Data management method and system based on optomagnetic fusion storage

Abstract

The invention provides a data management method and system based on optomagnetic fusion storage, and belongs to the field of data management.The method comprises the steps that target data are obtained, data feature information is collected, and data popularity analysis is carried out; a basic storage scheme is determined according to data popularity, and intelligent decision-making of optical storage, magnetic storage or fusion storage is realized through data criticality analysis and popularity critical coefficient evaluation; predicting a storage loss parameter according to the data popularity and the storage scheme, and optimizing an error correction code or erasure code parameter in combination with the data criticality; and finally, implementing data management according to the determined storage scheme and optimization parameters. The technical problems that a traditional storage system cannot intelligently distribute storage media according to data characteristics, and access efficiency and storage safety are difficult to balance are solved, self-adaptive selection of a storage scheme and optimization of storage quality are achieved, and efficiency and reliability of data management are improved.

Description

Technical Field

[0001] The present invention relates to the field of data management, and particularly to a data management method and system based on optical-magnetic fusion storage. Background Art

[0002] With the rapid development of information technology, the demand for data storage has shown an explosive growth, posing higher requirements for the security, reliability, and access efficiency of data storage. Traditional storage methods are mainly divided into two categories: optical storage and magnetic storage. Optical storage (such as optical discs, Blu-ray storage, etc.) has the advantages of stable media, long storage life, and low cost, and is suitable for long-term archival storage; while magnetic storage (such as hard disks, magnetic tapes, etc.) has the characteristics of fast read / write speed and strong random access ability, and is suitable for hot data with frequent access.

[0003] Currently, most data storage adopts a single storage medium or a simple hierarchical storage architecture, which has obvious deficiencies in data management. On the one hand, a single storage medium is difficult to meet the requirements of both data access efficiency and long-term preservation at the same time; on the other hand, traditional hierarchical storage usually relies on simple time rules or manual intervention for data migration, lacking the ability of intelligent analysis and judgment of data characteristics. In addition, the prediction and compensation mechanisms for data loss in the existing technology often adopt a unified redundancy strategy, failing to perform differential processing according to the importance and access patterns of different data, resulting in waste of storage resources or insufficient data security. Therefore, the existing technology has the defects that data storage cannot intelligently allocate storage media according to data characteristics, and it is difficult to balance data access efficiency and long-term preservation security. Summary of the Invention

[0004] The present invention aims at the technical problems in the existing technology that data storage cannot intelligently allocate storage media according to data characteristics and is difficult to balance data access efficiency and long-term preservation security, and provides a data management method and system based on optical-magnetic fusion storage to solve them.

[0005] The technical solutions of the present invention to solve the above technical problems are as follows:

[0006] In a first aspect, the present invention provides a data management method based on optical-magnetic fusion storage, including: obtaining target data to be stored, collecting data feature information of the target data, performing data heat analysis to obtain data heat; making a decision according to the data heat to obtain a basic storage scheme for optical storage or magnetic storage, and performing data criticality analysis according to the data feature information to obtain data criticality, combining the data heat to analyze the heat critical coefficient, making a fusion storage decision on the target data to obtain a storage scheme; according to the data heat and the storage scheme, performing data storage access loss prediction to obtain at least one of optical storage loss parameters and magnetic storage loss parameters, combining the data criticality, performing storage loss compensation optimization to obtain at least one of error correction coding parameters and erasure coding parameters; performing optical-magnetic storage data management on the target data according to the storage scheme combined with at least one of the error correction coding parameters and the erasure coding parameters.

[0007] In a second aspect, the present invention provides a data management system based on optical-magnetic fusion storage, including: a data acquisition and analysis module, configured to obtain target data to be stored, collect data feature information of the target data, and perform data heat analysis to obtain data heat; a storage decision and optimization module, configured to make a decision according to the data heat to obtain a basic storage scheme for optical storage or magnetic storage, and perform data criticality analysis according to the data feature information to obtain data criticality, combine the data heat to analyze the heat critical coefficient, and make a fusion storage decision on the target data to obtain a storage scheme; a loss prediction and compensation module, configured to perform data storage access loss prediction according to the data heat and the storage scheme to obtain at least one of optical storage loss parameters and magnetic storage loss parameters, combine the data criticality, and perform storage loss compensation optimization to obtain at least one of error correction coding parameters and erasure coding parameters; an optical-magnetic data management module, configured to perform optical-magnetic storage data management on the target data according to the storage scheme combined with at least one of the error correction coding parameters and the erasure coding parameters.

[0008] The beneficial effects of the present invention are:

[0009] Obtain the target data to be stored, collect the data feature information of the target data, conduct data heat analysis to obtain the data heat, which serves as the basis for subsequent storage decisions and lays the foundation for intelligent allocation of storage media. Make decisions based on the data heat to obtain the basic storage scheme for optical storage or magnetic storage, and conduct data criticality analysis based on the data feature information to obtain the data criticality. Combine the data heat analysis with the heat critical coefficient to make a fusion storage decision for the target data and obtain the storage scheme. Determine the preliminary selection of storage media according to the data heat (hot data tends to use magnetic storage for frequent access, and cold data tends to use optical storage for long-term archiving), and then further consider the critical degree of the data and the heat critical state to make a fusion storage decision for the data in the critical state and with high criticality, realizing the refined adjustment of the storage scheme.

[0010] According to the data heat and the storage scheme, conduct prediction of data storage access loss to obtain at least one of the optical storage loss parameters and magnetic storage loss parameters. Combine the data criticality to conduct optimization of storage loss compensation to obtain at least one of the error correction coding parameters and erasure coding parameters, so as to avoid storage waste caused by excessive redundancy while ensuring data integrity. Manage the optical and magnetic storage data of the target data according to the storage scheme combined with at least one of the error correction coding parameters and erasure coding parameters, so as to actually apply the decision-making storage scheme and the optimized error correction / erasure parameters to the storage process of the target data and complete the intelligent management of the data.

[0011] Through the above technical solutions, this application realizes the intelligent analysis of data characteristics and the adaptive allocation of storage media, improves the security of long-term preservation while ensuring data access efficiency, and enhances the reliability and resource utilization efficiency of the overall storage through targeted loss prediction and compensation optimization. BRIEF DESCRIPTION OF THE DRAWINGS

[0012] Figure 1 It is a schematic flowchart of the data management method based on optical and magnetic fusion storage provided by the present invention;

[0013] Figure 2 It is a schematic structural diagram of the data management system based on optical and magnetic fusion storage provided by the present invention.

[0014] In the drawings, the components represented by each reference numeral are as follows:

[0015] Data acquisition and analysis module 11, storage decision optimization module 12, loss prediction and compensation module 13, optical and magnetic data management module 14. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0016] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative efforts belong to the scope of protection of the present invention.

[0017] In the description of the present invention, the terms "first" and "second" are only used for descriptive purposes and cannot be construed as indicating or implying relative importance or implicitly specifying the quantity of the indicated technical features. Thus, the features defined with "first" and "second" may explicitly or implicitly include one or more of the described features. In the description of the present invention, "a plurality" means two or more, unless otherwise specifically defined.

[0018] In the description of the present invention, the term "for example" is used to mean "serving as an example, illustration, or explanation". Any embodiment described as "for example" in the present invention is not necessarily construed as being more preferred or having more advantages than other embodiments. In order for any person skilled in the art to implement and use the present invention, the following description is given. In the following description, details are set forth for the purpose of explanation. It should be understood that those of ordinary skill in the art can recognize that the present invention can be implemented without using these specific details. In other instances, well-known structures and processes are not elaborated in detail to avoid unnecessary details from obscuring the description of the present invention. Therefore, the present invention is not intended to be limited to the embodiments shown, but is consistent with the broadest scope that conforms to the principles and features disclosed in the present invention.

[0019] Embodiment 1, as Figure 1 shown, the embodiment of the present invention provides a data management method based on optical magnetic fusion storage, including:

[0020] S1. Obtain target data to be stored, collect data feature information of the target data, perform data heat analysis, and obtain data heat.

[0021] Specifically, first, obtain the target data to be stored. The target data can be any form of digital information, including but not limited to text files, images, videos, audios, or database records, etc.

[0022] After obtaining the target data, collect the data feature information of the target data. The data feature information includes key attributes such as the data type, data size, and ownership entity of the target data. Among them, the data type includes classifications such as documents, media, or structured data; the data size is recorded in units such as bytes, kilobytes, megabytes, etc.; the ownership entity identifies the owner or creator information of the target data. These data feature information constitute the basic parameters for data heat analysis.

[0023] Then, subsequently, based on the collected data feature information, data heat analysis is performed to determine the possible future access frequency of the target data, which is used as the data heat. The data heat is a quantitative indicator reflecting the activity level of the target data. By analyzing the characteristics of the target data and the historical access patterns, the future usage frequency of the target data can be determined, thereby providing a basis for the selection of the storage medium.

[0024] By obtaining the data heat of the target data, a basis for subsequent storage decisions is provided. The higher the data heat, the greater the likelihood that the target data will be frequently accessed in the future. Conversely, it indicates that the data is more suitable for long-term archival storage.

[0025] S2. Make a decision based on the data heat to obtain a basic storage plan for optical storage or magnetic storage, and perform data criticality analysis based on the data feature information to obtain the data criticality. Combine the data heat analysis with the heat critical coefficient to make a fusion storage decision for the target data and obtain the storage plan.

[0026] Specifically, first, a preliminary storage decision is made based on the obtained data heat. For example, compare the data heat of the target data with a preset data heat threshold. If the data heat is greater than or equal to the data heat threshold, it is determined to adopt magnetic storage as the basic storage plan; if the data heat is less than the data heat threshold, it is determined to adopt optical storage as the basic storage plan. This decision-making mechanism is based on the fact that high-heat data usually needs to be accessed more frequently, and magnetic storage has faster read and write speeds; low-heat data is suitable for long-term archiving, and optical storage has a longer data retention period and more stable storage characteristics.

[0027] After obtaining the basic storage plan, data criticality analysis is performed based on the data feature information of the target data. Input the data feature information into a pre-established data criticality classification table, and obtain the data criticality of the target data through index matching. The data criticality represents the degree of importance of the target data. The higher the criticality, the more important the data is, and the more serious the consequences will be if the data is lost. Therefore, higher-level data protection measures are required.

[0028] At the same time, calculate the relationship between the data heat and the data heat threshold to determine the heat critical coefficient. Specifically, first calculate the absolute difference between the data heat and the data heat threshold, then calculate the ratio of the absolute difference to the data heat threshold, and finally subtract this ratio from 1 to obtain the heat critical coefficient. The larger the heat critical coefficient, the closer the heat of the target data is to the critical threshold, and at the same time, it has higher access frequency requirements and long-term storage requirements.

[0029] Based on the data criticality and heat critical coefficient, the corrected data criticality is obtained through calculation. The corrected data criticality takes both factors into consideration, enabling the importance and access characteristics of the data to jointly influence the final storage decision. The corrected data criticality is compared with a preset data criticality threshold. If the corrected data criticality is greater than or equal to the data criticality threshold, it is decided to adopt a fusion storage scheme for the target data, that is, optical storage and magnetic storage are carried out simultaneously to provide a higher level of data security guarantee; if the corrected data criticality is less than the data criticality threshold, the previously determined basic storage scheme is directly adopted.

[0030] Through the multi-factor comprehensive decision-making mechanism, it is possible to adaptively determine the most suitable storage scheme according to the characteristics and importance of the target data, which not only meets the requirements of data access efficiency but also guarantees the security of important data, realizing intelligent data management.

[0031] S3. According to the data heat and storage scheme, predict the data storage access loss, obtain at least one of the optical storage loss parameter and the magnetic storage loss parameter, and combine the data criticality to optimize the storage loss compensation to obtain at least one of the error correction coding parameter and the erasure code parameter.

[0032] Specifically, first, based on the determined data heat and storage scheme, conduct a prediction analysis of the data storage access loss. In the actual storage process, different storage media will generate data loss due to various factors, affecting data integrity and reliability.

[0033] For the different characteristics of optical storage and magnetic storage, conduct loss predictions respectively. When the storage scheme includes optical storage, according to the data heat of the target data, establish a mapping relationship between the data heat and the data access time. Through this mapping relationship, predict the access time distribution of the target data on the optical storage medium. Since the optical storage medium will be affected by physical environmental factors such as dust and scratches during long-term storage, the longer the access time, the greater the possibility of physical damage accumulated on the medium surface. Based on this, conduct an optical storage loss mapping to obtain the optical storage loss parameter, where the important indicator is the optical storage data loss ratio, which represents the proportion of data that may be lost during optical storage.

[0034] When the storage scheme includes magnetic storage, according to the data heat of the target data, establish a mapping relationship between the data heat and the number of data accesses within a preset time period. The data loss of the magnetic storage medium mainly comes from magnetic attenuation and frequent read and write operations. The higher the access frequency, the greater the possibility of demagnetization of the magnetic material. Based on this, conduct a magnetic storage loss mapping to obtain the magnetic storage loss parameter, where the important indicator is the magnetic storage data loss ratio, which represents the proportion of data that may be lost during magnetic storage.

[0035] After obtaining the loss parameters, combined with the data criticality determined above, perform storage loss compensation optimization to determine the corresponding error correction strategy parameters. For the characteristics of different storage media, different data redundancy protection mechanisms are adopted: error correction coding strategy for optical storage; erasure coding strategy for magnetic storage. Through an optimization algorithm, determine the optimal error correction coding parameters or erasure coding parameters, such as key indicators like the redundancy rate. These parameters directly determine the recovery ability in case of data loss, affecting the degree of data integrity guarantee and storage space utilization efficiency. The optimization process comprehensively considers data criticality, loss parameters, and storage resource limitations to achieve a balance between data security and storage efficiency.

[0036] Through the above prediction and optimization mechanism, it is possible to accurately evaluate the possible loss risks before data storage according to the characteristics and importance of the target data, and formulate corresponding compensation strategies, significantly improving the reliability and security of long-term data storage.

[0037] S4. According to the storage scheme and in combination with at least one of the error correction coding parameters and erasure coding parameters, perform optical and magnetic storage data management on the target data.

[0038] In a preferred embodiment, first, according to the determined storage scheme, determine whether the target data should adopt optical storage, magnetic storage, or hybrid storage. When using optical storage, write the target data to an optical storage medium, such as an optical disc, Blu-ray storage device, etc.; when using magnetic storage, write the target data to a magnetic storage medium, such as a hard disk drive, magnetic tape, etc.; when using hybrid storage, write the target data to both an optical storage medium and a magnetic storage medium simultaneously to achieve dual protection of the data.

[0039] During the data writing process, apply the optimized error correction coding parameters or erasure coding parameters to encode the target data. For the data written to the optical storage medium, apply error correction coding parameters, such as the redundancy rate, etc., to generate appropriate redundant data and store it together with the original data. This coding mechanism can calculate and recover the original information through the remaining data and redundant data when part of the data is lost due to physical damage to the optical storage medium. For the data written to the magnetic storage medium, apply erasure coding parameters, divide the target data into multiple data blocks, and generate corresponding parity blocks to jointly form a complete encoded data set. This coding mechanism can reconstruct the original data using the remaining data blocks and parity blocks when some data blocks are lost due to demagnetization or other reasons in the magnetic storage medium, effectively preventing permanent data loss.

[0040] Preferably, a metadata index of the target data is established to record the data storage location, storage method, encoding parameters, and other key information, facilitating subsequent data access and management. When data needs to be read, the data is obtained from the corresponding storage medium according to the storage scheme and metadata index, and the corresponding decoding algorithm is applied to restore the original data. In addition, a dynamic data management strategy can be implemented according to the data popularity of the target data. Specifically, for high-popularity data, it is preferentially read from the magnetic storage medium to ensure access efficiency; for data with reduced popularity after a period of time, it may be migrated from the magnetic storage medium to the optical storage medium to optimize the storage resource allocation.

[0041] Through intelligent optical-magnetic fusion data storage management, it is possible to achieve efficient data storage and reliable protection according to the characteristics, importance, and access patterns of the target data, improving the comprehensive performance and security of data management.

[0042] Furthermore, the target data to be stored is obtained, the data characteristic information of the target data is collected, and data popularity analysis is performed to obtain the data popularity, including:

[0043] S11. Collect the data characteristic information of the target data, where the data characteristic information includes data type, data size, and ownership entity;

[0044] S12. According to the data characteristic information, use machine learning to perform data popularity analysis on the target data to obtain the data popularity.

[0045] In a feasible implementation manner, first, the data characteristic information of the target data is collected. The data characteristic information is a set of parameters describing the attributes of the target data, including but not limited to data type, data size, and ownership entity. Among them, the data type represents the content form of the target data, which can be a text document, tabular data, image, video, audio, database file, or other special formats; the data size represents the storage space occupied by the target data, which can be a value accurate to bytes; the ownership entity represents the information of the owner or creator of the target data, which can be an individual user, department unit, or application, etc. These characteristic information together constitute a multi-dimensional characteristic describing the characteristics of the target data.

[0046] After obtaining the data feature information, machine learning techniques are used for data popularity analysis. Machine learning methods can learn the internal relationship between data features and popularity from historical data access patterns, so as to predict the popularity of new target data. The collected data feature information is used as input and processed through a pre-trained machine learning model. The machine learning model can be in the form of a multi-layer perceptron, support vector machine, decision tree, random forest, or deep neural network, etc., which can capture the non-linear relationship between feature information and data popularity. This machine learning model learns the popularity performance rules under different data feature combinations through the analysis and training of historical data storage and access records. For example, small documents of a specific type have a high access frequency under a specific owner, while large media files may show different access patterns under different owners. Through this data-driven learning process, the popularity level of new target data can be accurately predicted. After being processed by the machine learning model, a quantitative index representing the popularity of the target data, that is, data popularity, is output, which reflects the expected frequency of the target data being accessed in a future period of time and provides a basis for realizing intelligent data storage management.

[0047] Through the above method of combining feature collection and machine learning, the access characteristics of different types of data can be analyzed adaptively, and the accurate prediction of data popularity can be realized, laying a foundation for the subsequent selection of storage schemes.

[0048] Furthermore, according to the data feature information, machine learning is used to perform data popularity analysis on the target data to obtain data popularity, including:

[0049] S111. Use the data popularity analysis unit of machine learning to perform data popularity analysis. The training steps of the data popularity analysis unit include:

[0050] S1111. According to the historical log data of data storage management, collect a set of sample data feature information, and collect the ratio of the access frequency of the data after storage to the preset access frequency under different sample data feature information, and label to obtain a set of sample data popularity;

[0051] S1112. Use machine learning to construct a data popularity analysis unit with data feature information and data popularity as input features and output features respectively;

[0052] S1113. Use the set of sample data feature information and the set of sample data popularity to perform iterative training and verification on the data popularity analysis unit, and complete the training after convergence.

[0053] In a preferred embodiment, a data heat analysis unit is constructed using machine learning techniques, and data heat analysis is performed through the data heat analysis unit. The data heat analysis unit outputs predicted data heat metrics based on the input data feature information. The training steps of the data heat analysis unit are as follows:

[0054] First, extract historical log data from the data storage management system. This historical log data records the detailed information of past data storage and access. Based on these historical log data, collect a set of sample data feature information that constitutes the training set. At the same time, calculate the ratio between the actual access frequency and the preset access frequency of each sample data after storage, and use this ratio as the heat metric of the sample data to form a sample data heat set. Through this ratio representation method, the access frequencies of different types of data can be standardized, facilitating cross-type comparison and analysis. Among them, the set of sample data feature information contains the feature information of a large amount of historical data, and each feature information includes dimensions such as data type, data size, and owner; the corresponding sample data heat set contains heat metric values corresponding one by one to the sample data feature information. These two sets together constitute the training data set for machine learning, providing a data basis for the subsequent training of the data heat analysis unit.

[0055] Then, use a machine learning algorithm to construct an initial structure of the data heat analysis unit with the data feature information as the input feature and the corresponding data heat as the output feature. The data heat analysis unit can be implemented using various machine learning frameworks, such as multi-layer perceptron, support vector regression, random forest regression, or deep neural network, etc. The specific structure design of the data heat analysis unit is optimized according to the feature dimensions and complexity to capture the complex non-linear relationship between features and heat. Next, use the constructed set of sample data feature information and the sample data heat set to perform iterative training on the data heat analysis unit. During the training process, input the sample data feature information into the data heat analysis unit, calculate the error by comparing it with the actual sample data heat in the sample data heat set, and adjust the parameters of the data heat analysis unit through backpropagation or other optimization algorithms to continuously reduce the prediction error. At the same time, use techniques such as cross-validation to evaluate the performance of the data heat analysis unit and prevent overfitting. When the prediction error of the data heat analysis unit is stable and reaches the preset convergence condition, the training process is completed. After a complete training process, the data heat analysis unit can accurately capture the internal relationship between data features and heat, and achieve high-precision prediction of the heat of new target data. In practical applications, only need to input the feature information of the new target data into the trained data heat analysis unit to obtain its predicted heat value, which is used as the data heat to provide a basis for subsequent storage decisions.

[0056] Through a machine learning method driven by historical data, intelligent learning and prediction of data access patterns are realized, improving the accuracy and adaptability of data heat analysis and laying a foundation for intelligent decision-making in optical-magnetic fusion storage.

[0057] Furthermore, make a decision according to the data heat to obtain a basic storage scheme for optical storage or magnetic storage, and perform data criticality analysis according to the data feature information to obtain the data criticality. Combine the heat criticality coefficient analyzed by the data heat to make a fusion storage decision on the target data and obtain a storage scheme, including:

[0058] S21. Judge whether the data heat is greater than or equal to the data heat threshold. If so, make a decision to obtain a basic storage scheme for magnetic storage; if not, make a decision to obtain a basic storage scheme for optical storage;

[0059] S22. Input the data feature information into the data criticality classification table and index and classify to obtain the data criticality. Among them, the data criticality classification table includes the index relationship between sample data feature information and sample data criticality;

[0060] S23. Calculate the ratio of the absolute difference between the data heat and the data heat threshold to the data heat threshold, and calculate to obtain the heat criticality coefficient of the data heat;

[0061] S24. Calculate and obtain the corrected data criticality of the target data according to the data criticality and the heat criticality coefficient;

[0062] S25. Judge whether the corrected data criticality is greater than or equal to the data criticality threshold. If so, make a decision to obtain a storage scheme for fusing and storing the target data; if not, use the basic storage scheme as the storage scheme.

[0063] In a preferred embodiment, first, a preliminary storage decision is made based on the obtained data heat. Specifically, a data heat threshold is set, which is the demarcation point for distinguishing high-heat data and low-heat data. This data heat threshold is determined by selecting its inflection point or characteristic value through statistical analysis of the access frequency distribution curve of historical data, combined with the experience of domain experts and business requirement assessment. Compare the data heat of the target data with this data heat threshold: If the data heat is greater than or equal to this data heat threshold, it indicates that the target data has a relatively high access frequency requirement, and the decision is to use magnetic storage as the basic storage solution to provide faster data access speed; if the data heat is less than this data heat threshold, it indicates that the target data has a relatively low access frequency and is suitable for long-term archiving, and the decision is to use optical storage as the basic storage solution to provide a longer data preservation capacity. Then, perform a data criticality analysis on the target data to evaluate the importance level of the target data. Specifically, first, establish a data criticality classification table, which contains the index relationship between sample data characteristic information and the corresponding sample data criticality. Input the data characteristic information of the target data into this data criticality classification table, and through index matching, obtain the sample data criticality most similar to it as the data criticality index of the target data. Data criticality is a quantitative indicator reflecting the importance of the target data. The higher the data criticality, the more important the data, and the more serious the impact that data loss may cause.

[0064] Subsequently, calculate the heat critical coefficient to evaluate the heat characteristics of the target data. The heat critical coefficient is used to represent the proximity of the heat of the target data to the data heat threshold, and the calculation formula is: heat critical coefficient = 1 - (|data heat - data heat threshold| / data heat threshold). Among them, |data heat - data heat threshold| represents the absolute difference between the data heat and the heat threshold. According to this formula, when the data heat of the target data is close to the data heat threshold, the difference is small, and the heat critical coefficient is close to 1; when the heat is far from the threshold, the difference is large, and the heat critical coefficient is close to 0. The larger the heat critical coefficient, the more ambiguous the access characteristics of the target data, and the greater the possibility of having both high-frequency access and long-term storage requirements. Then, based on the data criticality and the heat critical coefficient, calculate the corrected data criticality. The calculation method of the corrected data criticality is to multiply the data criticality by the heat critical coefficient, that is: corrected data criticality = data criticality × heat critical coefficient. This calculation method comprehensively considers the importance and access characteristics of the data. When the data is both important and in a heat critical state, the corrected criticality will reach the highest value, indicating that the data requires a higher level of protection.

[0065] After that, based on the criticality of the corrected data, a final storage scheme decision is made. Specifically, a pre-determined data criticality threshold is obtained, which is comprehensively set based on storage resources and security level requirements, combined with expert experience and historical data analysis results. The calculated criticality of the corrected data is compared with this data criticality threshold. If the criticality of the corrected data is greater than or equal to this data criticality threshold, it indicates that the target data has high importance and protection requirements, and the decision is to adopt a fusion storage scheme, that is, optical storage and magnetic storage are carried out simultaneously to provide a dual data protection mechanism; if the criticality of the corrected data is less than this data criticality threshold, it indicates that the basic storage scheme can already meet the data protection requirements, and the determined basic storage scheme is directly adopted as the final storage scheme.

[0066] Through a multi-level and multi-factor decision-making mechanism, the access requirements, importance level, and characteristic ambiguity of the data can be comprehensively considered, and the most suitable storage scheme can be adaptively determined, which not only ensures the security of important data but also optimizes the allocation efficiency of storage resources, realizing intelligent data storage management.

[0067] Furthermore, according to the data heat and the storage scheme, a prediction of data storage access loss is carried out to obtain at least one of the optical storage loss parameter and the magnetic storage loss parameter, including:

[0068] S31. When the storage scheme includes optical storage, map the data access time according to the data heat;

[0069] S32. According to the data access time, perform optical storage loss mapping to obtain the optical storage loss parameter, where the optical storage loss parameter includes the optical storage data loss ratio;

[0070] S33. When the storage scheme includes magnetic storage, map the number of data accesses within a preset time period according to the data heat;

[0071] S34. According to the number of data accesses, perform magnetic storage loss mapping to obtain the magnetic storage loss parameter, where the magnetic storage loss parameter includes the magnetic storage data loss ratio.

[0072] In a preferred embodiment, when the storage scheme includes optical storage, first, according to the data popularity of the target data, the data access time is obtained through the mapping relationship, for example, 100 days. Specifically, a mapping table or mapping function between data popularity and data access time is established. Target data with lower data popularity usually indicates a longer access time interval, that is, a longer time period from when the data is stored until it is accessed next time; target data with higher data popularity indicates a shorter access time interval. This mapping process can be implemented through a table lookup method, function calculation, or other mathematical models, and a value representing the expected data access time is output. After obtaining the expected data access time, an optical storage loss mapping is performed to calculate the optical storage loss parameter. During long-term storage, the optical storage medium is affected by many physical environmental factors, such as dust accumulation, surface scratches, material aging, etc. These factors intensify over time, resulting in an increased risk of data loss. By establishing a mapping relationship between the data access time and the optical storage loss parameter, the possible optical storage loss parameter under the expected data access time is evaluated. The core indicator of the optical storage loss parameter is the optical storage data loss ratio, which represents the proportion of data that may be lost on the optical medium within the expected time period. The higher the loss ratio, the greater the difficulty of data recovery and the stronger the error correction ability required.

[0073] When the storage scheme includes magnetic storage, first, according to the data popularity of the target data, the expected number of data accesses within a preset time period is obtained through the mapping relationship, for example, 1 time per day, and the preset time period is, for example, 1 day. Specifically, a mapping table or mapping function between data popularity and access frequency is established. Target data with higher data popularity usually indicates a higher access frequency, that is, a greater number of accesses within the preset time period; target data with lower data popularity indicates a lower access frequency. This mapping process can also be implemented through a table lookup method, function calculation, or other mathematical models, and a value representing the expected number of data accesses is output. After obtaining the expected number of data accesses, a magnetic storage loss mapping is performed to calculate the magnetic storage loss parameter. During frequent read and write operations, the magnetic storage medium is affected by factors such as magnetic attenuation and mechanical wear. These factors intensify as the number of accesses increases, resulting in an increased risk of data loss. By establishing a mapping relationship between the number of accesses and the magnetic storage loss parameter, the possible magnetic storage loss parameter under the expected number of data accesses is evaluated. The core indicator of the magnetic storage loss parameter is the magnetic storage data loss ratio, which represents the proportion of data that may be lost on the magnetic medium under the expected access frequency. Similarly, the higher the loss ratio, the greater the difficulty of data recovery and the stronger the error correction ability required.

[0074] Through the storage loss prediction mechanism based on data heat, it is possible to accurately evaluate the loss risk of target data under different storage schemes according to the physical characteristics of different storage media, providing a basis for subsequent error correction coding and erasure code strategy optimization, and improving the reliability and security of long-term data storage.

[0075] Furthermore, in combination with the data criticality, storage loss compensation optimization is performed to obtain at least one of the error correction coding parameters and erasure code parameters, including:

[0076] S31. When the storage scheme includes optical storage, randomly configure the first error correction coding parameters, where the first error correction coding parameters include the redundancy rate of error correction coding in optical storage;

[0077] S32. Calculate the first error correction fitness of the first error correction coding parameters according to the data criticality and optical storage loss parameters;

[0078] S33. Continuously randomly configure error correction coding parameters and calculate the error correction fitness, optimize the error correction coding parameters until convergence, and retain the error correction coding parameters with the maximum error correction fitness;

[0079] S34. When the storage scheme includes magnetic storage, randomly configure the first erasure code parameters, where the first erasure code parameters include the redundancy rate of erasure code in magnetic storage;

[0080] S35. Calculate the first erasure fitness of the first erasure code parameters according to the data criticality and magnetic storage loss parameters;

[0081] S36. Continuously randomly configure erasure code parameters and calculate the erasure fitness, optimize the erasure code parameters until convergence, and retain the erasure code parameters with the maximum erasure fitness.

[0082] In a preferred embodiment, when determining the optimal error correction coding parameters for an optical storage scheme, first, the initial error correction coding parameters, i.e., the first error correction coding parameters, are randomly configured. The core index of the error correction coding parameters is the redundancy rate of error correction coding in optical storage, which is expressed as the ratio of the redundant data stored additionally to ensure data integrity to the original data, for example, 15%. The higher the redundancy rate, the stronger the ability to resist data loss, but the greater the storage space occupation. The initial parameter configuration is random to cover a wider parameter space and avoid falling into a local optimal solution. Secondly, according to the obtained data criticality and optical storage loss parameters, the fitness value of the first error correction coding parameters, i.e., the first error correction fitness, is calculated. The fitness calculation comprehensively considers two aspects: data protection ability and storage efficiency. The higher the data criticality, the greater the data protection requirement; the higher the optical storage loss parameter, the greater the difficulty of data recovery. By combining these factors with the error correction coding parameters, the fitness value is calculated, and the higher this value indicates the more reasonable the parameter configuration. Then, an iterative optimization process is entered. New error correction coding parameters are continuously randomly configured, their fitness values are calculated, compared with the current optimal error correction coding parameters, and the error correction coding parameter configuration with a higher fitness is retained. This process is repeated continuously until the preset convergence condition is met, such as the fitness change being less than the threshold or reaching the maximum number of iterations. Finally, the error correction coding parameters with the highest fitness in all iterations are retained as the final result.

[0083] When determining the optimal erasure code parameters for a magnetic storage scheme, first, the initial erasure code parameters, i.e., the first erasure code parameters, are randomly configured. The core of the erasure code parameters is the redundancy rate of the erasure code in magnetic storage, and its implementation mechanism is different from that of error correction coding. Data protection is mainly achieved through data block segmentation and parity block generation. The initial parameter configuration is also random to increase the diversity of parameter search. Secondly, according to the obtained data criticality and magnetic storage loss parameters, the fitness value of the first erasure code parameters, i.e., the first erasure fitness, is calculated. The fitness calculation also comprehensively considers two aspects: data protection ability and storage efficiency, but is optimized and adjusted according to the characteristics of magnetic storage. Then, an iterative optimization process is entered. New erasure code parameters are continuously randomly configured, their fitness values are calculated, compared with the current optimal erasure code parameters, and the parameter configuration with a higher fitness is retained. This process is repeated continuously until the preset convergence condition is met. Finally, the erasure code parameters with the highest fitness in all iterations are retained as the final result.

[0084] Through the iterative optimization method based on random search and fitness evaluation, the best balance point can be found between ensuring data security and storage efficiency, providing differentiated protection strategies for data of different importance levels, and improving the data management efficiency of optical-magnetic fusion storage. The convergence characteristics of this method ensure that the optimization process can be completed within a limited time, meeting the application requirements of actual storage.

[0085] Further, calculating a first error correction fitness of the first error correction coding parameter according to the data criticality and the optical storage loss parameter includes:

[0086] S321. Calculate a ratio of the first error correction coding parameter to the optical storage loss parameter, and in combination with the data criticality, perform a correction calculation on the ratio to obtain a first error correction data fitness;

[0087] S322. Calculate a ratio of a preset error correction coding parameter to the first error correction coding parameter to obtain a first error correction redundancy fitness;

[0088] S323. Calculate a first error correction fitness according to the first error correction data fitness and the first error correction redundancy fitness.

[0089] In a preferred embodiment, first, calculate the first error correction data fitness, which measures the adaptability of the error correction coding parameter to data protection. Specifically, first, calculate a ratio of the first error correction coding parameter to the optical storage loss parameter. This ratio represents the matching degree between the error correction ability and the expected loss risk. The larger the ratio, the higher the error correction redundancy degree is higher than the expected loss risk, and the stronger the data protection ability. Then, combine this ratio with the data criticality for correction calculation. The correction method is to perform a multiplication operation on the ratio and the data criticality, so that the data criticality directly affects the fitness result. Through the multiplication relationship, when the data criticality increases, the first error correction data fitness also increases accordingly. Thus, in subsequent optimization, it tends to select a higher error correction redundancy rate for important data to provide stronger data protection ability. Through the correction calculation combining the criticality, obtain the first error correction data fitness characterizing the adaptability of the error correction parameter to data protection.

[0090] Secondly, calculate the first error correction redundancy fitness, which measures the storage efficiency of the error correction coding parameter. Specifically, calculate a ratio of the preset error correction coding parameter to the first error correction coding parameter as the first error correction redundancy fitness. Among them, the preset error correction coding parameter is a pre-determined reference parameter, representing a reasonable redundancy level. This ratio represents the redundancy degree of the current error correction coding parameter relative to the reference parameter. The larger the ratio, the lower the redundancy degree of the current error correction coding parameter is lower than the reference level, and the higher the storage efficiency. Then, comprehensively obtain the first error correction data fitness and the first error correction redundancy fitness to obtain the final first error correction fitness. The comprehensive calculation can adopt weighted average, geometric average or other functional relationships. By adjusting the weight coefficient, the importance of data protection ability and storage efficiency can be flexibly balanced, and the error correction strategy can be optimized for different application scenarios.

[0091] Through multi-level error correction fitness calculation, the performance of error correction coding parameters can be comprehensively evaluated, and the optimal parameter configuration can be selected during the iterative optimization process. This method not only considers the data protection requirements but also takes into account the efficient utilization of storage resources, achieving comprehensive optimization of storage performance. In addition, this fitness-based evaluation mechanism has good scalability and can adjust the calculation model according to specific application requirements to adapt to the characteristic requirements of different storage systems.

[0092] Further, according to the data criticality and magnetic storage loss parameters, calculating the first error correction fitness of the first erasure code parameter includes:

[0093] S351. Calculate the ratio of the first erasure code parameter to the magnetic storage loss parameter, combine the data criticality, and perform a correction calculation on the ratio to obtain the first erasure data fitness;

[0094] S352. Calculate the ratio of the preset erasure code parameter to the first erasure code parameter to obtain the first erasure redundancy fitness;

[0095] S353. Calculate the first erasure fitness based on the first erasure data fitness and the first erasure redundancy fitness.

[0096] In a preferred implementation manner, first, calculate the first erasure data fitness, which measures the adaptability of the erasure code parameter to magnetic storage data protection. Specifically, first calculate the ratio of the first erasure code parameter to the magnetic storage loss parameter. This ratio represents the matching degree between the erasure ability and the expected magnetic storage loss risk. The larger the ratio, the higher the erasure redundancy degree is higher than the expected loss risk, and the stronger the data recovery ability. Then, combine this ratio with the data criticality for correction calculation. The correction method is to perform a multiplication operation on the ratio and the data criticality, so that the data criticality directly affects the fitness evaluation result. Through the multiplication relationship, when the data criticality increases, the first erasure data fitness also increases accordingly, so that in subsequent optimization, a higher erasure redundancy rate tends to be selected for important data to provide stronger data recovery ability.

[0097] Secondly, calculate the first erasure redundancy fitness, which measures the storage efficiency of erasure code parameters. Specifically, calculate the ratio of the preset erasure code parameters to the first erasure code parameters as the first erasure redundancy fitness. Among them, the preset erasure code parameters are pre-determined reference parameters, representing a reasonable redundancy level under magnetic storage media. This ratio represents the redundancy degree of the current erasure code parameters relative to the reference erasure code parameters. The larger the ratio, the lower the redundancy degree of the current parameters compared to the reference level, and the higher the storage efficiency, avoiding waste of storage space caused by overly redundant storage. After that, combine the obtained first erasure data fitness and the first erasure redundancy fitness to obtain the final first erasure fitness. The combined calculation can adopt the weighted average method, and by adjusting the weights of different indicators, flexibly balance the importance of data protection capabilities and storage efficiency. For different application scenarios and different levels of magnetic storage systems, the weight configuration can be dynamically adjusted to optimize the erasure strategy.

[0098] Through multi-level calculation of erasure fitness, the performance of erasure code parameters can be comprehensively evaluated, and the parameter configuration most suitable for the characteristics of magnetic storage can be selected during the iterative optimization process. This method takes into account the particularity of magnetic storage media, optimizes specifically for the demagnetization risk caused by frequent access, and also takes into account the efficient utilization of storage resources, achieving comprehensive optimization of the performance of magnetic storage systems. In addition, this fitness evaluation mechanism and the aforementioned optical storage error correction fitness calculation method form a unified evaluation system, providing a consistent optimization framework for the optical-magnetic fusion storage system, enhancing the overall coordination and manageability.

[0099] Embodiment 2, as Figure 2 shown, based on the same inventive concept as the data management method for optical-magnetic fusion storage provided in Embodiment 1, the embodiment of the present invention also provides a data management system for optical-magnetic fusion storage, including:

[0100] A data acquisition and analysis module 11, configured to acquire target data to be stored, collect data feature information of the target data, perform data popularity analysis, and obtain data popularity;

[0101] A storage decision optimization module 12, configured to make a decision based on the data popularity to obtain a basic storage scheme for optical storage or magnetic storage, perform data criticality analysis based on the data feature information to obtain data criticality, combine the data popularity analysis to analyze the popularity critical coefficient, and make a fusion storage decision on the target data to obtain a storage scheme;

[0102] A loss prediction and compensation module 13, configured to perform data storage access loss prediction based on the data popularity and the storage scheme to obtain at least one of optical storage loss parameters and magnetic storage loss parameters, combine the data criticality, and perform storage loss compensation optimization to obtain at least one of error correction coding parameters and erasure code parameters;

[0103] The optical magnetic data management module 14 is used to perform optical magnetic storage data management on the target data according to at least one of the error correction coding parameters and the erasure code parameters in combination with the storage scheme.

[0104] Furthermore, the data acquisition and analysis module 11 includes the following execution steps:

[0105] Collect the data feature information of the target data, where the data feature information includes data type, data size, and belonging entity;

[0106] According to the data feature information, use machine learning to perform data heat analysis on the target data to obtain the data heat.

[0107] Furthermore, the data acquisition and analysis module 11 includes the following execution steps:

[0108] Use the data heat analysis unit of machine learning to perform data heat analysis, and the training steps of the data heat analysis unit include:

[0109] According to the historical log data of data storage management, collect the set of sample data feature information, and collect the ratio of the access frequency after data storage to the preset access frequency under different sample data feature information, and label to obtain the set of sample data heat;

[0110] Use machine learning to construct a data heat analysis unit with the data feature information and the data heat as the input feature and the output feature respectively;

[0111] Use the set of sample data feature information and the set of sample data heat to perform iterative training and verification on the data heat analysis unit, and complete the training after convergence.

[0112] Furthermore, the storage decision optimization module 12 includes the following execution steps:

[0113] Judge whether the data heat is greater than or equal to the data heat threshold. If so, decide to obtain the basic storage scheme for magnetic storage. If not, decide to obtain the basic storage scheme for optical storage;

[0114] Input the data feature information into the data criticality classification table, and index and classify to obtain the data criticality, where the data criticality classification table includes the index relationship between the sample data feature information and the sample data criticality;

[0115] Calculate the ratio of the absolute difference between the data heat and the data heat threshold to the data heat threshold, and calculate to obtain the heat criticality coefficient of the data heat;

[0116] Calculate the corrected data criticality of the target data according to the data criticality and the heat critical coefficient;

[0117] Determine whether the corrected data criticality is greater than or equal to the data criticality threshold. If so, decide to obtain a storage scheme for fusing and storing the target data. If not, use the basic storage scheme as the storage scheme.

[0118] Further, the loss prediction and compensation module 13 includes the following execution steps:

[0119] When the storage scheme includes optical storage, map to obtain the data access time according to the data heat;

[0120] Perform optical storage loss mapping according to the data access time to obtain optical storage loss parameters, where the optical storage loss parameters include the optical storage data loss ratio;

[0121] When the storage scheme includes magnetic storage, map to obtain the number of data accesses within a preset time period according to the data heat;

[0122] Perform magnetic storage loss mapping according to the number of data accesses to obtain magnetic storage loss parameters, where the magnetic storage loss parameters include the magnetic storage data loss ratio.

[0123] Further, the loss prediction and compensation module 13 includes the following execution steps:

[0124] When the storage scheme includes optical storage, randomly configure the first error correction coding parameters, where the first error correction coding parameters include the redundancy rate of error correction coding in optical storage;

[0125] Calculate the first error correction fitness of the first error correction coding parameters according to the data criticality and the optical storage loss parameters;

[0126] Continue to randomly configure error correction coding parameters and calculate the error correction fitness, optimize the error correction coding parameters until convergence, and retain the error correction coding parameters with the maximum error correction fitness;

[0127] When the storage scheme includes magnetic storage, randomly configure the first erasure code parameters, where the first erasure code parameters include the redundancy rate of erasure code in magnetic storage;

[0128] Calculate the first erasure fitness of the first erasure code parameters according to the data criticality and the magnetic storage loss parameters;

[0129] Continue to randomly configure erasure code parameters and calculate the erasure fitness, optimize the erasure code parameters until convergence, and retain the erasure code parameters with the maximum erasure fitness.

[0130] Further, the loss prediction compensation module 13 further includes the following execution steps:

[0131] Calculate the ratio of the first error correction coding parameter to the optical storage loss parameter, and in combination with the data criticality, perform a correction calculation on the ratio to obtain the first error correction data fitness;

[0132] Calculate the ratio of the preset error correction coding parameter to the first error correction coding parameter to obtain the first error correction redundancy fitness;

[0133] Calculate and obtain the first error correction fitness according to the first error correction data fitness and the first error correction redundancy fitness.

[0134] Further, the loss prediction compensation module 13 further includes the following execution steps:

[0135] Calculate the ratio of the first erasure code parameter to the magnetic storage loss parameter, and in combination with the data criticality, perform a correction calculation on the ratio to obtain the first erasure data fitness;

[0136] Calculate the ratio of the preset erasure code parameter to the first erasure code parameter to obtain the first erasure redundancy fitness;

[0137] Calculate and obtain the first erasure fitness according to the first erasure data fitness and the first erasure redundancy fitness.

[0138] It should be noted that in the above embodiments, the descriptions of the various embodiments have their own emphases. For the parts not described in detail in a certain embodiment, reference may be made to the relevant descriptions of other embodiments.

[0139] Those skilled in the art should understand that the embodiments of the present invention can be provided as a method, a system, or a computer program product. Therefore, the present invention can adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present invention can adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0140] The present invention is described with reference to the flowcharts and / or block diagrams of methods, apparatuses (systems), and computer program products according to the embodiments of the present invention. It should be understood that each process and / or block in the flowcharts and / or block diagrams, and the combination of processes and / or blocks in the flowcharts and / or block diagrams, can be realized by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded computer, or other programmable data processing devices to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing devices generate for realizing the processFigure 1 means for the functions specified in one process or multiple processes and / or blocks Figure 1 or multiple blocks.

[0141] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer-readable memory produce a manufactured article including an instruction means, and the instruction means implements the functions specified in Figure 1 one process or multiple processes and / or blocks Figure 1 or multiple blocks.

[0142] These computer program instructions can also be loaded onto a computer or other programmable data processing device, so that a series of operation steps are executed on the computer or other programmable device to produce a computer-implemented process. Thus, the instructions executed on the computer or other programmable device provide steps for implementing the functions specified in Figure 1 one process or multiple processes and / or blocks Figure 1 or multiple blocks.

[0143] Although the preferred embodiments of the present invention have been described, those skilled in the art can make additional changes and modifications to these embodiments once they know the basic creative concepts.

[0144] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the present invention and its equivalent technologies, the present invention also intends to include these changes and modifications.

Claims

1. A data management method based on optical magnetic fusion storage, characterized in that The method includes: Obtain target data to be stored, collect data feature information of the target data, perform data heat analysis, and obtain data heat; Make a decision according to the data heat to obtain a basic storage scheme for optical storage or magnetic storage, perform data criticality analysis according to the data feature information to obtain data criticality, combine the data heat to analyze the heat critical coefficient, and make a fusion storage decision on the target data to obtain a storage scheme; According to the data heat and the storage scheme, predict the data storage access loss, obtain at least one of the optical storage loss parameter and the magnetic storage loss parameter, combine the data criticality, and perform storage loss compensation optimization to obtain at least one of the error correction coding parameter and the erasure coding parameter; According to the storage scheme and in combination with at least one of the error correction coding parameter and the erasure coding parameter, perform optical-magnetic storage data management on the target data.

2. The data management method based on optical magnetic fusion storage according to claim 1, wherein Obtain target data to be stored, collect data feature information of the target data, perform data heat analysis, and obtain data heat, including: Collect data feature information of the target data, where the data feature information includes data type, data size, and belonging entity; According to the data feature information, use machine learning to perform data heat analysis on the target data to obtain data heat.

3. The data management method based on optical-magnetic fusion storage according to claim 2, characterized in that According to the data feature information, use machine learning to perform data heat analysis on the target data to obtain data heat, including performing data heat analysis using a data heat analysis unit of machine learning. The training steps of the data heat analysis unit include: According to the historical log data of data storage management, collect a set of sample data feature information, and collect the ratio of the access frequency after data storage to the preset access frequency under different sample data feature information, and label to obtain a set of sample data heat; Use machine learning to construct a data heat analysis unit with data feature information and data heat as input features and output features respectively; Use the set of sample data feature information and the set of sample data heat to perform iterative training and verification on the data heat analysis unit, and complete the training after convergence.

4. The data management method based on optical magnetic fusion storage according to claim 1, characterized in that Make a decision according to the data heat to obtain a basic storage scheme for optical storage or magnetic storage, perform data criticality analysis according to the data feature information to obtain data criticality, combine the data heat to analyze the heat critical coefficient, and make a fusion storage decision on the target data to obtain a storage scheme, including: Judge whether the data heat is greater than or equal to the data heat threshold. If so, make a decision to obtain a basic storage scheme for magnetic storage. If not, make a decision to obtain a basic storage scheme for optical storage; Input the data feature information into the data criticality classification table, and index and classify to obtain data criticality, where the data criticality classification table includes the index relationship between sample data feature information and sample data criticality; Calculate the ratio of the absolute difference between the data heat and the data heat threshold to the data heat threshold, and calculate the heat critical coefficient of the data heat; Calculate the corrected data criticality of the target data according to the data criticality and the heat critical coefficient; Determine whether the corrected data criticality is greater than or equal to the data criticality threshold. If so, decide to obtain a storage scheme for fusing and storing the target data. If not, use the basic storage scheme as the storage scheme.

5. The data management method based on optical-magnetic fusion storage according to claim 1, wherein Perform data storage access loss prediction according to the data heat and the storage scheme to obtain at least one of an optical storage loss parameter and a magnetic storage loss parameter, including: When the storage scheme includes optical storage, map the data access time according to the data heat; Perform optical storage loss mapping according to the data access time to obtain an optical storage loss parameter, where the optical storage loss parameter includes an optical storage data loss ratio; When the storage scheme includes magnetic storage, map the number of data accesses within a preset time period according to the data heat; Perform magnetic storage loss mapping according to the number of data accesses to obtain a magnetic storage loss parameter, where the magnetic storage loss parameter includes a magnetic storage data loss ratio.

6. The data management method based on optical magnetic fusion storage according to claim 1, wherein Combine the data criticality to perform storage loss compensation optimization to obtain at least one of an error correction coding parameter and an erasure code parameter, including: When the storage scheme includes optical storage, randomly configure a first error correction coding parameter, where the first error correction coding parameter includes the redundancy rate of error correction coding in optical storage; Calculate the first error correction fitness of the first error correction coding parameter according to the data criticality and the optical storage loss parameter; Continue to randomly configure the error correction coding parameter and calculate the error correction fitness, optimize the error correction coding parameter until convergence, and retain the error correction coding parameter with the maximum error correction fitness; When the storage scheme includes magnetic storage, randomly configure a first erasure code parameter, where the first erasure code parameter includes the redundancy rate of the erasure code in magnetic storage; Calculate the first erasure fitness of the first erasure code parameter according to the data criticality and the magnetic storage loss parameter; Continue to randomly configure the erasure code parameter and calculate the erasure fitness, optimize the erasure code parameter until convergence, and retain the erasure code parameter with the maximum erasure fitness.

7. The data management method based on optical-magnetic fusion storage according to claim 6, wherein Calculate the first error correction fitness of the first error correction coding parameter according to the data criticality and the optical storage loss parameter, including: Calculate the ratio of the first error correction coding parameter to the optical storage loss parameter, combine the data criticality, and perform a correction calculation on the ratio to obtain the first error correction data fitness; Calculate the ratio of the preset error correction coding parameter to the first error correction coding parameter to obtain the first error correction redundancy fitness; Calculate the first error correction fitness according to the first error correction data fitness and the first error correction redundancy fitness.

8. The data management method based on optical-magnetic fusion storage according to claim 6, characterized in that Calculate the first erasure fitness of the first erasure code parameter according to the data criticality and the magnetic storage loss parameter, including: Calculate the ratio of the first erasure code parameter to the magnetic storage loss parameter, combine the data criticality, and perform a correction calculation on the ratio to obtain the first erasure data fitness; Calculate the ratio of the preset erasure code parameter to the first erasure code parameter to obtain the first erasure redundancy fitness; Based on the first erasure data fitness and the first erasure redundancy fitness, the first erasure fitness is calculated and obtained.

9. A data management system based on optical-magnetic fusion storage, characterized in that, For implementing the data management method based on optical-magnetic fusion storage according to any one of claims 1 to 7, the system includes: A data acquisition and analysis module, configured to acquire target data to be stored, collect data feature information of the target data, perform data heat analysis, and obtain data heat. A storage decision optimization module, configured to make a decision based on the data heat to obtain a basic storage scheme for optical storage or magnetic storage, perform data criticality analysis according to the data feature information to obtain data criticality, combine the data heat analysis to analyze the heat critical coefficient, and make a fusion storage decision on the target data to obtain a storage scheme. A loss prediction and compensation module, configured to perform data storage access loss prediction according to the data heat and the storage scheme to obtain at least one of an optical storage loss parameter and a magnetic storage loss parameter, and combine the data criticality to perform storage loss compensation optimization to obtain at least one of an error correction coding parameter and an erasure code parameter. An optical-magnetic data management module, configured to perform optical-magnetic storage data management on the target data according to the storage scheme in combination with at least one of the error correction coding parameter and the erasure code parameter.

Images