A differentiated configuration management method and device for distributed application services

Abstract

The application discloses a kind of distributed application service's differentiated configuration management method and device, it is related to data management technical field;Method includes, obtaining multiple identification data in application service, extracting feature vector, feature vector is input into the data evaluation model of pre-set and obtains data configuration score;According to configuration score threshold, identification data is divided into different levels, and different configuration management modes are adopted to identification data of different levels;Analysis is carried out through first level data, and according to preset determination condition, first level data is divided into variable data and immutable data;Through variable data of first level data, second level data, the level coefficient of each identification data is obtained by dynamic adjustment analysis calculation;According to level coefficient, data grade is reclassified, and corresponding configuration management mode is carried out to the data after classification.This scheme can reasonably allocate management resources by dynamic adjustment analysis, avoid resource waste, improve the overall efficiency of system.

Description

Technical Field

[0001] This invention relates to the field of data management technology, specifically to a method and apparatus for differentiated configuration management of distributed application services. Background Technology

[0002] Configuration management in current data management technologies is undergoing a transformation to address the increasingly complex IT architectures and the need for automated operations and maintenance. Traditional centralized CMDBs face issues such as low data accuracy and delayed updates, while emerging graph database technologies, through node and relationship modeling, significantly improve the visualization and management efficiency of configuration items and their dynamic relationships. Simultaneously, automated configuration management tools, combined with cloud-native technologies, support rapid deployment and consistent maintenance of large-scale infrastructure, reducing errors from manual intervention.

[0003] Existing technology (publication number: CN113852968A) provides a core network data configuration management method and system, including: obtaining a set of configuration information of a network element to be configured through Operation and Maintenance Management (OAM); obtaining a configuration file generated by the network element to be configured based on the set of configuration information; displaying the configuration file through a preset format file; modifying at least one configuration information in the set of configuration information to generate a modified configuration file; distributing the modified configuration file to the network element to be configured through OAM for configuration modification by the network element to be configured; and obtaining the differences generated after differential comparison of the network element to be configured. The configuration data management method for core network control plane network elements in 5G networks proposed in this invention achieves unified management of configuration files and modification processes by displaying configuration files on the OMC network management system and by modifying the configuration and distributing the modified configuration data to the local network element through the OMC, triggering different service processes.

[0004] However, in distributed application services, traditional static resource allocation methods often adopt a unified configuration management strategy for different types of identification data, which may cause important data to be delayed due to insufficient resources, while secondary data consumes too many resources and is wasted. Summary of the Invention

[0005] The purpose of this invention is to solve the problem that traditional static resource allocation methods often adopt a uniform configuration management strategy for different types of identification data, which may lead to response delays for important data due to insufficient resources, while secondary data consumes too many resources and causes waste; and to propose a differentiated configuration management method and device for distributed application services.

[0006] The objective of this invention can be achieved through the following technical solutions:

[0007] First, a differentiated configuration management method for distributed application services is proposed, the method including:

[0008] Obtain multiple identification data from the application service, extract feature parameters based on the identification data, construct feature vectors based on the feature parameters, and input the feature vectors into a preset data evaluation model to obtain a data configuration score;

[0009] The identification data is divided into different levels based on the configuration scoring threshold, and different configuration management methods are adopted for the identification data of different levels; the identification data levels are divided into first-level data and second-level data.

[0010] The data at the first level is analyzed and divided into variable data and immutable data according to preset judgment conditions.

[0011] The level coefficients of each identifier data are obtained by dynamically adjusting and analyzing the variable data of the first level data and the second level data; the data levels are reclassified according to the level coefficients, and the corresponding configuration management methods are applied to the classified data.

[0012] Optionally, the identification data can be divided into different levels based on the configuration scoring threshold, and different configuration management methods can be adopted for the identification data at different levels; the identification data levels are divided into first-level data and second-level data, including:

[0013] The configuration score of the identification data is compared with the preset configuration score threshold;

[0014] If the configuration score is less than the configuration score threshold, the identification data will be classified as the first level.

[0015] If the configuration score is greater than the configuration score threshold, the identification data will be classified as the second level.

[0016] When the identification data level is Level 1, manual configuration management is used;

[0017] When the identification data level is the second level, a script-based configuration management method is used.

[0018] Optionally, the first-level data can be divided into variable and immutable data according to preset judgment criteria, including:

[0019] For the identification data corresponding to the first level, determine whether its level can be changed according to the preset judgment conditions. If not, the corresponding identification data is recorded as immutable data, and the original configuration management method is maintained; if yes, the corresponding data is recorded as variable data, and the configuration management method of variable data is changed.

[0020] Optionally, the level coefficients for each identifier data can be obtained through dynamic adjustment analysis and calculation using the variable data of the first level and the second level data, including:

[0021] Using the variable data of the first-level data and the second-level data, obtain the number of detections of the identifier data within a preset period and a preset time period, and construct a detection sequence; sort and map the elements in the detection sequence to obtain the order column Rj; construct the rank Rt of the standard time series;

[0022] The Spearman method is used to measure the monotonic correlation θ between ordered columns Rj and Rt. The formula is as follows:

[0023] ;

[0024] Where θ represents monotonic correlation; the order column R j (i) and the rank R of the time series t (i) represent R respectively j and R t The rank of the i-th sample value; n represents the total number of samples;

[0025] The formula for calculating the level coefficient is:

[0026] ;

[0027] Where JB represents the level coefficient. Indicates the optimization factor at the level.

[0028] Optionally, the data levels can be reclassified based on level coefficients, and the corresponding configuration management methods for the reclassified data can be implemented, including:

[0029] Compare the level coefficient of the identified data with the preset upper limit threshold of the level coefficient;

[0030] If the level coefficient of the identification data is less than the preset lower limit threshold of the level coefficient, the original level of the identification data will be reduced by one level, the new level of the identification data after the reduction will be set, and the corresponding configuration management method will be performed according to the new level.

[0031] If the level coefficient of the identification data is greater than the preset lower limit threshold and less than the preset upper limit threshold, the level of the identification data remains unchanged, and the corresponding configuration management method is performed according to the original level.

[0032] If the level coefficient of the identification data is greater than the preset upper limit threshold of the level coefficient, the original level of the identification data will be raised by one level, the new level of the identification data after the level has been lowered will be set, and the corresponding configuration management method will be performed according to the new level.

[0033] Secondly, a differentiated configuration management device for distributed application services is proposed, the device comprising:

[0034] Data evaluation module: acquires multiple identification data from the application service, extracts feature parameters based on the identification data, constructs feature vectors based on the feature parameters, and inputs the feature vectors into a preset data evaluation model to obtain a data configuration score;

[0035] Level Classification Module: Based on the configured scoring threshold, the identification data is divided into different levels, and different configuration management methods are adopted for the identification data of different levels; the identification data levels are divided into first-level data and second-level data; the first-level data is analyzed and divided into variable data and immutable data according to preset judgment conditions;

[0036] Dynamic adjustment module: It calculates the level coefficient of each identifier data by dynamically adjusting the variable data of the first level data and the second level data; it reclassifies the data levels according to the level coefficients and performs corresponding configuration management on the classified data.

[0037] Optionally, the level division module includes: Level division module:

[0038] The level classification module is used to compare the configuration score of the identification data with the preset configuration score threshold.

[0039] If the configuration score is less than the configuration score threshold, the identification data will be classified as the first level.

[0040] If the configuration score is greater than the configuration score threshold, the identification data will be classified as the second level.

[0041] When the identification data level is Level 1, manual configuration management is used;

[0042] When the identification data level is the second level, a script-based configuration management method is used.

[0043] Optionally, the level classification module includes a determination module:

[0044] The determination module is used to determine whether the level of the identification data corresponding to the first level can be changed according to preset determination conditions. If not, the corresponding identification data is recorded as immutable data, and the original configuration management method is maintained; if yes, the corresponding data is recorded as variable data, and the configuration management method of the variable data is changed.

[0045] Optional, a dynamic adjustment module, including a detection sequence module and a level coefficient module:

[0046] The detection sequence module is used to obtain the number of times the identifier data is detected within a preset period and a preset time period using the variable data of the first-level data and the second-level data, and to construct a detection sequence; and to sort and map the elements in the detection sequence to obtain an order column R.j Construct the rank R of a standard time series t ;

[0047] The level coefficient module is used by the Spearman method to measure the order of the column R. j and R t The monotonic correlation θ is calculated using the following formula:

[0048] ;

[0049] Where θ represents monotonic correlation; the order column R j (i) and the rank R of the time series t (i) represent R respectively j and R t The rank of the i-th sample value; n represents the total number of samples;

[0050] The formula for calculating the level coefficient is:

[0051] ;

[0052] Where JB represents the level coefficient. Indicates the optimization factor at the level.

[0053] Optional, dynamic adjustment modules, including a level reset module:

[0054] The reset level module is used to compare the level coefficient of the identification data with the preset level coefficient upper limit threshold.

[0055] If the level coefficient of the identification data is less than the preset lower limit threshold of the level coefficient, the original level of the identification data will be reduced by one level, the new level of the identification data after the reduction will be set, and the corresponding configuration management method will be performed according to the new level.

[0056] If the level coefficient of the identification data is greater than the preset lower limit threshold and less than the preset upper limit threshold, the level of the identification data remains unchanged, and the corresponding configuration management method is performed according to the original level.

[0057] If the level coefficient of the identification data is greater than the preset upper limit threshold of the level coefficient, the original level of the identification data will be raised by one level, the new level of the identification data after the level has been lowered will be set, and the corresponding configuration management method will be performed according to the new level.

[0058] The beneficial effects of this invention are:

[0059] This invention acquires multiple identifier data from application services, extracts feature parameters from the identifier data, constructs feature vectors based on these parameters, and inputs these feature vectors into a preset data evaluation model to obtain a data configuration score. Based on the configuration score threshold, the identifier data is divided into different levels, and different configuration management methods are applied to identifier data at different levels. Identifier data levels are divided into first-level data and second-level data. First-level data is analyzed, and based on preset judgment conditions, it is divided into variable and immutable data. Dynamic adjustment analysis is performed on the variable data of the first-level data and the second-level data to calculate the level coefficient for each identifier data. Based on the level coefficient, the data levels are reclassified, and corresponding configuration management methods are applied to the reclassified data. By dividing the identifier data of the target company's application services into different levels and determining different configuration management methods based on different levels, resource waste and application service response latency are reduced. Through dynamic adjustment analysis, management resources can be rationally allocated, resource waste is avoided, and the overall efficiency of the system is improved. Attached Figure Description

[0060] Figure 1 A flowchart illustrating a differentiated configuration management method for distributed application services provided in an embodiment of the present invention;

[0061] Figure 2 This is a schematic diagram of the structure of a differentiated configuration management device for distributed application services provided in an embodiment of the present invention. Detailed Implementation

[0062] 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.

[0063] This invention provides a method for differentiated configuration management of distributed application services. See also... Figure 1 , Figure 1 A flowchart illustrating a differentiated configuration management method for distributed application services provided in an embodiment of the present invention. The method includes the following steps:

[0064] Based on the differentiated configuration management method for distributed application services provided in this embodiment of the invention, the following is achieved:

[0065] Obtain multiple identification data from the application service, extract feature parameters based on the identification data, construct feature vectors based on the feature parameters, and input the feature vectors into a preset data evaluation model to obtain a data configuration score;

[0066] The identification data is divided into different levels based on the configuration scoring threshold, and different configuration management methods are adopted for the identification data of different levels; the identification data levels are divided into first-level data and second-level data.

[0067] The data at the first level is analyzed and divided into variable data and immutable data according to preset judgment conditions.

[0068] The level coefficients of each identifier data are obtained by dynamically adjusting and analyzing the variable data of the first level data and the second level data; the data levels are reclassified according to the level coefficients, and the corresponding configuration management methods are applied to the classified data.

[0069] Based on the differentiated configuration management method for distributed application services provided in this embodiment of the invention, the identification data of the target company's application services is divided into different levels, and different configuration management methods are determined according to the data of different levels, thereby reducing resource waste and application service response latency. Through dynamic adjustment and analysis, management resources can be reasonably allocated, resource waste can be avoided, and the overall efficiency of the system can be improved.

[0070] In one embodiment, multiple identification data in the application service are obtained, feature parameters are extracted based on the identification data, and a feature vector is constructed based on the feature parameters. The feature vector is then input into a preset data evaluation model to obtain a data configuration score.

[0071] In the application of AI software within enterprises, multiple identifier data points are acquired from application services. These identifier data include employee IDs, department / position information, permission levels, frequency of micro-application openings, function usage duration, approval process operation nodes, VPN access records, etc. The collected identifier data undergoes preprocessing, including data cleaning to remove invalid records such as VPN access records; data transformation is used to standardize timestamps of different formats, and VPN access records and other data are encoded for subsequent analysis. Feature parameters are extracted from this data, such as average daily message processing volume, cross-departmental communication ratio, non-working-hour response rate, multi-device login frequency, access records during abnormal periods, and sensitive file operation logs, thereby constructing a multi-dimensional feature vector that reflects employee behavioral characteristics.

[0072] The constructed feature vectors are input into a pre-defined data evaluation model. This model, based on the analytic hierarchy process (AHP) combined with machine learning algorithms, comprehensively evaluates the data from multiple dimensions, including employee activity and interaction capabilities, ultimately outputting a data configuration score. This score will serve as an important basis for subsequent resource allocation and precise management.

[0073] For example, a company uses internal AI software to collect employee behavior data, including employee IDs, departments, approval operation records, and VPN login logs. After data cleaning (such as removing invalid access records from test accounts) and standardization (unifying the time format), key features such as average daily approval processing volume (efficiency), login frequency outside of working hours (work habits), and access frequency to sensitive files (security risks) are extracted. These key features are then imported into a data evaluation model to obtain a data configuration score. This solution improves management efficiency and enhances data security and risk control.

[0074] In one embodiment, the identification data is divided into different levels based on a configuration scoring threshold, and different configuration management methods are applied to the identification data at different levels; the identification data levels are divided into first-level data and second-level data, including:

[0075] The configuration score of the identification data is compared with the preset configuration score threshold;

[0076] If the configuration score is less than the configuration score threshold, the identification data will be classified as the first level.

[0077] If the configuration score is greater than the configuration score threshold, the identification data will be classified as the second level.

[0078] When the identification data level is Level 1, manual configuration management is used;

[0079] When the identification data level is the second level, a script-based configuration management method is used.

[0080] Specifically, the manual configuration management process involves the following steps: First, extract the first-level identifier data from the specified data source. Then, manually check the data's integrity and accuracy, looking for any missing or incorrect information. Next, based on business needs and data characteristics, manually formulate configuration strategies and parameters. Operators then adjust each configuration item one by one through the management interface or command line. Afterward, test the configuration changes to ensure data processing meets expectations. Finally, record and archive the configuration process and results in detail for future auditing or traceability, while continuously monitoring the configuration effect and intervening manually in case of problems.

[0081] The script-based configuration management method involves the following steps: First, an automated configuration script is written according to the characteristics of the second-level identifier data, clearly defining the configuration logic and parameters. Then, the script's accuracy is verified in a test environment, and any potential anomalies are handled. Next, the script is incorporated into a version control system to record change history. The script is then executed automatically using configuration management tools, dynamically adjusting the configuration strategy based on different data characteristics using a parameterized approach. The configuration results are automatically checked and a detailed report is generated. If configuration fails, a pre-set rollback script is initiated to restore the previous state. The configuration scoring threshold is set by staff based on experience.

[0082] In one embodiment, by configuring scoring thresholds to divide the identified data into different levels using the above method, data characteristics and importance can be accurately identified, and a matching configuration management approach can be adopted. Manual configuration management for first-level data ensures flexibility and accuracy in special or complex scenarios; scripted configuration management for second-level data improves the processing efficiency and consistency of large amounts of standardized data. This hierarchical processing model optimizes resource allocation, ensuring the quality of critical data configuration, improving overall management efficiency, reducing manual operation costs and error rates, and enhancing the traceability and stability of configuration through automation mechanisms.

[0083] In one embodiment, dividing first-level data into variable data and immutable data according to preset judgment criteria includes:

[0084] For the identification data corresponding to the first level, determine whether its level can be changed according to the preset judgment conditions. If not, the corresponding identification data is recorded as immutable data, and the original configuration management method is maintained; if yes, the corresponding data is recorded as variable data, and the configuration management method of variable data is changed.

[0085] Specifically, it should be noted that the preset judgment conditions are set by staff based on experience. For example, in a cloud server (ECS) scenario, instance IDs and private IP addresses are judged as immutable data due to the need for uniqueness and audit trail requirements, prohibiting modification and maintaining the original configuration of the system's automatic management; while instance names and security group IDs, due to management needs (such as dynamic scaling or security policy adjustments), are allowed to be modified and are classified as mutable data, supporting real-time updates via API or console and recording audit logs. This hierarchical strategy strictly distinguishes between the two types of data through technical means (such as permission verification and audit logs in pseudocode): immutable data ensures the reliability and compliance of the system's core identifiers (such as the permanence of instance IDs in financial scenarios), while mutable data supports business agility (such as on-demand renaming of Kubernetes nodes). Ultimately, this mechanism logically achieves a balance between security constraints and operational flexibility, preventing the accidental manipulation of critical data while meeting dynamic management needs.

[0086] One implementation approach significantly enhances system security by strictly distinguishing between mutable and immutable data. For immutable data (such as instance IDs and private IPs), the unmodifiable nature ensures the long-term stability of core identifiers, preventing data inconsistencies or audit trail failures due to accidental operations. For mutable data (such as instance names and security group IDs), dynamic adjustments and audit log recording are supported, enabling management to quickly adapt to changing needs (such as automatic scaling and security policy optimization), improving operational efficiency and reducing the risk of human error. Simultaneously, this mechanism balances security control and business agility through permission verification and automated auditing (such as anomaly interception and logging in pseudocode), ultimately achieving a dual optimization of system reliability and operational convenience.

[0087] In one implementation, the level coefficients of each identifier data are obtained through dynamic adjustment analysis and calculation using variable data of the first level and second level data, including:

[0088] Using the variable data from the first level and the second level, the number of detections of the identifier data within a preset period and time is obtained, and a detection sequence is constructed. The elements in the detection sequence are sorted and mapped to obtain the order column R. j Construct the rank R of a standard time series t ;

[0089] Using the Spearman method to measure the order of column R j and R t The monotonic correlation θ is calculated using the following formula:

[0090] ;

[0091] Where θ represents monotonic correlation; the order column R j (i) and the rank R of the time series t (i) represent R respectively j and R t The rank of the i-th sample value; n represents the total number of samples;

[0092] The formula for calculating the level coefficient is:

[0093] ;

[0094] Where JB represents the level coefficient. Indicates the optimization factor at the level.

[0095] Specifically, it should be noted that the preset time period includes one day, one week, etc.; the level optimization factor is a factor that dynamically adjusts the level. The value range is (0,1). The core advantage of the Spearman method lies in its flexibility and robustness, making it particularly suitable for complex and non-ideal data environments. As a non-parametric method, it does not rely on data distribution assumptions (such as normality) but is based solely on the ranking order of variables, thus being insensitive to outliers and yielding more stable results. Spearman can capture non-linear monotonic relationships (such as exponential or logarithmic trends) and is not limited to linear associations, while also being applicable to continuous, discrete, and ordinal data (such as rankings or ratings).

[0096] Specifically, by combining first-level variable data and second-level baseline data for dynamic analysis, the system can more accurately identify the differences in importance of different labeled data; the introduction of Spearman rank correlation analysis ensures a quantitative assessment of data fluctuation patterns, thereby avoiding classification bias caused by relying solely on static thresholds.

[0097] One implementation method involves reclassifying data levels based on level coefficients and then managing the corresponding configurations of the reclassified data.

[0098] Compare the level coefficient of the identified data with the preset upper limit threshold of the level coefficient;

[0099] If the level coefficient of the identification data is less than the preset lower limit threshold of the level coefficient, the original level of the identification data will be reduced by one level, the new level of the identification data after the reduction will be set, and the corresponding configuration management method will be performed according to the new level.

[0100] If the level coefficient of the identification data is greater than the preset lower limit threshold and less than the preset upper limit threshold, the level of the identification data remains unchanged, and the corresponding configuration management method is performed according to the original level.

[0101] If the level coefficient of the identification data is greater than the preset upper limit threshold of the level coefficient, the original level of the identification data will be raised by one level, the new level of the identification data after the level has been lowered will be set, and the corresponding configuration management method will be performed according to the new level.

[0102] Specifically, it should be noted that the upper and lower thresholds of the preset level coefficients are obtained by relevant staff based on historical experience. By comparing the level coefficients with the preset thresholds, the system can automatically adjust the data level to ensure that high-value data receives more resources (such as storage, computing, and transmission priority), while low-value data is reasonably downgraded, thereby achieving dynamic data hierarchical management and improving resource utilization.

[0103] Based on the same inventive concept, embodiments of the present invention also provide a differentiated configuration management device for distributed application services. See also Figure 2 , Figure 2A schematic diagram of a differentiated configuration management device for distributed application services provided in an embodiment of the present invention includes:

[0104] A differentiated configuration management device for distributed application services provided by an embodiment of the present invention includes: a data evaluation module: acquiring multiple identification data in the application service, extracting feature parameters based on the identification data; constructing a feature vector based on the feature parameters, and inputting the feature vector into a preset data evaluation model to obtain a data configuration score;

[0105] Level Classification Module: Based on the configured scoring threshold, the identification data is divided into different levels, and different configuration management methods are adopted for the identification data of different levels; the identification data levels are divided into first-level data and second-level data; the first-level data is analyzed and divided into variable data and immutable data according to preset judgment conditions;

[0106] Dynamic adjustment module: It calculates the level coefficient of each identifier data by dynamically adjusting the variable data of the first level data and the second level data; it reclassifies the data levels according to the level coefficients and performs corresponding configuration management on the classified data.

[0107] Based on the differentiated configuration management device for distributed application services provided in this embodiment of the invention, the identification data of the target company's application services is divided into different levels in the above manner, and different configuration management methods are determined according to the data of different levels, thereby reducing resource waste and application service response latency. Through dynamic adjustment and analysis, management resources can be reasonably allocated, resource waste can be avoided, and the overall efficiency of the system can be improved.

[0108] It should be noted that, in this document, terms such as “comprising,” “including,” or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus.

[0109] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention.

Claims

1. A differentiated configuration management method for distributed application services, characterized in that, The method includes: Obtain multiple identification data from the application service, extract feature parameters based on the identification data, construct feature vectors based on the feature parameters, and input the feature vectors into a preset data evaluation model to obtain a data configuration score; The identification data is divided into different levels based on the configuration scoring threshold, and different configuration management methods are adopted for the identification data of different levels. The identification data levels are divided into first-level data and second-level data. The identification data includes employee ID, department / position information, permission level, micro-application opening frequency, function usage duration, approval process operation nodes, and VPN access records. The data at the first level is analyzed and divided into variable data and immutable data according to preset judgment conditions. The level coefficients of each identifier data are obtained by dynamically adjusting and analyzing the variable data of the first level data and the second level data; the data levels are reclassified according to the level coefficients, and the corresponding configuration management methods are applied to the classified data. The step of dividing the identification data into different levels based on the configuration scoring threshold and adopting different configuration management methods for the identification data at different levels includes: The configuration score of the identification data is compared with the preset configuration score threshold; If the configuration score is less than the configuration score threshold, the identification data will be classified as the first level. If the configuration score is greater than the configuration score threshold, the identification data will be classified as the second level. When the identification data level is Level 1, manual configuration management is used; When the identification data level is the second level, a script-based configuration management method is used.

2. The differentiated configuration management method for distributed application services according to claim 1, characterized in that, The step of dividing the first-level data into variable data and immutable data according to preset judgment conditions includes: For the identification data corresponding to the first level, determine whether its level can be changed according to the preset judgment conditions. If not, the corresponding identification data is recorded as immutable data, and the original configuration management method is maintained; if yes, the corresponding data is recorded as variable data, and the configuration management method of variable data is changed.

3. The differentiated configuration management method for distributed application services according to claim 1, characterized in that, The level coefficients for each identifier data obtained through dynamic adjustment analysis and calculation using variable data from the first level and data from the second level include: Using the variable data from the first level and the second level, the number of detections of the identifier data within a preset period and time is obtained, and a detection sequence is constructed. The elements in the detection sequence are sorted and mapped to obtain the order column R. j Construct the rank R of a standard time series t ; Using the Spearman method to measure the order of column R j and R t The monotonic correlation θ is calculated using the following formula: ； Where θ represents monotonic correlation; the order column R j (i) and the rank R of the time series t (i) represent R respectively j and R t The rank of the i-th sample value; n represents the total number of samples; The formula for calculating the level coefficient is: ； Where JB represents the level coefficient. Indicates the level of optimization factor.

4. The differentiated configuration management method for distributed application services according to claim 1, characterized in that, The method of reclassifying data levels based on level coefficients and then configuring and managing the classified data accordingly includes: Compare the level coefficient of the identified data with the preset upper limit threshold of the level coefficient; If the level coefficient of the identification data is less than the preset lower limit threshold of the level coefficient, the original level of the identification data will be reduced by one level, the new level of the identification data after the reduction will be set, and the corresponding configuration management method will be performed according to the new level. If the level coefficient of the identification data is greater than the preset lower limit threshold and less than the preset upper limit threshold, the level of the identification data remains unchanged, and the corresponding configuration management method is performed according to the original level. If the level coefficient of the identification data is greater than the preset upper limit threshold of the level coefficient, the original level of the identification data will be raised by one level, the new level of the identification data after the level has been lowered will be set, and the corresponding configuration management method will be performed according to the new level.

5. A differentiated configuration management device for distributed application services, characterized in that, The device includes: Data evaluation module: acquires multiple identification data from the application service, extracts feature parameters based on the identification data, constructs feature vectors based on the feature parameters, and inputs the feature vectors into a preset data evaluation model to obtain a data configuration score; Level Classification Module: Based on the configured scoring threshold, the identification data is divided into different levels, and different configuration management methods are adopted for the identification data of different levels; the identification data levels are divided into first-level data and second-level data; the first-level data is analyzed and divided into variable data and immutable data according to preset judgment conditions; the identification data includes employee ID, department / position information, permission level, micro-application opening frequency, function usage duration, approval process operation nodes, and VPN access records; Dynamic adjustment module: It calculates the level coefficient of each identifier data by dynamically adjusting the variable data of the first level data and the second level data; it reclassifies the data levels according to the level coefficients and performs corresponding configuration management on the classified data; The level division module includes: Level division module: The level classification module is used to compare the configuration score of the identification data with the preset configuration score threshold. If the configuration score is less than the configuration score threshold, the identification data will be classified as the first level. If the configuration score is greater than the configuration score threshold, the identification data will be classified as the second level. When the identification data level is Level 1, manual configuration management is used; When the identification data level is the second level, a script-based configuration management method is used.

6. The differentiated configuration management device for distributed application services according to claim 5, characterized in that, The level classification module includes a determination module: The determination module is used to determine whether the level of the identification data corresponding to the first level can be changed according to preset determination conditions. If not, the corresponding identification data is recorded as immutable data, and the original configuration management method is maintained; if yes, the corresponding data is recorded as variable data, and the configuration management method of the variable data is changed.

7. The differentiated configuration management device for distributed application services according to claim 5, characterized in that, The dynamic adjustment module includes a detection sequence module and a level coefficient module: The detection sequence module is used to obtain the number of times the identifier data is detected within a preset period and a preset time period using the variable data of the first-level data and the second-level data, and to construct a detection sequence; and to sort and map the elements in the detection sequence to obtain an order column R. j Construct the rank R of a standard time series t ; The level coefficient module is used by the Spearman method to measure the order of the column R. j and R t The monotonic correlation θ is calculated using the following formula: ； Where θ represents monotonic correlation; the order column R j (i) and the rank R of the time series t (i) represent R respectively j and R t The rank of the i-th sample value; n represents the total number of samples; The formula for calculating the level coefficient is: ； Where JB represents the level coefficient. Indicates the level of optimization factor.

8. The differentiated configuration management device for distributed application services according to claim 5, characterized in that, The dynamic adjustment module includes a level reset module: The reset level module is used to compare the level coefficient of the identification data with the preset upper limit threshold of the level coefficient. If the level coefficient of the identification data is less than the preset lower limit threshold of the level coefficient, the original level of the identification data will be reduced by one level, the new level of the identification data after the reduction will be set, and the corresponding configuration management method will be performed according to the new level. If the level coefficient of the identification data is greater than the preset lower limit threshold and less than the preset upper limit threshold, the level of the identification data remains unchanged, and the corresponding configuration management method is performed according to the original level. If the level coefficient of the identification data is greater than the preset upper limit threshold of the level coefficient, the original level of the identification data will be raised by one level, the new level of the identification data after the level has been lowered will be set, and the corresponding configuration management method will be performed according to the new level.

Images