File level migration problem-oriented solution generation method

Abstract

The invention discloses a method for generating a solution for a file-level migration problem, which is used for judging whether migration is blocked or not on the basis of extracting part of data for migration evaluation test. The specific position of the low-efficiency file is locked in a traversal mode, and the corresponding blocking coefficient in the efficiency diagnosis library is called by matching the preset blocking factor database to accurately predict the migration duration. And on the basis, a targeted intelligent solution is provided for the locked low-efficiency file information in combination with a knowledge graph technology. According to the method, the accuracy of migration evaluation can be improved, the prediction result can better serve for improving the migration efficiency, tasks are assisted in achieving low-risk and high-efficiency migration, and the user migration experience feeling is improved.

Description

Technical Field

[0001] The present invention relates to the field of data processing technology, and in particular to a method for generating a solution to a file level migration problem. Background Art

[0002] File-level migration refers to the process of migrating data from an old system to a new system in units of files. It usually involves moving files from one directory structure or file system to another, and the association between files needs to be maintained during the move. The main purpose of file migration is to achieve resource integration and optimization, meet business needs and adapt to business needs, and improve the flexibility and responsiveness of business systems.

[0003] Compared with other migration methods, file migration can be selectively migrated according to actual needs, and can be performed when the system is operating normally, with less impact on the business system. Overall, it has the advantages of strong migration flexibility, high security, and relatively transparent process, and is widely used in various scenarios.

[0004] File migration is a full-process operation, including preliminary migration research and evaluation, target environment preparation, data transmission, and consistency verification of data before and after. Among them, the preliminary evaluation is crucial in the overall migration process. A thorough investigation and evaluation can not only help users determine the technical feasibility, but also clarify the migration cost and duration, and identify potential risks in the migration process, so as to better formulate countermeasures based on the evaluation results. This key step is the basic prerequisite for ensuring a smooth migration process, and is of great significance to improving the accuracy and efficiency of file migration.

[0005] With the rapid development of information technology in today's world, along with the rise of high-tech technologies such as big data, the Internet of Things, and cloud computing, the data information generated by individuals and corporate organizations has also shown explosive growth. In this information context, data migration technology has played a huge role. At present, there are many tools or software developed for data migration work, which can provide a systematic migration process, including migration assessment, resource allocation, data automation migration, etc., to achieve visualization of the migration process. Among them, the full assessment in the early stage can provide accurate guidance for the subsequent migration plan and is a key stage for controlling risks. However, most migration solutions focus more on the stability and sustainability of the migration process, and invest energy in improving the level of migration automation. The assessment stage in the early stage of migration needs to be improved. The migration duration is dynamically changing. It is affected by many factors such as network, configuration, and business data. The existing assessment methods cannot obtain the accurate completion time of the migration. In addition, only the estimated duration is obtained in the pre-assessment stage, and the ability to lock inefficient files is relatively lacking, and it is impossible to assist in making migration decisions based on the assessment results.

[0006] There are several problems with the file migration solutions currently on the market: 1. The design of the pre-assessment phase is relatively simple, and the predicted migration duration lacks sufficient consideration of hardware configuration, network anomalies, and other factors that hinder migration efficiency, resulting in a low accuracy rate for the estimated duration. 2. After obtaining the estimated migration duration, there is a lack of a locking mechanism for inefficient project information, and the inefficient file location cannot be quickly locked, which is not conducive to the formulation of solutions. 3. After obtaining the assessment results, it is impossible to provide users with intelligent time-consuming improvement suggestions based on the prediction results, which has little effect on improving migration efficiency. Summary of the invention

[0007] The purpose of the present invention is to provide a method for generating solutions to file-level migration problems, which can not only improve the accuracy of migration evaluation, but also enable the prediction results to better serve the improvement of migration efficiency, assist tasks in achieving low-risk and high-efficiency migration, and enhance user migration experience.

[0008] The technical solution adopted by the present invention is:

[0009] A method for generating a solution to a file-level migration problem includes the following steps:

[0010] Step 1: Preset the migration efficiency blocking factor directory to build a migration efficiency blocking factor database, use historical migration tasks as a sample set, and train a migration efficiency diagnosis model;

[0011] Furthermore, the migration efficiency blocking factors preset in step 1 are designed according to different types of factors to construct a corresponding database. As a feasible implementation, the migration efficiency blocking factors preset include network status, system resources, business systems, storage systems, and application services.

[0012] Specifically, factors that may affect the migration efficiency are pre-screened, including network status, system resources, business systems, storage systems, application services, etc., and a blocking factor catalog is designed according to different types of factors to build a database.

[0013] The preset blocking factors are preliminarily designed as follows: 1) Network status: including but not limited to network bandwidth, number of temporary network interruptions, network delay, network congestion, etc.; 2) System resources: including but not limited to operating system type, operating system version, CPU usage, memory occupancy, number of concurrent processing, etc.; 3) Storage system: including but not limited to data volume, database type, number of database connections, disk read and write speed, file mount directory tree, etc.; 4) Business system: including but not limited to data transmission priority, etc.; 5) Application service: including but not limited to application software type, application software version, software dependency, port service, etc.

[0014] It should be noted that: the contents of the preset blocking factor catalog can be modified accordingly according to the actual situation, and the selection of the specific implementation model of the correlation analysis and importance analysis in the performance diagnosis library is not unique. Any specific content modification, method replacement, etc. made within the logical framework of the present invention should be included in the scope of protection of the technology of the present invention.

[0015] Step 2: In the pre-migration assessment phase of a single migration, part of the data to be migrated is extracted through the migration efficiency diagnosis module for migration assessment testing, and all blocking source information of the source system is extracted and stored;

[0016] Furthermore, the blocking source information in step 2 includes system resource information, network information, storage information, business information, and application service information.

[0017] Step 3: In a single migration, determine whether there will be any blockage in the migration task through the migration efficiency diagnosis module; determine the type of blocking factors according to the actual situation, calculate the estimated migration duration under the current environment status as the first output result; and obtain the location and detailed information of the inefficient file as the second output result;

[0018] Furthermore, in step 3, whether the monitored data transmission rate is lower than the network bandwidth is judged to determine whether there is an efficiency blocking phenomenon in the migration task; when the monitored data transmission rate is lower than the network bandwidth, it is determined that there is an efficiency blocking phenomenon in the migration task and the location of the blocking factor is judged to determine the type of factor causing the blocking phenomenon, and the appropriate data traversal method is selected according to different factor categories; when the monitored data transmission rate is not lower than the network bandwidth, the blocking phenomenon judgment process is terminated, and the subsequent migration duration calculation is directly performed.

[0019] Step 3 specifically includes the following steps:

[0020] Step 3-1, determine whether the monitored data transmission rate is lower than the network bandwidth; if so, determine that there is an efficiency blockage phenomenon in the migration task, and determine the location of the blocking factor to determine the type of factor causing the blocking phenomenon, and select a suitable data traversal method according to different factor categories and then execute step 3-2; otherwise, end the blocking phenomenon judgment process, and directly perform the subsequent migration duration calculation to execute step 3-4;

[0021] Step 3-2, using the selected data traversal method to traverse and lock the specific location information of the inefficient file, and write it into the performance diagnosis library;

[0022] Step 3-3, match the obtained blocking factor with the preset migration efficiency blocking factor database to determine whether there is a match; if there is a match, directly retrieve the corresponding blocking factor coefficient in the performance diagnosis library; if there is no match, write the current blocking factor as a new blocking factor into the preset blocking factor database, and do not consider the blocking effect of the corresponding blocking factor when calculating the migration duration;

[0023] Step 3-4: Calculate the migration duration to obtain the estimated migration duration.

[0024] Furthermore, the factor types of the blocking phenomenon in step 3-1 include system configuration factors, storage factors or other factor categories, and a suitable data traversal method is selected according to different factor categories.

[0025] Furthermore, the calculation formula for the migration duration in step 3-4 is as follows:

[0026] y=β0+β1x1+β2x2+…+β n x n

[0027]

[0028] Among them: y represents the final predicted migration duration; β0 represents the basic migration duration when no blocking phenomenon is detected, and β0 is mainly affected by the amount of data, bandwidth and data transmission rate; x1, x2…x n Represent n migration efficiency blocking factors, β1, β2…β n They respectively represent the size of the blocking coefficient corresponding to each blocking factor.

[0029] Specifically, the degree of influence between each blocking factor and the migration efficiency is quantitatively analyzed in the performance diagnosis library, and the calculated importance characteristic value is used as the blocking coefficient of the corresponding blocking factor to support the prediction calculation of the migration duration.

[0030] Step 4: During the migration implementation process, the migration blocking status information is monitored and recorded in real time;

[0031] Step 5: Supplement the historical sample database according to the monitoring record results, and iteratively update the performance diagnosis library after each evaluation is completed;

[0032] Furthermore, in step 5, based on the monitoring record results, for the blocking factors already in the preset database, the actual data recorded this time is input into the performance diagnosis library and recalculated with the historical data to obtain the updated blocking factor coefficient as the corresponding coefficient for subsequent calls; for the newly added blocking factors, the actual data this time is directly input into the performance diagnosis library, and the corresponding coefficient for subsequent calls is calculated, so as to realize the iterative update of the diagnosis library after each evaluation is completed.

[0033] Step 6: Combine the estimated migration duration as the first output result and the location and detailed information of the inefficient files as the second output result with the knowledge graph technology to generate a solution to the problem of file-level migration.

[0034] Furthermore, the performance diagnosis library update steps in step 5 are as follows:

[0035] Step 5-1: Add all the actual data information of the blocking factors recorded in the migration real-time monitoring as the independent variable information of the newly added sample data to the historical sample database, and then import the updated historical sample database data into the performance diagnosis library;

[0036] Step 5-2: Clean the sample data to eliminate the influence of outliers on the results;

[0037] Step 5-3: Perform correlation analysis on the preprocessed data, remove the blocking factors whose correlation exceeds the set threshold (too strong) to obtain the blocking factor set, and ensure the independence of the input factors;

[0038] Step 5-4: Import the set of blocking factors diagnosed by correlation analysis into the ensemble learning model for importance analysis to obtain the importance characteristic value of each blocking factor;

[0039] Furthermore, in order to improve the reliability of the model output, the mean of the importance coefficients obtained by the three importance models is used to represent the final importance characteristic value of the blocking factor.

[0040] Step 5-5: Set the coefficient threshold of the factor to be retained, and traverse and screen all independent variable blocking factors; for blocking factors whose importance characteristic values ​​are less than the preset threshold, it is determined that the blocking effect on the migration duration is small and is not considered in this calculation;

[0041] Step 5-6: Write the importance characteristic value of the blocking factor that passes the threshold screening into the performance diagnosis library as the blocking coefficient directly called in the subsequent duration calculation of the factor.

[0042] Specifically, the historical sample database is responsible for storing the actual monitoring data of the blocking factors monitored in the historical migration tasks, and the performance diagnosis library is responsible for calculating and storing the blocking coefficients corresponding to the factors in the migration efficiency blocking factor database. When the collected blocking factors match the preset blocking factor database, it can directly call the corresponding relative importance characteristic coefficient in the diagnosis library as the blocking coefficient for the estimated migration duration. In addition, after the migration is completed, the actual monitoring data of all blocking factors in this evaluation is received and input into the performance diagnosis library. The weight coefficient is updated through the specified algorithm to achieve iterative optimization of the model.

[0043] Furthermore, the specific steps for generating a solution in step 6 are as follows:

[0044] Step 6-1, determine the source of supporting data in the knowledge base, use expert experience as basic data, combine the problems and solutions encountered in the actual migration process, and use AI to make judgments based on the training of historical tasks to obtain knowledge information;

[0045] Step 6-2, after obtaining all source information, extract the information set to construct the question library and path library respectively;

[0046] Step 6-3, integrating the extracted and classified problem library with the path library to establish a knowledge base for subsequent solution output;

[0047] Step 6-4, importing the output results of the performance diagnosis library into the knowledge model as input data of the knowledge base;

[0048] Step 6-5, matching the input data with relevant problem information in the knowledge base, and then matching the corresponding solution path according to the problem;

[0049] Step 6-6: Output targeted solutions based on the matching results.

[0050] The present invention adopts the above technical solution, which has the following technical advantages: 1. Improving the accuracy of migration time prediction: The present invention collects and screens migration efficiency blocking factors, and uses a specified algorithm to efficiently predict the migration time. In addition, the bottleneck status information is monitored and recorded in real time during each migration implementation process, and the preset blocking factor database and performance diagnosis library are updated in combination with the bottleneck information to achieve continuous optimization and iteration of the algorithm, thereby significantly improving the accuracy of migration time prediction. 2. Providing an inefficient project locking mechanism: Based on the judgment of whether there is a blocking phenomenon, the present invention can further lock the information of inefficient projects, including file location, file size, etc., to provide users with an accurate problem path. 3. Providing an intelligent solution: The present invention provides users with targeted intelligent solutions based on the obtained migration prediction time and inefficient file information, combined with knowledge graph technology, and can better apply the results of the evaluation stage to improve migration efficiency.

[0051] The present invention can not only predict the migration duration more efficiently and accurately, but also provide intelligent solutions based on the prediction results, which is of great significance for the planning and design of migration tasks and risk control. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] The present invention is further described in detail below with reference to the accompanying drawings and specific embodiments;

[0053] Figure 1 A schematic diagram of the principle architecture of a method for generating a solution to a file level migration problem according to the present invention;

[0054] Figure 2 This is a schematic diagram of the working principle of the migration efficiency diagnosis module of the present invention;

[0055] Figure 3 Schematic diagram of the updating process of the blocking factors of the efficacy diagnosis library of the present invention;

[0056] Figure 4 A schematic diagram of the process of generating intelligent solutions based on knowledge graph technology in the present invention. DETAILED DESCRIPTION

[0057] In order to make the purpose, technical solutions and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application.

[0058] File migration technology is widely used in data optimization management and system integration business in various fields. The following are some common application scenarios: (1) Data center migration: File migration technology can help data or applications migrate across data centers to meet the needs of data centers in business expansion and integration. Ensure the security and effectiveness of data during transmission, minimize interruption time and achieve the lowest impact and smooth migration of business. (2) Cloud migration: Cloud migration is the process of migrating data, applications and related services from the local environment to the cloud environment. File migration technology can make the cloud migration process more efficient, help enterprise organizations achieve flexible management of data resources, reduce management costs and optimize data scalability. (3) Storage system migration: Storage system migration refers to the process of migrating data from one storage system to another storage device or system. File migration technology can help data migrate between different storage devices, storage networks and protocols, and plays an important role in achieving efficient and complete data migration. (4) Data backup and recovery: In the current information technology context, relevant organizations are paying more and more attention to data backup and recovery in disaster recovery scenarios. This involves the process of copying data from local storage devices to backup storage devices. File migration technology can help achieve data backup and recovery capabilities, and play an important role in improving the overall disaster resistance of data. (5) Multi-cloud environment deployment: In a multi-cloud deployment scenario, file migration technology can help users move data between different cloud service providers, achieve flexible data management across platforms, and have high security and availability.

[0059] like Figures 1 to 4As shown in one, the present invention discloses a method for generating a solution to the file-level migration problem. Based on a preset migration efficiency blocking factor directory, a migration efficiency diagnostic model is constructed using historical migration tasks as a sample data source. Through the above diagnostic model, the estimated migration duration and the location of inefficient files are judged in the pre-migration evaluation stage, and the knowledge graph technology is combined to provide users with targeted intelligent solutions; specifically, the following steps are included:

[0060] Step 1: Preset the migration efficiency blocking factor directory to build a migration efficiency blocking factor database, use historical migration tasks as a sample set, and train a migration efficiency diagnosis model;

[0061] Specifically, factors that may affect the migration efficiency are pre-screened, including network status, system resources, business systems, storage systems, application services, etc., and a blocking factor catalog is designed according to different types of factors to build a database.

[0062] The preset blocking factors are preliminarily designed as follows: 1) Network status: including but not limited to network bandwidth, number of temporary network interruptions, network delay, network congestion, etc.; 2) System resources: including but not limited to operating system type, operating system version, CPU usage, memory occupancy, number of concurrent processing, etc.; 3) Storage system: including but not limited to data volume, database type, number of database connections, disk read and write speed, file mount directory tree, etc.; 4) Business system: including but not limited to data transmission priority, etc.; 5) Application service: including but not limited to application software type, application software version, software dependency, port service, etc.

[0063] It should be noted that: the contents of the preset blocking factor catalog can be modified accordingly according to the actual situation, and the selection of the specific implementation model of the correlation analysis and importance analysis in the performance diagnosis library is not unique. Any specific content modification, method replacement, etc. made within the logical framework of the present invention should be included in the scope of protection of the technology of the present invention.

[0064] Step 2: In the pre-migration assessment phase of a single migration, part of the data to be migrated is extracted through the migration efficiency diagnosis module for migration assessment testing, and all blocking source information of the source system is extracted and stored;

[0065] Furthermore, the blocking source information in step 2 includes system resource information, network information, storage information, business information, and application service information.

[0066] Step 3: In a single migration, determine whether there will be any blockage in the migration task through the migration efficiency diagnosis module; determine the type of blocking factors according to the actual situation, calculate the estimated migration duration under the current environment status as the first output result; and obtain the location and detailed information of the inefficient file as the second output result;

[0067] Furthermore, in step 3, whether the monitored data transmission rate is lower than the network bandwidth is judged to determine whether there is an efficiency blocking phenomenon in the migration task; when the monitored data transmission rate is lower than the network bandwidth, it is determined that there is an efficiency blocking phenomenon in the migration task and the location of the blocking factor is judged to determine the type of factor causing the blocking phenomenon, and the appropriate data traversal method is selected according to different factor categories; when the monitored data transmission rate is not lower than the network bandwidth, the blocking phenomenon judgment process is terminated, and the subsequent migration duration calculation is directly performed.

[0068] Furthermore, step 3 specifically includes the following steps:

[0069] Step 3-1, determine whether the monitored data transmission rate is lower than the network bandwidth; if so, determine that there is an efficiency blockage phenomenon in the migration task, and determine the location of the blocking factor to determine the type of factor causing the blocking phenomenon, and select a suitable data traversal method according to different factor categories and then execute step 3-2; otherwise, end the blocking phenomenon judgment process, and directly perform the subsequent migration duration calculation to execute step 3-4;

[0070] Step 3-2, using the selected data traversal method to traverse and lock the specific location information of the inefficient file, and write it into the performance diagnosis library;

[0071] Step 3-3, match the obtained blocking factor with the preset migration efficiency blocking factor database to determine whether there is a match; if there is a match, directly retrieve the corresponding blocking factor coefficient in the performance diagnosis library; if there is no match, write the current blocking factor as a new blocking factor into the preset blocking factor database, and do not consider the blocking effect of the corresponding blocking factor when calculating the migration duration;

[0072] Step 3-4: Calculate the migration duration to obtain the estimated migration duration.

[0073] Furthermore, the factor types of the blocking phenomenon in step 3-1 include system configuration factors, storage factors or other factor categories, and a suitable data traversal method is selected according to different factor categories.

[0074] Furthermore, the calculation formula for the migration duration in step 3-4 is as follows:

[0075] y=β0+β1x1+β2x2+…+β n xn

[0076]

[0077] Among them: y represents the final predicted migration duration; β0 represents the basic migration duration when no blocking phenomenon is detected, and β0 is mainly affected by the amount of data, bandwidth and data transmission rate; x1, x2…x n Represent n migration efficiency blocking factors, β1, β2…β n They respectively represent the size of the blocking coefficient corresponding to each blocking factor.

[0078] Specifically, the degree of influence between each blocking factor and the migration efficiency is quantitatively analyzed in the performance diagnosis library, and the calculated importance characteristic value is used as the blocking coefficient of the corresponding blocking factor to support the prediction calculation of the migration duration.

[0079] Step 4: During the migration implementation process, the migration blocking status information is monitored and recorded in real time;

[0080] Step 5: Supplement the historical sample database according to the monitoring record results, and iteratively update the performance diagnosis library after each evaluation is completed;

[0081] Furthermore, in step 5, based on the monitoring record results, for the blocking factors already in the preset database, the actual data recorded this time is input into the performance diagnosis library and recalculated with the historical data to obtain the updated blocking factor coefficient as the corresponding coefficient for subsequent calls; for the newly added blocking factors, the actual data this time is directly input into the performance diagnosis library, and the corresponding coefficient for subsequent calls is calculated, so as to realize the iterative update of the diagnosis library after each evaluation is completed.

[0082] Furthermore, if Figure 3 As shown, the steps for updating the performance diagnosis library in step 5 are as follows:

[0083] Step 5-1: Add all the actual data information of the blocking factors recorded in the migration real-time monitoring as the independent variable information of the newly added sample data to the historical sample database, and then import the updated historical sample database data into the performance diagnosis library;

[0084] Step 5-2: Clean the sample data to eliminate the influence of outliers on the results;

[0085] Step 5-3: Perform correlation analysis on the preprocessed data, remove the blocking factors whose correlation exceeds the set threshold (too strong) to obtain the blocking factor set, and ensure the independence of the input factors;

[0086] Step 5-4: Import the set of blocking factors diagnosed by correlation analysis into the ensemble learning model for importance analysis to obtain the importance characteristic value of each blocking factor;

[0087] Furthermore, in order to improve the reliability of the model output, the mean of the importance coefficients obtained by the three importance models is used to represent the final importance characteristic value of the blocking factor.

[0088] Step 5-5: Set the coefficient threshold of the factor to be retained, and traverse and screen all independent variable blocking factors; for blocking factors whose importance characteristic values ​​are less than the preset threshold, it is determined that the blocking effect on the migration duration is small and is not considered in this calculation;

[0089] Step 5-6: Write the importance characteristic value of the blocking factor that passes the threshold screening into the performance diagnosis library as the blocking coefficient directly called in the subsequent duration calculation of the factor.

[0090] Specifically, the historical sample database is responsible for storing the actual monitoring data of the blocking factors monitored in the historical migration tasks, and the performance diagnosis library is responsible for calculating and storing the blocking coefficients corresponding to the factors in the migration efficiency blocking factor database. When the collected blocking factors match the preset blocking factor database, it can directly call the corresponding relative importance characteristic coefficient in the diagnosis library as the blocking coefficient for the estimated migration duration. In addition, after the migration is completed, the actual monitoring data of all blocking factors in this evaluation is received and input into the performance diagnosis library. The weight coefficient is updated through the specified algorithm to achieve iterative optimization of the model.

[0091] Step 6: Combine the estimated migration duration as the first output result and the location and detailed information of the inefficient files as the second output result with the knowledge graph technology to generate a solution to the problem of file-level migration.

[0092] Furthermore, if Figure 4 As shown, the specific steps for generating the solution in step 6 are as follows:

[0093] Step 6-1, determine the source of supporting data in the knowledge base, use expert experience as basic data, combine the problems and solutions encountered in the actual migration process, and use AI to make judgments based on the training of historical tasks to obtain knowledge information;

[0094] Step 6-2, after obtaining all source information, extract the information set to construct the question library and path library respectively;

[0095] Step 6-3, integrating the extracted and classified problem library with the path library to establish a knowledge base for subsequent solution output;

[0096] Step 6-4, importing the output results of the performance diagnosis library into the knowledge model as input data of the knowledge base;

[0097] Step 6-5, matching the input data with relevant problem information in the knowledge base, and then matching the corresponding solution path according to the problem;

[0098] Step 6-6: Output targeted solutions based on the matching results.

[0099] The solution for file-level migration proposed in the present invention can improve the pre-migration assessment process, better meet the assessment requirements in complex migration scenarios, and improve the accuracy of migration duration assessment and user migration experience; the present invention exposes and locks the inefficient file locations during the migration process, which can solve the problem of ambiguous inefficient file paths in the assessment phase in the scenario of rapid migration of massive data, and further provides targeted intelligent solutions, which not only reduces the labor input cost in the migration assessment phase, but also greatly improves the efficiency of migration implementation, serving multiple cloud migration scenarios.

[0100] The present invention designs a set of intelligent solutions of "identifying problems - matching paths - providing suggestions", and determines whether there is a blockage in migration based on the extraction of part of the data for migration assessment testing. The specific location of the inefficient file is locked by traversal, and the corresponding blocking coefficient in the performance diagnosis library is called by matching the preset blocking factor database to accurately predict the migration duration. On this basis, the knowledge graph technology is combined to provide a targeted intelligent solution for the locked inefficient file information. Any specific content modification, method replacement, etc. carried out within the logical framework of the present invention should be included in the scope of protection of the technology of the present invention.

[0101] The present invention reduces the manual input in the migration assessment process and effectively improves the accuracy of the pre-migration assessment. The solution can improve the assessment process in the pre-migration stage, and provide users with customized intelligent solutions for specific inefficient projects based on predictions, to assist in the efficient and stable migration of files. The present invention fully considers the main factors that may affect the performance of migration, uses relevant algorithms to accurately predict the migration duration, and combines knowledge graph technology to provide users with intelligent solutions in the pre-assessment stage. This solution can not only improve the accuracy of migration assessment, but also enable the prediction results to better serve the improvement of migration efficiency, assist tasks in achieving low-risk and high-efficiency migration, and enhance the user's migration experience.

[0102] The present invention adopts the above technical solution, which has the following technical advantages: 1. Improving the accuracy of migration time prediction: The present invention collects and screens migration efficiency blocking factors, and uses a specified algorithm to efficiently predict the migration time. In addition, the bottleneck status information is monitored and recorded in real time during each migration implementation process, and the preset blocking factor database and performance diagnosis library are updated in combination with the bottleneck information to achieve continuous optimization and iteration of the algorithm, thereby significantly improving the accuracy of migration time prediction. 2. Providing an inefficient project locking mechanism: Based on the judgment of whether there is a blocking phenomenon, the present invention can further lock the information of inefficient projects, including file location, file size, etc., to provide users with an accurate problem path. 3. Providing an intelligent solution: The present invention provides users with targeted intelligent solutions based on the obtained migration prediction time and inefficient file information, combined with knowledge graph technology, and can better apply the results of the evaluation stage to improve migration efficiency.

[0103] Obviously, the described embodiments are part of the embodiments of the present application, rather than all of the embodiments. In the absence of conflict, the embodiments in the present application and the features in the embodiments can be combined with each other. The components of the embodiments of the present application generally described and shown in the drawings here can be arranged and designed in various different configurations. Therefore, the detailed description of the embodiments of the present application is not intended to limit the scope of the application claimed for protection, but merely represents the selected embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians of the art without making creative work are within the scope of protection of the present application.

Claims

1. A method for generating a solution to a file level migration problem, characterized in that: It includes the following steps: Step 1: Preset the migration efficiency blocking factor directory to build a migration efficiency blocking factor database, use historical migration tasks as a sample set, and train a migration efficiency diagnosis model; Step 2: In the pre-migration assessment phase of a single migration, part of the data to be migrated is extracted through the migration efficiency diagnosis module for migration assessment testing, and all blocking source information of the source system is extracted and stored; Step 3: In a single migration, determine whether there will be any blockage in the migration task through the migration efficiency diagnosis module; determine the type of blocking factors according to the actual situation, calculate the estimated migration duration under the current environment status as the first output result; and obtain the location and detailed information of the inefficient file as the second output result; Step 4: During the migration implementation process, the migration blocking status information is monitored and recorded in real time; Step 5: Supplement the historical sample database according to the monitoring record results, and iteratively update the performance diagnosis library after each evaluation is completed; Step 6: Combine the estimated migration duration as the first output result and the location and detailed information of the inefficient files as the second output result with the knowledge graph technology to generate a solution to the problem of file-level migration.

2. A method for generating a solution to a file level migration problem according to claim 1, characterized in that: In step 1, the migration efficiency blocking factors are preset, and a blocking factor catalog is designed according to different types of factors to construct a corresponding database.

3. The method for generating a solution to a file level migration problem according to claim 1, characterized in that: The migration efficiency blocking factors preset in step 1 include network status, system resources, business system, storage system, and application services; the blocking source information in step 2 includes system resource information, network information, storage information, business information, and application service information.

4. The method for generating a solution to a file level migration problem according to claim 1, characterized in that: In step 3, whether the monitored data transmission rate is lower than the network bandwidth is judged to determine whether there is an efficiency blockage in the migration task; when the monitored data transmission rate is lower than the network bandwidth, it is determined that there is an efficiency blockage in the migration task and the location of the blocking factor is judged to determine the type of factor causing the blockage, and the appropriate data traversal method is selected according to different factor categories; when the monitored data transmission rate is not lower than the network bandwidth, the blocking phenomenon judgment process is terminated and the subsequent migration duration calculation is directly performed.

5. A method for generating a solution to a file level migration problem according to claim 1 or 4, characterized in that: Step 3 specifically includes the following steps: Step 3-1, determine whether the monitored data transmission rate is lower than the network bandwidth; if so, determine that there is an efficiency blockage phenomenon in the migration task, and determine the location of the blocking factor to determine the type of factor causing the blocking phenomenon, and select a suitable data traversal method according to different factor categories and then execute step 3-2; otherwise, end the blocking phenomenon judgment process, and directly perform the subsequent migration duration calculation to execute step 3-4; Step 3-2, using the selected data traversal method to traverse the specific location information of the locked inefficient file and write it into the performance diagnosis library; Step 3-3, matching the obtained blocking factors with the preset migration efficiency blocking factor database to determine whether there is a match; if there is a match, directly retrieve the corresponding blocking factor coefficient in the performance diagnosis library; if there is no match, write the current blocking factor as a new blocking factor into the preset blocking factor database, and do not consider the blocking effect of the corresponding blocking factor when calculating the migration duration; Step 3-4: Calculate the migration duration to get the estimated migration duration. The calculation formula is as follows: y=β0+β1x1+β2x2+…+β n x n Among them: y represents the final predicted migration duration; β0 represents the basic migration duration when no blocking phenomenon is detected, and β0 is mainly affected by the amount of data, bandwidth and data transmission rate; x1, x2…x n Represent n migration efficiency blocking factors, β1, β2…β n They respectively represent the size of the blocking coefficient corresponding to each blocking factor.

6. The method for generating a solution to a file level migration problem according to claim 5, characterized in that: The types of factors causing the blocking phenomenon in step 3-1 include system configuration factors and storage factors.

7. The method for generating a solution to a file level migration problem according to claim 1, characterized in that: In step 5, based on the monitoring record results, for the blocking factors already in the preset database, the actual data recorded this time is input into the performance diagnosis library and recalculated with the historical data to obtain the updated blocking factor coefficient as the corresponding coefficient for subsequent calls; for the newly added blocking factors, the actual data this time is directly input into the performance diagnosis library, and the corresponding coefficient for subsequent calls is calculated to achieve iterative update of the diagnosis library after each evaluation is completed.

8. A method for generating a solution to a file level migration problem according to claim 1 or 7, characterized in that: Step 5: Update the performance diagnostic library as follows: Step 5-1: Add all the actual data information of the blocking factors recorded in the migration real-time monitoring as the independent variable information of the newly added sample data to the historical sample database, and then import the updated historical sample database data into the performance diagnosis library; Step 5-2: Clean the sample data to eliminate the influence of outliers on the results; Step 5-3: Perform correlation analysis on the preprocessed data, remove the blocking factors whose correlation exceeds the set threshold, and obtain the blocking factor set to ensure the independence of the input factors; Step 5-4: Import the set of blocking factors diagnosed by correlation analysis into the ensemble learning model for importance analysis to obtain the importance characteristic value of each blocking factor; Step 5-5: Set the coefficient threshold size for factor retention, and traverse and screen all independent variable blocking factors; for blocking factors whose importance characteristic values ​​are less than the preset threshold, it is determined that the blocking effect on the migration duration is small and is not considered in this calculation; Step 5-6: Write the importance characteristic values ​​corresponding to the blocking factors that pass the threshold screening into the performance diagnosis library as the blocking coefficient directly called by the factor in the subsequent duration calculation.

9. The method for generating a solution to a file level migration problem according to claim 1, characterized in that: In order to improve the reliability of the model output, the average of the importance coefficients obtained by the three importance models is used to represent the final importance characteristic value of the blocking factor.

10. The method for generating a solution to a file level migration problem according to claim 1, characterized in that: The specific steps to generate the solution in step 6 are as follows: Step 6-1, determine the source of supporting data in the knowledge base, use expert experience as basic data, combine the problems and solutions encountered in the actual migration process, and use AI to make judgments based on the training of historical tasks to obtain knowledge information; Step 6-2, after obtaining all source information, extract the information set to construct the question library and path library respectively; Step 6-3, integrating the extracted and classified problem library with the path library to establish a knowledge base for subsequent solution output; Step 6-4, importing the output results of the performance diagnosis library into the knowledge model as input data of the knowledge base; Step 6-5, matching the input data with relevant problem information in the knowledge base, and then matching the corresponding solution path according to the problem; Step 6-6, outputting targeted solutions according to the matching results.

Images