by a failure, with the highest weight), resource cost (e.g., points are deducted if additional server purchases are required), operational risk (e.g., restarting core services may cause temporary interruptions), and historical success rate (whether similar solutions have been effective in the past), and preset weights are assigned to ensure that key factors dominate the decision; the four-dimensional scores of each solution are "flattened" (e.g., 0-100 points are uniformly mapped to the 0-1 range); and a preset formula is used. (Dimensional Scoring) k ×weight W k The system generates a "health index" - a comprehensive score for each solution, and finally ranks the optimal solutions according to their comprehensive scores.

[0122] It should be understood that the sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0123] In one embodiment, a big data-based information system integrated operation and maintenance management platform system is provided, which corresponds to the big data-based information system integrated operation and maintenance management platform method in the above embodiment. The big data-based information system integrated operation and maintenance management platform control system includes:

[0124] Data Acquisition Module: Collects system operation data of a specified enterprise information system in real time, stores and preprocesses the system operation data, and outputs a system operation dataset;

[0125] Fusion module: Performs time axis calibration and fusion processing on the system operation dataset, and outputs a multi-dimensional correlated dataset;

[0126] Output module: Based on the multi-dimensional associated dataset, a dynamic association model is established to monitor abnormal situations in real time and output the early warning level;

[0127] Matching module: Based on the warning level, matches a set of candidate treatment plans;

[0128] Filtering module: Based on preset filtering rules, it filters the optimal treatment plan from the candidate treatment plan set and executes it;

[0129] Feedback module: Outputs the evaluation results of the execution effect of the optimal solution and feeds the evaluation results back to the preset big data platform.

[0130] Optional, also includes:

[0131] The first data acquisition module collects system operation data of the specified enterprise information system in real time through sensor acquisition, log file extraction, and API interface reception.

[0132] The first module includes: the system operation data includes server performance indicators and network device status data;

[0133] The second module includes: the service performance indicators include CPU utilization, memory usage, and disk space;

[0134] The third module includes: the network device status data includes network traffic, number of network connections, packet loss rate, and application runtime logs;

[0135] First output module: Based on a preset big data platform, it stores and preprocesses the system operation data and outputs the system operation dataset.

[0136] Optional, also includes:

[0137] Alignment module: Based on a preset data interpolation algorithm, the system operation data of two adjacent related nodes in the system operation dataset are aligned to eliminate the acquisition time deviation.

[0138] Extraction module: Extracts temporal and spatial features from the aligned system runtime data;

[0139] The fifth module includes: the time characteristics include the collection time, trend of change, and periodicity of system operation data;

[0140] The sixth module includes: the spatial features include the source device, device location, device type, and data association path of the system operation data;

[0141] The first fusion module: inputs the extracted temporal and spatial features into a pre-trained weighted spatiotemporal correlation matrix for fusion;

[0142] First calculation module: Based on the fusion results, perform multi-dimensional data fusion calculations, including but not limited to device operation status fusion, service response latency fusion, and business health indicator fusion;

[0143] The seventh module includes: generating a multi-dimensional related dataset by taking the results of the fusion calculation and generating it according to a predefined data structure and format. The multi-dimensional dataset includes device dimension, service dimension, business dimension, and time dimension.

[0144] Optional, also includes:

[0145] The first module establishes a dynamic correlation model between hardware performance indicators and service response quality for the multi-dimensional associated dataset.

[0146] First monitoring module: Based on the dynamic correlation model, monitor hardware parameter anomalies in real time and simultaneously detect corresponding service response time degradation;

[0147] The first matching module: Once abnormal hardware parameters and deteriorated service response time are detected simultaneously, it will match and output the corresponding warning level.

[0148] Optional, also includes:

[0149] The second calculation module calculates the deviation L1 for abnormal hardware parameter values.

[0150] The third calculation module: calculates the degradation rate L2 based on the service response time degradation;

[0151] First formula module: Based on the preset formula S=α·L1+β·L2, where α+β=1, a dynamic score S is generated;

[0152] Second matching module: When S≥δ1 and δ1 is the first warning threshold, match the first-level warning;

[0153] Third matching module: When δ2≤S<δ1 and δ2 is the second warning threshold, match the second-level warning;

[0154] Fifth matching module: When δ3≤S<δ2 and δ3 is the third warning threshold, match the third-level warning.

[0155] Optional, also includes:

[0156] First screening module: Based on the handling priority corresponding to the warning level, screen candidate solutions that meet the following conditions from a preset historical case database:

[0157] First judgment module: The candidate solution type must match the current warning level with a preset similarity threshold or higher, and the historical execution success rate of the candidate solution must not be lower than a preset success rate threshold;

[0158] Second output module: After filtering, output the Top-N candidate solution set that meets the conditions;

[0159] The third output module: In the Top-N candidate solution set, for different warning levels, execute the preset differentiated handling strategy and output the candidate handling solution set;

[0160] The eighth module includes: the preset differentiated handling strategy includes:

[0161] The sixth matching module: When it is a Level 1 alert, it matches a set of emergency solutions that need to be executed immediately;

[0162] The seventh matching module: When it is a level 2 warning, it matches a set of optimized solutions that need to be addressed within a time limit;

[0163] The eighth matching module: When it is a level 3 warning, match the handling plan for the observation period.

[0164] Optional, also includes:

[0165] Settings module: Set business impact dimension, resource cost dimension, operation complexity dimension and historical success dimension, and assign preset weights W1, W2, W3 and W4 to each dimension;

[0166] The fifth calculation module: After normalizing the scores of each candidate solution across all dimensions, it calculates the scores based on a preset formula. (Dimensional Scoring) k ×weight W k ), calculate the overall score;

[0167] Sorting module: Sort the comprehensive scores from high to low, and select the candidate solution with the highest comprehensive score as the optimal solution.

[0168] In one embodiment, a computer device is provided, which may be a server, and its internal structure diagram may be as follows: Figure 5 As shown, the computer device includes a processor, memory, network interface, and database connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and database. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The database stores system runtime datasets. The network interface communicates with external terminals via a network connection. When the computer program is executed by the processor, it implements a method for a comprehensive operation and maintenance management platform for an information system based on big data.

[0169] In one embodiment, a computer device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements a method for a comprehensive operation and maintenance management platform for an information system based on big data.

[0170] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon, wherein the computer program, when executed by a processor, is a method for a comprehensive operation and maintenance management platform for an information system based on big data.

[0171] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), Rambus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0172] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of functional units and modules is used as an example. In practical applications, the above functions can be assigned to different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above.

Claims

1. A method for a big data-based information system comprehensive operation and maintenance management platform, characterized in that, Including the following steps: The system collects system operation data from a specified enterprise information system in real time, stores and preprocesses the system operation data, and outputs a system operation dataset. The system runtime dataset is subjected to time axis calibration and fusion processing to output a multi-dimensional correlated dataset; For the system operation data of two adjacent related nodes in the system operation dataset, the system operation data is aligned by eliminating the acquisition time deviation based on the preset data interpolation algorithm; Extract temporal and spatial features from the aligned system operation data; The time characteristics include the collection time, trend of change, and periodicity of system operation data; The spatial characteristics include the source device, device location, device type, and data association path of the system operation data; The extracted temporal and spatial features are input into a pre-trained weighted spatiotemporal correlation matrix for fusion. Based on the fusion results, multi-dimensional data fusion calculations are performed, including but not limited to device operation status fusion, service response latency fusion, and business health indicator fusion. The results of the fusion computing are used to generate a multi-dimensional associated dataset according to a predefined data structure and format. The multi-dimensional associated dataset includes device dimension, service dimension, business dimension, and time dimension. A dynamic association model is established based on the multi-dimensional association dataset to monitor abnormal situations in real time and output early warning levels. For the aforementioned multi-dimensional associated dataset, a dynamic correlation model between hardware performance indicators and service response quality is established; Based on the aforementioned dynamic correlation model, abnormal hardware parameters are monitored in real time, and corresponding service response time degradation is detected simultaneously. Once both abnormal hardware parameters and deteriorating service response time are detected, the corresponding warning level will be output. For hardware parameter outliers, calculate the deviation degree ; for service response time degradation, calculating a degradation rate ; Based on preset formula ,in Generate dynamic ratings ; when , When the threshold is the first warning threshold, a level one warning is triggered. When , is the second early warning threshold, match the secondary early warning; When , is the third early warning threshold, match the third level early warning; Based on the aforementioned warning level, a set of candidate response plans will be matched; Based on preset screening rules, the optimal treatment plan is selected from the set of candidate treatment plans and then executed. Set up business impact dimension, resource cost dimension, operational complexity dimension, and historical success dimension, and assign preset weights W1, W2, W3, and W4 to each dimension; After normalizing the scores of each dimension of each candidate solution, a comprehensive score score is calculated based on a preset formula score . The comprehensive scores are sorted from high to low, and the candidate solution with the highest comprehensive score is selected as the optimal solution. Output the evaluation results of the optimal handling solution and feed them back to the preset big data platform. 2.The big data-based information system comprehensive operation and maintenance management platform method of claim 1, wherein, The steps of real-time collection of system operation data from a designated enterprise information system, storage and preprocessing of the system operation data, and outputting a system operation dataset include: The system operation data of the specified enterprise information system is collected in real time through sensor acquisition, log file extraction, and API interface reception. The system operation data includes server performance indicators and network device status data; The server performance metrics include CPU utilization, memory usage, and disk space. The network device status data includes network traffic, number of network connections, packet loss rate, and application runtime logs. Based on a pre-set big data platform, the system operation data is stored and preprocessed to output a system operation dataset. 3.The big data-based information system comprehensive operation and maintenance management platform method of claim 1, wherein, The step of matching a set of candidate response plans based on the warning level includes: Based on the handling priority corresponding to the warning level, candidate solutions that meet the following conditions are selected from a preset historical case database: The candidate solution type must match the current warning level with a preset similarity threshold, and the historical execution success rate of the candidate solution must not be lower than a preset success rate threshold. After the filtering is complete, output the Top-N candidate solutions that meet the criteria; Within the Top-N candidate solution set, for different warning levels, a pre-defined differentiated handling strategy is executed, and a candidate handling solution set is output. The preset differentiated handling strategy includes: When it is a Level 1 warning, match the set of emergency plans that need to be implemented immediately; When it is a Level II warning, match the optimized solution set that requires a time limit for handling; When it is a Level 3 warning, a corresponding observation period response plan will be implemented.

4. A big data-based information system comprehensive operation and maintenance management platform system, characterized in that, include: Data Acquisition Module: Collects system operation data of a specified enterprise information system in real time, stores and preprocesses the system operation data, and outputs a system operation dataset; Fusion module: Performs time axis calibration and fusion processing on the system operation dataset, and outputs a multi-dimensional correlated dataset; Alignment module: Based on a preset data interpolation algorithm, the system operation data of two adjacent related nodes in the system operation dataset are aligned to eliminate the acquisition time deviation. Extraction module: Extracts temporal and spatial features from the aligned system runtime data; The fifth module includes: the time characteristics include the collection time, trend of change, and periodicity of system operation data; The sixth module includes: the spatial features include the source device, device location, device type, and data association path of the system operation data; The first fusion module: inputs the extracted temporal and spatial features into a pre-trained weighted spatiotemporal correlation matrix for fusion; First calculation module: Based on the fusion results, perform multi-dimensional data fusion calculations, including but not limited to device operation status fusion, service response latency fusion, and business health indicator fusion; The seventh module includes: generating a multi-dimensional associated dataset by taking the results of the fusion calculation and generating it according to a predefined data structure and format. The multi-dimensional associated dataset includes device dimension, service dimension, business dimension, and time dimension. Output module: Based on the multi-dimensional associated dataset, a dynamic association model is established to monitor abnormal situations in real time and output the early warning level; The first module establishes a dynamic correlation model between hardware performance indicators and service response quality for the multi-dimensional associated dataset. First monitoring module: Based on the dynamic correlation model, monitor hardware parameter anomalies in real time and simultaneously detect corresponding service response time degradation; First matching module: Once abnormal hardware parameters and deteriorated service response time are detected simultaneously, the corresponding warning level is matched and output. The second calculation module calculates the deviation from abnormal hardware parameter values. ; The third calculation module assesses service response time degradation and calculates the degradation rate. ; First formula module: based on a preset formula , wherein , generating a dynamic score ; Second matching module: when , When the threshold is the first warning threshold, a level one warning is triggered. The third matching module matches the second early warning when , is the second early warning threshold. The fifth matching module: when , is the third early warning threshold, match the third early warning; Matching module: Based on the warning level, matches a set of candidate treatment plans; Filtering module: Based on preset filtering rules, it filters the optimal treatment plan from the candidate treatment plan set and executes it; Settings module: Set business impact dimension, resource cost dimension, operation complexity dimension and historical success dimension, and assign preset weights W1, W2, W3 and W4 to each dimension; The fifth calculation module: after normalizing the scores of each dimension of each candidate scheme, the comprehensive score score is calculated based on a preset formula score . Sorting module: Sort the comprehensive scores from high to low, and select the candidate solution with the highest comprehensive score as the optimal solution; Feedback module: Outputs the evaluation results of the execution effect of the optimal solution and feeds the evaluation results back to the preset big data platform.

5. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the method for a comprehensive operation and maintenance management platform for an information system based on big data as described in any one of claims 1-3.

6. A computer-readable storage medium storing a computer program, the computer program comprising instructions that, when executed by a computer, cause the computer to perform the method of any one of claims 1 to 5. When the computer program is executed by the processor, it implements the steps of the method for a comprehensive operation and maintenance management platform for an information system based on big data as described in any one of claims 1-3.

Images