Big data-based database maintenance analysis method and system
By using big data analytics methods and systems, database object dependency graphs and performance models are constructed, solving the problem of insufficient human experience in traditional database maintenance and enabling efficient and automated database maintenance decisions and resource allocation.
Patent Information
- Application Number
- CN202610775457.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-06-01
- Publication Date
- 2026-08-25
AI Technical Summary
Traditional database maintenance and analysis methods rely on manual experience and lack systematic data collection and intelligent analysis. They cannot accurately assess the level of fault risk and the impact of related objects, resulting in inefficient maintenance decisions, unreasonable resource allocation, and delayed maintenance of critical objects.
Using a big data-based approach, through data acquisition, preprocessing, graph neural networks, and temporal convolutional networks, a database object dependency graph and performance degradation model are constructed. The fault risk index, correlation impact index, and performance degradation index are calculated to generate maintenance work orders and perform automated maintenance.
It enables the scientific classification of database risk levels, accurately identifies high-risk databases, improves the efficiency of maintenance resource allocation, accurately quantifies the scope of impact of objects, and optimizes the efficiency of maintenance decision-making and execution.
Smart Images
Figure CN122633654A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of database maintenance technology, and in particular to a database maintenance and analysis method and system based on big data. Background Technology
[0002] With the rapid development of the Internet, the Internet of Things, and enterprise informatization, the volume of data has exploded, giving rise to big data technology, which is now widely used across various industries. Databases, as the core infrastructure for data storage and management, are constantly increasing in scale and complexity, undertaking critical tasks such as real-time reading and writing, transaction processing, and analysis and querying of massive amounts of business data. In a big data environment, database systems typically contain hundreds or thousands of database objects, such as tables, views, indexes, and stored procedures. These objects have complex dependencies, such as foreign key relationships, query references, and data flow dependencies. Performance degradation or failure of any one object can propagate rapidly through the dependency chain, triggering a chain reaction that can lead to performance crashes or even service interruptions for the entire database system.
[0003] However, traditional database maintenance and analysis methods rely on manual experience for regular inspections and troubleshooting. They lack systematic collection and intelligent analysis of massive amounts of operational data, making it impossible to accurately assess the risk level of database failures. They also struggle to quantify the strength of the correlation between various database objects and the trend of performance degradation. This leads to inefficient maintenance decisions, unreasonable allocation of maintenance resources, and delayed maintenance of critical objects, ultimately affecting the stability and availability of the database system.
[0004] To address the aforementioned technical deficiencies, a solution is proposed. Summary of the Invention
[0005] The purpose of this invention is to address the problems of traditional database maintenance and analysis methods that rely on manual experience for regular inspections and troubleshooting, lack systematic collection and intelligent analysis of massive amounts of operational data, cannot accurately assess the fault risk level of the database, and are difficult to quantify the correlation and influence strength and performance degradation trend between various database objects, resulting in low maintenance decision efficiency and problems such as unreasonable allocation of maintenance resources and delayed maintenance of key objects.
[0006] To achieve the above objectives, the present invention adopts the following technical solution: a database maintenance and analysis method based on big data, comprising the following steps: Step 1: Through the database operation platform, collect the database operation data, database object association data and performance data, and preprocess the collected data. At the same time, obtain the database operation data, perform analysis and calculation, and obtain the database failure risk index. Divide the database operation risk level into high-risk databases and low-risk databases. Step 2: When the database is at a high risk level, construct a database object dependency graph, obtain related data, perform analysis and calculation, and derive the association impact index of the database objects. Step 3: Construct a performance degradation model for database objects, input real-time database object performance data into the performance degradation model, and output the performance degradation index of the database objects. Step 4: Obtain the correlation impact index and performance degradation index, and analyze and calculate them in conjunction with the fault risk index to obtain the priority maintenance index for each database object. Then, generate maintenance work orders according to the value and carry out corresponding maintenance measures for each database object.
[0007] Furthermore, the preprocessing of the collected data involves standardizing the collected runtime data, database object association data, and performance data using the Min-Max standardization method. ,in, These are standardized data values. For the first The numerical value of the data to be processed The minimum value of similar data. It represents the maximum value of the same type of data.
[0008] Furthermore, the calculation process for the database failure risk index is as follows: S11. Obtain the database's operational data and perform analysis and calculations. The operational data includes CPU utilization, memory usage, and disk I / O latency data. S12. Calculate the database failure risk index according to the following formula. : in, CPU utilization of the database server This refers to the memory usage of the database server. For the disk I / O latency of the database server, The preset safe threshold for CPU utilization. This is a preset safe threshold for memory usage. This is the preset upper limit for disk I / O latency. The preset CPU weight coefficient, The preset memory weighting coefficients, The preset delay weighting coefficient, and The database failure risk index is used to reflect the risk of database failures during operation; S13. Obtain the preset fault risk threshold. Database failure risk index Comparative analysis, when If the risk of database failure during operation is low, the database is classified as a low-risk database. S14, when If the risk level is high, it indicates that the database is at high risk of failure during operation. The database is classified as a high-risk database, and the maintenance of the high-risk database will continue to be analyzed.
[0009] Furthermore, the calculation process for the association influence index of database objects is as follows: S21. Identify all objects in the database and their dependencies. Represent objects as nodes and dependencies as directed edges. Construct a directed acyclic graph, where each node v carries its associated attribute feature vector. Each edge (u,v) is assigned a weight. , indicating the strength of dependency; S22. Using a graph neural network model, the features of each node in the dependency graph are aggregated along the edges in multiple rounds to capture the dependency propagation effect. For node v, the impact after aggregation is characterized. The correlation influence index is calculated by weighting and updating neighborhood information and combining it with the node's own characteristics. ; S23. Calculate the association influence index of database objects according to the following formula. : in, The association influence index of database object v. The set of all nodes that depend on v. For edge weights, Let u be the out-degree of node u. Let be the shortest path length from node u to v. This represents the number of times node v itself is referenced. The Euclidean norm is used to represent the aggregated influence, indicating the overall influence strength after dependency propagation. The association influence index of database objects is used to reflect the association influence strength of database objects in the dependency graph.
[0010] Furthermore, the process of constructing a performance degradation model for database objects is as follows: S31. Select the core performance characteristics of database objects, including query response time, lock wait time, transaction throughput, index hit rate, and disk space growth rate. Use a temporal convolutional network to build a performance degradation model. Train the model with historical performance data to capture the degradation trend of performance indicators over time. The input of the model is a sequence of performance characteristics in a continuous time window, and the output is a representation of the degradation rate of each indicator. S32. Collect real-time performance data of database objects at the current moment, including the instantaneous values and recent change rates of each feature index. Normalize the real-time data according to the standardization rules during model training to unify the data scale. Input the preprocessed real-time performance feature sequence into the trained temporal convolutional network model. After multiple layers of causal convolution and pooling operations, extract the deep temporal features of performance degradation and generate the performance state representation vector at the current moment. S33. Calculate the performance degradation index of a database object using the following formula. : in, This represents the actual response time of the database query at the current moment. The average lock wait time for the current database transaction. This represents the number of transactions successfully executed per second. This is the baseline response time during normal operation. The base lock wait time As a benchmark transaction throughput, The preset response weighting coefficients, This is a preset weighting coefficient for lock waiting time.
[0011] Furthermore, the calculation process for the priority maintenance index of database objects is as follows: S41. Obtain the correlation impact index and performance degradation index, and analyze and calculate them in conjunction with the failure risk index; S42. Calculate the priority maintenance index of database objects according to the following formula: in, The priority maintenance index for database object v. The failure risk index for high-risk databases. The preset fault risk threshold, The association influence index of database object v. D is the maximum allowed performance degradation index for a predefined database object.
[0012] Furthermore, the process of generating a maintenance work order is as follows: S51. Sort each database object from highest to lowest according to the value of its priority maintenance index, and generate a database object maintenance priority list. S52. Based on maintenance priority, filter out the database objects that need to be processed first, and automatically extract the basic information, dependencies, performance degradation details and historical maintenance records of the objects. S53. Based on the database administrator's preset maintenance resource configuration, generate a corresponding maintenance work order for each object to be maintained. The work order includes the maintenance object identifier, suggested maintenance type, estimated maintenance duration, required resource list and operation precautions, and push the work order to the relevant maintenance personnel in real time. S54. After a maintenance work order is generated, the work order status is automatically recorded, the maintenance process is tracked throughout, and the performance data of the maintained object and the dependency graph are updated after the maintenance is completed.
[0013] The present invention also provides a database maintenance and analysis system based on big data, including a data acquisition unit, a risk assessment unit, a correlation analysis unit, a performance monitoring unit, and a maintenance decision unit; The data acquisition unit is used to collect database operation data, database object association data, and performance data through the database operation platform, and to preprocess the collected data. The risk assessment unit is used to acquire the database's operational data, perform analysis and calculations, derive the database's failure risk index, and classify the database's operational risk level into high-risk and low-risk databases. The association analysis unit is used to construct a database object dependency graph, obtain association data, perform analysis and calculation, and derive the association influence index of database objects when the database is at a high risk level. The performance monitoring unit is used to build a performance degradation model for database objects, input real-time performance data of database objects into the performance degradation model, and output the performance degradation index of database objects. The maintenance decision unit is used to obtain the correlation impact index and performance degradation index, and combine them with the failure risk index for analysis and calculation to obtain the priority maintenance index of each database object. Then, it generates maintenance work orders according to the value and performs corresponding maintenance measures for each database object.
[0014] In summary, due to the adoption of the above technical solution, the beneficial effects of the present invention are: This database maintenance analysis method and system based on big data quantifies and calculates a fault risk index through multi-dimensional operational data, enabling the scientific classification of database risk levels. It accurately identifies high-risk databases and triggers in-depth analysis, preventing low-risk databases from excessively consuming maintenance resources and significantly improving the efficiency of maintenance resource allocation. Simultaneously, it constructs a database object dependency graph and uses graph neural networks to calculate the correlation impact index, accurately quantifying the impact range of each object within the entire system, providing a reliable basis for subsequent maintenance decisions. Furthermore, it utilizes a temporal convolutional network to construct a performance degradation model, achieving accurate prediction of database object performance degradation trends. In addition, it calculates a priority maintenance index by comprehensively considering the correlation impact index, performance degradation index, and fault risk index, enabling scientific ranking of maintenance priorities and ensuring that the most impactful objects receive priority maintenance. Finally, it automatically generates maintenance work orders containing complete information and tracks the entire process, significantly improving the automation level and execution efficiency of maintenance work. Attached Figure Description
[0015] Figure 1 A schematic diagram of the method flow of the present invention is shown; Figure 2 A schematic diagram of the system flow of the present invention is shown. Detailed Implementation
[0016] 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.
[0017] Example 1: like Figure 1 As shown, the database maintenance and analysis method based on big data includes the following steps: Step 1: Through the database operation platform, collect the database operation data, database object association data and performance data, and preprocess the collected data. At the same time, obtain the database operation data, perform analysis and calculation, and obtain the database failure risk index. Divide the database operation risk level into high-risk databases and low-risk databases. The preprocessing of the collected data involves standardizing the collected runtime data, database object association data, and performance data using the Min-Max standardization method. ,in, These are standardized data values. For the first The numerical value of the data to be processed The minimum value of similar data. It represents the maximum value of the same type of data.
[0018] The calculation process for the database failure risk index is as follows: S11. Obtain the database's operational data and perform analysis and calculations. The operational data includes CPU utilization, memory usage, and disk I / O latency data. S12. Calculate the database failure risk index according to the following formula. : in, CPU utilization of the database server This refers to the memory usage of the database server. For the disk I / O latency of the database server, This is a preset safe threshold for CPU utilization (e.g., 0.8, which represents 80%). This is a preset safe threshold for memory usage (e.g., 0.9, which represents 90%). This is the preset upper limit for disk I / O latency (e.g., 50ms). The preset CPU weight coefficient, The preset memory weighting coefficients, The preset delay weighting coefficient, and The database failure risk index is used to reflect the risk of database failure during operation. The higher the failure risk index value, the greater the risk of database failure during operation, and the lower the failure risk index value, the lower the risk of database failure during operation. S13. Obtain the preset fault risk threshold. Database failure risk index Comparative analysis, when If the risk of database failure during operation is low, the database is classified as a low-risk database. S14, when If the risk level is high, it indicates that the database is at high risk of failure during operation. The database is classified as a high-risk database, and the maintenance of the high-risk database will continue to be analyzed.
[0019] Step 2: When the database is at a high risk level, construct a database object dependency graph, obtain related data, perform analysis and calculation, and derive the association impact index of the database objects. The calculation process for the association impact index of database objects is as follows: S21. Identify all objects in the database (such as tables, views, indexes, and stored procedures) and their dependencies (such as foreign key relationships, query references, and data flow dependencies). Represent objects as nodes and directions of dependencies as directed edges. Construct a directed acyclic graph, with each node v carrying its associated attribute feature vector. Parameters such as the number of times an edge is referenced, the amount of data, and the update frequency; each edge (u,v) is assigned a weight. This indicates the strength of the dependency (such as the number of calls or data flow). S22. Using a graph neural network model, the features of each node in the dependency graph are aggregated along the edges in multiple rounds to capture the dependency propagation effect. For node v, the impact after aggregation is characterized. The correlation influence index is calculated by weighting and updating neighborhood information and combining it with the node's own characteristics. ; S23. Calculate the association influence index of database objects according to the following formula. : in, The association influence index of database object v. This is the set of all nodes that depend on v (i.e., nodes that point to v). For edge weights, Let u be the out-degree of node u (representing its influence breadth). This is the shortest path length from node u to v (used to mitigate the effects of indirect dependencies). This represents the number of times node v itself is referenced. Let be the Euclidean norm of the aggregation characterization, representing the strength of the combined influence after dependency propagation, where It is by The association influence index of a database object, obtained through graph neural network aggregation, reflects the strength of the association influence of the database object in the dependency graph. The larger the value of the association influence index, the stronger the association influence of the database object in the dependency graph, and the higher the maintenance priority. The smaller the value of the association influence index, the weaker the association influence of the database object in the dependency graph, and the lower the maintenance priority.
[0020] Step 3: Construct a performance degradation model for database objects, input real-time database object performance data into the performance degradation model, and output the performance degradation index of the database objects. The process of constructing a performance degradation model for database objects is as follows: S31. Select the core performance characteristics of database objects, including query response time, lock wait time, transaction throughput, index hit rate, and disk space growth rate. Use a temporal convolutional network (TCN) to build a performance degradation model. Train the model with historical performance data to capture the degradation trend of performance indicators over time. The input of the model is a sequence of performance characteristics in a continuous time window, and the output is a representation of the degradation rate of each indicator. Use the causal convolutional structure of the temporal convolutional network to ensure that the prediction depends only on historical data, avoid future information leakage, and lay the foundation for subsequent real-time evaluation. S32. Collect real-time performance data of database objects at the current moment, including the instantaneous values and recent change rates of each feature index. Normalize the real-time data according to the standardization rules during model training to unify the data scale. Input the preprocessed real-time performance feature sequence into the trained temporal convolutional network model. After multiple layers of causal convolution and pooling operations, extract the deep temporal features of performance degradation and generate the performance status representation vector at the current moment. This vector integrates the current level and historical trend information of each index. S33. Calculate the performance degradation index of a database object using the following formula. : in, This represents the actual response time of the database query at the current moment. The average lock wait time for the current database transaction. This represents the number of transactions successfully executed per second. This is the baseline response time during normal operation (historical best or design target value). The baseline lock wait time (expected value under normal concurrency). The baseline transaction throughput (the peak processing capacity of the system design). The preset response weighting coefficients, This is a preset weighting coefficient for lock waiting time.
[0021] Step 4: Obtain the correlation impact index and performance degradation index, and analyze and calculate them in conjunction with the fault risk index to obtain the priority maintenance index of each database object. Then, generate maintenance work orders according to the value and carry out corresponding maintenance measures for each database object. The calculation process for the priority maintenance index of database objects is as follows: S41. Obtain the correlation impact index and performance degradation index, and analyze and calculate them in conjunction with the failure risk index; S42. Calculate the priority maintenance index of database objects according to the following formula: in, The priority maintenance index for database object v. The failure risk index for high-risk databases. The preset fault risk threshold, The association influence index of database object v. D is the maximum allowed performance degradation index for a predefined database object.
[0022] The process of generating a maintenance work order is as follows: S51. Sort each database object from highest to lowest according to the value of its priority maintenance index, and generate a database object maintenance priority list. S52. Based on maintenance priority, filter out the database objects that need to be processed first, and automatically extract the basic information, dependencies, performance degradation details and historical maintenance records of the objects. S53. Based on the database administrator's preset maintenance resource configuration, generate a corresponding maintenance work order for each object to be maintained. The work order includes the maintenance object identifier, suggested maintenance type (such as index rebuilding, tablespace expansion, stored procedure optimization), estimated maintenance duration, required resource list and operation precautions, and push the work order to the relevant maintenance personnel in real time. S54. After a maintenance work order is generated, the work order status is automatically recorded, the maintenance process is tracked throughout, and the performance data of the maintained object and the dependency graph are updated after the maintenance is completed.
[0023] This invention quantifies and calculates a fault risk index using multi-dimensional operational data, enabling the scientific classification of database risk levels. It accurately identifies high-risk databases and triggers in-depth analysis, preventing low-risk databases from excessively consuming maintenance resources and significantly improving the efficiency of maintenance resource allocation. Simultaneously, it constructs a database object dependency graph and uses graph neural networks to calculate the correlation impact index, accurately quantifying the impact range of each object within the entire system, providing a reliable basis for subsequent maintenance decisions. Furthermore, it utilizes a temporal convolutional network to construct a performance degradation model, achieving accurate prediction of database object performance degradation trends. In addition, it calculates a priority maintenance index by comprehensively considering the correlation impact index, performance degradation index, and fault risk index, achieving a scientific ranking of maintenance priorities and ensuring that the most impactful objects receive priority maintenance. Finally, it automatically generates maintenance work orders containing complete information and tracks the entire process, significantly improving the automation level and execution efficiency of maintenance work.
[0024] Example 2: like Figure 2 As shown, the database maintenance and analysis system based on big data is applied to database maintenance and analysis methods based on big data, including a data acquisition unit, a risk assessment unit, a correlation analysis unit, a performance monitoring unit, and a maintenance decision-making unit. The data acquisition unit is used to collect database operation data, database object association data, and performance data through the database operation platform, and to preprocess the collected data. The risk assessment unit is used to acquire the database's operational data, perform analysis and calculations, derive the database's failure risk index, and classify the database's operational risk level into high-risk and low-risk databases. The association analysis unit is used to construct a database object dependency graph, obtain association data, perform analysis and calculation, and derive the association influence index of database objects when the database is at a high risk level. The performance monitoring unit is used to build a performance degradation model for database objects, input real-time performance data of database objects into the performance degradation model, and output the performance degradation index of database objects. The maintenance decision unit is used to obtain the correlation impact index and performance degradation index, and combine them with the failure risk index for analysis and calculation to obtain the priority maintenance index of each database object. Then, it generates maintenance work orders according to the value and performs corresponding maintenance measures for each database object.
[0025] The size of the interval and threshold is set to facilitate comparison. The size of the threshold depends on the amount of sample data and the number of bases set by those skilled in the art for each set of sample data; as long as it does not affect the ratio between the parameter and the quantized value.
[0026] The above formulas are all dimensionless calculations. The formulas are derived from software simulations based on a large amount of collected data to obtain the most recent real-world results. The preset parameters in the formulas are set by those skilled in the art according to the actual situation. The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.
Claims
1. A database maintenance and analysis method based on big data, characterized in that, Includes the following steps: Step 1: Through the database operation platform, collect the database operation data, database object association data and performance data, and preprocess the collected data. At the same time, obtain the database operation data, perform analysis and calculation, and obtain the database failure risk index. Divide the database operation risk level into high-risk databases and low-risk databases. Step 2: When the database is at a high-risk level, construct a database object dependency graph, obtain related data, perform analysis and calculation, and derive the association impact index of the database objects. Step 3: Construct a performance degradation model for database objects, input real-time database object performance data into the performance degradation model, and output the performance degradation index of the database objects. Step 4: Obtain the correlation impact index and performance degradation index, and analyze and calculate them in conjunction with the fault risk index to obtain the priority maintenance index for each database object. Then, generate maintenance work orders according to the value and carry out corresponding maintenance measures for each database object.
2. The database maintenance and analysis method based on big data according to claim 1, characterized in that, The preprocessing of the collected data involves standardizing the collected runtime data, database object association data, and performance data using the Min-Max standardization method. ,in, These are standardized data values. For the first The numerical value of the data to be processed The minimum value of similar data. It represents the maximum value of the same type of data.
3. The database maintenance and analysis method based on big data according to claim 1, characterized in that, The calculation process for the database failure risk index is as follows: S11. Obtain the database's operating data and perform analysis and calculations. The operating data includes CPU utilization, memory usage, and disk I / O latency data. S12. Calculate the database failure risk index according to the following formula. : in, CPU utilization of the database server This refers to the memory usage of the database server. For the disk I / O latency of the database server, The preset safe threshold for CPU utilization. This is a preset safe threshold for memory usage. This is the preset upper limit for disk I / O latency. The preset CPU weight coefficient, The preset memory weighting coefficients, The preset delay weighting coefficient, and The database failure risk index is used to reflect the risk of database failures during operation; S13. Obtain the preset fault risk threshold. Database failure risk index Comparative analysis, when If the risk of database failure during operation is low, the database is classified as a low-risk database. S14, when If the risk level is high, it indicates that the database is at high risk of failure during operation. The database is classified as a high-risk database, and the maintenance of the high-risk database will continue to be analyzed.
4. The database maintenance and analysis method based on big data according to claim 1, characterized in that, The calculation process for the association impact index of database objects is as follows: S21. Identify all objects in the database and their dependencies. Represent objects as nodes and dependencies as directed edges. Construct a directed acyclic graph, where each node v carries its associated attribute feature vector. Each edge (u,v) is assigned a weight. , indicating the strength of dependency; S22. Using a graph neural network model, the features of each node in the dependency graph are aggregated along the edges in multiple rounds to capture the dependency propagation effect. For node v, the impact after aggregation is characterized. The correlation influence index is calculated by weighting and updating neighborhood information and combining it with the node's own characteristics. ; S23. Calculate the association influence index of database objects according to the following formula. : in, The association influence index of database object v. The set of all nodes that depend on v. For edge weights, Let be the out-degree of node u. Let be the shortest path length from node u to v. This represents the number of times node v itself is referenced. The Euclidean norm is used to represent the aggregated influence, indicating the overall influence strength after dependency propagation. The association influence index of database objects is used to reflect the association influence strength of database objects in the dependency graph.
5. The database maintenance and analysis method based on big data according to claim 1, characterized in that, The process of constructing a performance degradation model for database objects is as follows: S31. Select the core performance characteristics of database objects, including query response time, lock wait time, transaction throughput, index hit rate, and disk space growth rate. Use a temporal convolutional network to build a performance degradation model. Train the model with historical performance data to capture the degradation trend of performance indicators over time. The input of the model is a sequence of performance characteristics in a continuous time window, and the output is a representation of the degradation rate of each indicator. S32. Collect real-time performance data of database objects at the current moment, including the instantaneous values and recent change rates of each feature index. Normalize the real-time data according to the standardization rules during model training to unify the data scale. Input the preprocessed real-time performance feature sequence into the trained temporal convolutional network model. After multiple layers of causal convolution and pooling operations, extract the deep temporal features of performance degradation and generate the performance state representation vector at the current moment. S33. Calculate the performance degradation index of a database object using the following formula. : in, This represents the actual response time of the database query at the current moment. The average lock wait time for the current database transaction. This represents the number of transactions successfully executed per second. This is the baseline response time during normal operation. The base lock wait time As a benchmark transaction throughput, The preset response weighting coefficients, This is a preset weighting coefficient for lock waiting time.
6. The database maintenance and analysis method based on big data according to claim 1, characterized in that, The calculation process for the priority maintenance index of database objects is as follows: S41. Obtain the correlation impact index and performance degradation index, and analyze and calculate them in conjunction with the failure risk index; S42. Calculate the priority maintenance index of database objects according to the following formula: in, The priority maintenance index for database object v. The failure risk index for high-risk databases. The preset fault risk threshold, The association influence index of database object v. D is the maximum allowed performance degradation index for a predefined database object.
7. The database maintenance and analysis method based on big data according to claim 1, characterized in that, The process of generating a maintenance work order is as follows: S51. Sort each database object from highest to lowest according to its priority maintenance index, and generate a database object maintenance priority list. S52. Based on maintenance priority, filter out the database objects that need to be processed first, and automatically extract the basic information, dependencies, performance degradation details and historical maintenance records of the objects; S53. Based on the database administrator's preset maintenance resource configuration, generate a corresponding maintenance work order for each object to be maintained. The work order includes the maintenance object identifier, suggested maintenance type, estimated maintenance duration, required resource list and operation precautions, and push the work order to the relevant maintenance personnel in real time. S54. After a maintenance work order is generated, the work order status is automatically recorded, the maintenance process is tracked throughout, and the performance data of the maintained object and the dependency graph are updated after the maintenance is completed.
8. A database maintenance and analysis system based on big data, applied to the database maintenance and analysis method based on big data as described in any one of claims 1-7, characterized in that, It includes a data acquisition unit, a risk assessment unit, a correlation analysis unit, a performance monitoring unit, and a maintenance decision-making unit; The data acquisition unit is used to collect database operation data, database object association data, and performance data through the database operation platform, and to preprocess the collected data. The risk assessment unit is used to acquire the database's operational data, perform analysis and calculations, derive the database's failure risk index, and classify the database's operational risk level into high-risk and low-risk databases. The association analysis unit is used to construct a database object dependency graph, obtain association data, perform analysis and calculation, and derive the association influence index of database objects when the database is at a high risk level. The performance monitoring unit is used to build a performance degradation model for database objects, input real-time performance data of database objects into the performance degradation model, and output the performance degradation index of database objects. The maintenance decision unit is used to obtain the correlation impact index and performance degradation index, and combine them with the failure risk index for analysis and calculation to obtain the priority maintenance index of each database object. Then, it generates maintenance work orders according to the value and performs corresponding maintenance measures for each database object.