Cross-cloud platform disaster recovery data migration and recovery method and system

Abstract

The present invention relates to the field of computer technology and provides a cross-cloud platform disaster recovery data migration and recovery method and system thereof. The method comprises: for each type of business data to be disaster recovered in a primary device, determining at least two target cloud platforms for each type of business data based on data characteristic information of each type of business data and platform characteristic information of each cloud platform; transmitting each type of business data to its corresponding at least two target cloud platforms for storage; when a primary device fails, determining the optimal data recovery cloud platform for each type of business data based on platform status information of each of its corresponding at least two target cloud platforms; and restoring each type of business data to a backup device of the primary device based on each optimal data recovery cloud platform. The embodiments of the present invention improve the stability and reliability of enterprise business operations by performing disaster recovery data migration and recovery across cloud platforms.

Description

Technical Field

[0001] The present invention relates to the field of computer technology, and in particular to a cross-cloud platform disaster recovery data migration and recovery method and system thereof. Background Art

[0002] In today's digital age, data is crucial to the operations of businesses and organizations. Disaster recovery and data migration and recovery methods have emerged to address the risks of data loss or system failure. Currently, the most widely adopted disaster recovery and data migration and recovery methods primarily rely on a single cloud platform. This approach can, to a certain extent, ensure data security and recoverability, providing a fundamental guarantee for the normal operation of businesses.

[0003] However, because the entire disaster recovery process relies on a single cloud platform, data migration and recovery may not proceed normally if a serious failure occurs (e.g., a natural disaster or cyberattack). If a problem occurs on the single cloud platform hosting the primary device and disaster recovery server, backup data may not be properly transferred and restored, exposing the enterprise to the risk of data loss and business interruption, impacting the stability and reliability of business operations. Summary of the Invention

[0004] The present invention provides a cross-cloud platform disaster recovery data migration and recovery method and system thereof, for improving the stability and reliability of enterprise business operations.

[0005] In a first aspect, the present invention provides a cross-cloud platform disaster recovery data migration and recovery method, comprising:

[0006] For each type of business data to be restored in the primary device, determine at least two target cloud platforms for each type of business data based on the data characteristics of each type of business data and the platform characteristics of each cloud platform.

[0007] Transmit each type of business data to at least two corresponding target cloud platforms for storage;

[0008] When a master device fails, for each type of business data, the optimal data recovery cloud platform is determined based on the platform status information of each of the at least two corresponding target cloud platforms;

[0009] Restore each type of business data to the backup device of the primary device based on each optimal data recovery cloud platform.

[0010] In a second aspect, the present invention further provides a cross-cloud platform disaster recovery data migration and recovery system, which is applied to the cross-cloud platform disaster recovery data migration and recovery method as described in the first aspect; the cross-cloud platform disaster recovery data migration and recovery system includes:

[0011] A cross-cloud platform determination module is used to determine at least two target cloud platforms for each type of business data to be restored in the primary device based on the data characteristics of each type of business data and the platform characteristics of each cloud platform;

[0012] Disaster recovery data migration module, used to transfer each type of business data to at least two corresponding target cloud platforms for storage;

[0013] A source cloud platform positioning module is used to determine the optimal data recovery cloud platform for each type of business data based on the platform status information of each of the at least two corresponding target cloud platforms when a primary device fails;

[0014] The disaster recovery data recovery module is used to restore each type of business data to a backup device of the primary device based on each optimal data recovery cloud platform.

[0015] In a third aspect, the present invention also provides an electronic device comprising: a memory for storing a computer software program; and a processor for reading and executing the computer software program, thereby implementing a disaster recovery data migration and recovery method across cloud platforms as described in any one of the above.

[0016] In a fourth aspect, the present invention also provides a non-transitory computer-readable storage medium, in which a computer software program is stored. When the computer software program is executed by a processor, it implements a cross-cloud platform disaster recovery data migration and recovery method as described in any one of the above.

[0017] In a fifth aspect, the present invention also provides a computer program product, including a computer program, which, when executed by a processor, implements any of the above-mentioned cross-cloud platform disaster recovery data migration and recovery methods.

[0018] The cross-cloud platform disaster recovery data migration and recovery method provided by the embodiment of the present invention selects at least two target cloud platforms as disaster recovery storage locations for each type of business data, thereby avoiding the risk of business data being centrally stored on a single cloud platform. Even if a cloud platform fails in the subsequent process, the business data can still be recovered from other cloud platforms, thus avoiding the problem of business data loss and improving the reliability of the enterprise's business operations. On the other hand, when the main device fails, the optimal data recovery cloud platform is selected from multiple target cloud platforms based on the platform status information to recover each type of business data, thus avoiding the risk of business interruption, reducing the time during the business data recovery process, and improving the stability of the enterprise's business operations. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 This is a flow chart of a cross-cloud platform disaster recovery data migration and recovery method provided by an embodiment of the present invention;

[0020] Figure 2 This is a structural diagram of a cross-cloud platform disaster recovery data migration and recovery system provided by an embodiment of the present invention;

[0021] Figure 3 An embodiment diagram of an electronic device provided by an embodiment of the present invention;

[0022] Figure 4 An embodiment diagram of a computer-readable storage medium provided for an embodiment of the present invention. DETAILED DESCRIPTION

[0023] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without making any creative efforts shall fall within the scope of protection of the present invention.

[0024] In the description of the present invention, the terms "first" and "second" are used for descriptive purposes only and should not be understood to indicate or imply relative importance or implicitly specify the number of the technical features indicated. Therefore, a feature specified as "first" or "second" may explicitly or implicitly include one or more of the specified features. In the description of the present invention, "plurality" means two or more, unless otherwise specifically defined.

[0025] In the description of the present invention, the term "for example" is used to mean "used as an example, illustration or illustration". Any embodiment of the present invention described as "for example" is not necessarily to be construed as being more preferred or advantageous than other embodiments. The following description is given to enable any person skilled in the art to implement and use the present invention. In the following description, details are listed for the purpose of explanation. It should be understood that a person of ordinary skill in the art can recognize that the present invention can be implemented without using these specific details. In other examples, well-known structures and processes are not elaborated in detail to avoid obscuring the description of the present invention with unnecessary details. Therefore, the present invention is not intended to be limited to the embodiments shown, but is consistent with the widest scope consistent with the principles and features disclosed herein.

[0026] Optional, see Figure 1 , Figure 1 This is a flow chart of a cross-cloud platform disaster recovery data migration and recovery method provided by the present invention. In an embodiment of the present invention, the cross-cloud platform disaster recovery data migration and recovery method is executed by a cross-cloud platform disaster recovery data migration and recovery system. One form of the cross-cloud platform disaster recovery data migration and recovery system is a data management system. Therefore, the cross-cloud platform disaster recovery data migration and recovery method includes:

[0027] Step 10: For each type of business data to be backed up in the primary device, determine at least two target cloud platforms for each type of business data based on data characteristic information of each type of business data and platform characteristic information of each cloud platform.

[0028] Optionally, data characteristic information in embodiments of the present invention includes business type, importance, and change frequency, and platform characteristic information includes storage capacity, read / write speed, and security level. Business type determines the data's application scenario and access mode, importance reflects the impact of data loss or damage on the business, and change frequency reflects the frequency of data updates. Storage capacity determines the amount of data that can be accommodated, read / write speed affects data access efficiency, and security level ensures data storage security.

[0029] Therefore, for each master device, the data management system obtains each type of business data requiring disaster recovery from the master device, including customer information, order data, log files, and so on. Furthermore, based on the data characteristics of each type of business data and the platform characteristics of each cloud platform, the data management system matches at least two target cloud platforms for each type of business data in the cloud platform library.

[0030] In one embodiment, for financial transaction data, the business type is high-real-time transaction processing, which is extremely important and frequently changes. The cloud platform library includes Cloud Platforms A, B, and C, with the following platform characteristics: Cloud Platform A: 100TB storage capacity, 10GB / s read / write speed, and high security level. Cloud Platform B: 50TB storage capacity, 5GB / s read / write speed, and extremely high security level. Cloud Platform C: 200TB storage capacity, 3GB / s read / write speed, and medium security level. First, based on the extremely high importance, cloud platforms with high security levels and above are selected, namely Cloud Platforms A and B. Next, read / write speed is considered. Due to the high real-time requirements of financial transaction data, faster read / write speeds are preferred, and Cloud Platform A has the best read / write speeds. Next, storage capacity is considered. Although the current amount of financial transaction data is relatively small, considering future growth, Cloud Platform C's large capacity makes it a suitable long-term storage option. Ultimately, Cloud Platforms A and C are selected as the target cloud platforms for financial transaction data.

[0031] Step 20: Transmit each type of business data to at least two corresponding target cloud platforms for storage.

[0032] Furthermore, the data management system encapsulates and packages each type of business data according to the interface specification of the target cloud platform, and transmits the encapsulated and packaged business data of each type to the target cloud platform for storage.

[0033] It should be noted that before data transmission, in order to ensure data security, the data management system encrypts each type of business data, and then transmits the encrypted data of each type of business data to at least two target cloud platforms corresponding to each type of business data for storage. The specific data encryption process is described in detail in steps 21 to 23.

[0034] At the same time, during the data transmission process, it is necessary to ensure the integrity and accuracy of the data. Therefore, after receiving the transmitted encrypted data, the target cloud platform decrypts the encrypted data and stores it. The decryption process is the reverse process of the encryption process and will not be described here. After the storage is completed, the target cloud platform verifies the integrity of the data to determine whether the data is damaged during the transmission and storage process, and returns a data transmission feedback signal to the data management system. Therefore, the data management system receives the data transmission feedback signal fed back by the target cloud platform in real time. If the data transmission feedback signal indicates that the data is damaged, each type of business data is re-encrypted and retransmitted to at least two target cloud platforms corresponding to each type of business data for storage.

[0035] Continuing with the example of financial transaction data, the data management system first packages the financial transaction data and encapsulates it according to the interface specifications of Cloud Platform A and Cloud Platform C. Using encrypted channels and multi-threading technology, the data is transmitted to both Cloud Platform A and Cloud Platform C simultaneously.

[0036] Step 30: When a master device fails, for each type of business data, determine the optimal data recovery cloud platform based on the platform status information of each of the at least two corresponding target cloud platforms.

[0037] Optionally, the platform status information in the embodiment of the present invention includes server load status, network delay status and data backup integrity status, wherein the server load status represents the proportion of resources currently used by the server to total resources, affecting the ability of the cloud platform to process data recovery requests. The network delay status represents the average network delay time between the cloud platform and the failed main device, and determines the speed at which data is transmitted back to the backup device. The data backup integrity status represents the difference rate between the backup data and the latest business data, ensuring that the recovered data is the latest and complete. Therefore, when the main device fails, for each type of business data, the data management system evaluates the capabilities of each cloud platform based on the platform status information of each cloud platform in the target cloud platform, obtains the data recovery capability of each cloud platform, and determines the cloud platform with the largest data recovery capability as the optimal data recovery cloud platform for each type of business data, as described in steps 301 to 304.

[0038] Continuing with the above example, the target cloud platforms A and C corresponding to the financial transaction data have the following platform status information:

[0039] Target Cloud Platform A: Server load is 60%, network latency is 50ms, and data backup integrity is 99%. Target Cloud Platform C: Server load is 30%, network latency is 80ms, and data backup integrity is 95%. First, evaluate server load. The lower the load, the greater the processing power, so Target Cloud Platform C has the lowest load. Regarding network latency, the lower the better, so Target Cloud Platform A has the lowest latency. In terms of data backup integrity, Target Cloud Platform A is superior. For example, calculations show that Target Cloud Platform A has higher data recovery capabilities, so Target Cloud Platform A is determined to be the optimal data recovery cloud platform.

[0040] Step 40: Restore each type of business data to a backup device of the primary device based on each optimal data recovery cloud platform.

[0041] Furthermore, for each type of business data, the data management system retrieves the business data from the optimal data recovery cloud platform and restores the business data to a backup device connected to the primary device according to the backup device's data format and storage structure. Of course, the business data can also be restored to the repaired primary device. During the writing process, data verification is required to ensure the accuracy of each block of data. For example, a CRC verification algorithm is used to verify the transmitted data block. If the verification fails, the data block is retransmitted. Once all data is recovered, the backup device immediately takes over the business, ensuring business continuity.

[0042] The embodiment of the present invention selects at least two target cloud platforms as disaster recovery storage locations for each type of business data, avoiding the risk of business data being centrally stored on a single cloud platform. Even if a cloud platform fails in the subsequent process, the business data can still be recovered from other cloud platforms, avoiding the problem of business data loss and improving the reliability of the enterprise's business operations. On the other hand, when the main device fails, the optimal data recovery cloud platform is selected from multiple target cloud platforms based on platform status information to recover each type of business data, avoiding the risk of business interruption, while reducing the time required for business data recovery and improving the stability of the enterprise's business operations.

[0043] In one embodiment, steps 101 to 104 are described as follows:

[0044] Step 101 : Based on the security level of each cloud platform and the business type and importance of each type of business data, a first candidate cloud platform in the cloud platform library is determined.

[0045] Optionally, for each type of business data, the data management system selects cloud platforms with security levels that match the importance from the cloud platform library based on the security level of each cloud platform combined with the business type and importance, and determines them as the first candidate cloud platforms for each type of business data.

[0046] Different business types have different requirements for data security, with more important data requiring higher levels of security. Security levels can be categorized as low, medium, high, and very high.

[0047] In one embodiment, there are three types of business data: financial transaction data (business type: high-real-time transaction processing, extremely important), general office document data (business type: daily office, medium important), and video surveillance data (business type: surveillance storage, low important). The cloud platform database includes Cloud Platform A (high security level), Cloud Platform B (extremely high security level), Cloud Platform C (medium security level), and Cloud Platform D (low security level). For financial transaction data, due to its extremely high importance, only cloud platforms with a security level of high or above meet the requirements. Therefore, Cloud Platform A and Cloud Platform B are the first candidate cloud platforms. For general office document data, due to its medium importance, cloud platforms with a security level of medium or above meet the requirements. Therefore, Cloud Platform A, Cloud Platform B, and Cloud Platform C are the first candidate cloud platforms. For video surveillance data, due to its low importance, cloud platforms with a security level of low or above are all eligible. Therefore, Cloud Platform A, Cloud Platform B, Cloud Platform C, and Cloud Platform D are the first candidate cloud platforms.

[0048] Step 102 : Determine a second candidate cloud platform in the first candidate cloud platform based on matching the read and write speed of each cloud platform with the change frequency of each type of business data.

[0049] Furthermore, based on the first candidate cloud platform, the data management system matches and screens the cloud platform's read and write speeds with the frequency of business data changes to determine a second candidate cloud platform within the first candidate cloud platform. Frequently changing business data requires a cloud platform with fast read and write speeds to ensure timely data updates and access. Read and write speed can be measured by the amount of data read and written per second, while change frequency can be expressed as the number of data updates per unit time.

[0050] Continuing with the above business data as an example, financial transaction data changes 100 times per minute, general office document data changes 10 times per day, and video surveillance data changes once per hour. The first candidate cloud platform A has a read and write speed of 10GB / s, cloud platform B has a read and write speed of 5GB / s, cloud platform C has a read and write speed of 3GB / s, and cloud platform D has a read and write speed of 1GB / s. Set the minimum read and write speed threshold , according to the empirical formula , For financial transaction data, due to its high frequency of change, the faster the read and write speed, the better. The minimum read and write speed threshold is Therefore, for financial transaction data, only cloud platform A meets the requirements, and the second candidate cloud platform is cloud platform A. For ordinary office document data, the minimum read and write speed threshold is , cloud platform A, cloud platform B and cloud platform C all meet the requirements, and the second candidate cloud platforms are cloud platform A, cloud platform B and cloud platform C. For video surveillance data, the minimum read and write speed threshold , cloud platform A, cloud platform B, cloud platform C and cloud platform D all meet the requirements, and the second candidate cloud platforms are cloud platform A, cloud platform B, cloud platform C and cloud platform D.

[0051] Step 103: Filter out a third candidate cloud platform from the second candidate cloud platforms based on the storage capacity of each cloud platform.

[0052] Furthermore, for each type of business data, the data management system estimates the business data volume and analyzes its growth trend based on the current volume and historically projected annual growth rate, thereby determining the storage capacity required for the business data. Furthermore, the data management system selects a third candidate cloud platform from the second candidate cloud platforms that meets the storage capacity requirements for the business data based on the storage capacity of each cloud platform in the second candidate cloud platform, where the storage capacity can be expressed in bytes.

[0053] In one embodiment, the current data volume of financial transaction data is 20TB, which is expected to grow by 20% each year; the current data volume of general office document data is 10TB, which is expected to grow by 10% each year; the current data volume of video surveillance data is 50TB, which is expected to grow by 5% each year. For ease of understanding, the second candidate cloud platform includes cloud platform A, cloud platform B, cloud platform C and cloud platform D as an example. The storage capacity of cloud platform A is 100TB, the storage capacity of cloud platform B is 50TB, the storage capacity of cloud platform C is 30TB, and the storage capacity of cloud platform D is 15TB. For financial transaction data, the data volume is expected to be 3 years later. , Cloud Platform A and Cloud Platform B can both meet the current and next three years' needs. The third candidate cloud platform is Cloud Platform A and Cloud Platform B. For ordinary office document data, the estimated data volume in three years will be , cloud platform A, cloud platform B, cloud platform C and cloud platform D can all meet the demand, and the third candidate cloud platform is cloud platform A, cloud platform B, cloud platform C and cloud platform D. For video surveillance data, the data volume is expected to be , only cloud platform A can meet the requirements, and the third candidate cloud platform is cloud platform A.

[0054] Step 104: Determine at least two target cloud platforms for each type of business data based on the coupling degree of each cloud platform pair in the third candidate cloud platform; the coupling degree of the cloud platform pair is determined based on the business association information and resource sharing interaction information of the cloud platform pair.

[0055] Furthermore, the data management system obtains the business association information and resource sharing interaction information of each cloud platform pair in the third candidate cloud platform, and calculates the coupling degree of each cloud platform pair in the third candidate cloud platform based on the business association information and resource sharing interaction information. The business association information includes whether the business types served by the cloud platforms are similar, and the resource sharing interaction information includes whether there is data sharing and computing resource borrowing between the cloud platforms. The higher the coupling degree, the better the synergy between the two cloud platforms in terms of business and resources. The specific formula for the coupling degree is as follows:

[0056] .in, Indicates the degree of coupling, Indicates business related information. Indicates resource sharing interaction information.

[0057] Furthermore, the data management system determines at least two final target cloud platforms according to the coupling degree of each cloud platform pair in the third candidate cloud platforms, as specifically described in steps 1041 to 1044 .

[0058] Business correlation information includes whether the business types served by the cloud platforms are similar, and resource sharing interaction information includes whether there is data sharing and computing resource borrowing between the cloud platforms. The higher the coupling degree, the better the synergy between the two cloud platforms in terms of business and resources.

[0059] The embodiment of the present invention can accurately match at least two target cloud platforms for each type of business data. The target cloud platforms can better meet the characteristic requirements of business data in terms of security, read and write speed, storage capacity and collaboration, providing a solid foundation for subsequent data transmission, storage and data recovery in the event of a main device failure, improving the reliability and effectiveness of the entire data disaster recovery process, thereby improving the stability and reliability of the enterprise business operation.

[0060] In one embodiment, steps 1041 to 1044 are described as follows:

[0061] Step 1041 , taking any first cloud platform among the third candidate cloud platforms as the initial clustering center, classifying the second cloud platform among the third candidate cloud platforms into the cluster corresponding to the initial clustering center, and obtaining the initial clustering result of the first cloud platform.

[0062] Optionally, the data management system randomly selects a cloud platform from the third candidate cloud platforms as the first cloud platform, sets the first cloud platform as the initial clustering center, and then traverses other cloud platforms (i.e., second cloud platforms) in the third candidate cloud platforms except the first cloud platform. For each second cloud platform, the coupling degree between the second cloud platform and the first cloud platform is calculated.

[0063] Furthermore, if the coupling degree between the second cloud platform and the first cloud platform is less than or equal to a preset coupling threshold, the data management system will classify the second cloud platform into the cluster corresponding to the first cloud platform, and obtain the initial clustering result of the first cloud platform, wherein the coupling degree reflects the closeness of the business association and resource sharing interaction between the first cloud platform and the second cloud platform, and the preset coupling threshold is a standard set according to business needs and experience, which is used to judge whether the association between cloud platforms is close enough to be classified into one category.

[0064] In one embodiment, the third candidate cloud platforms include cloud platform A, cloud platform B, cloud platform C, and cloud platform D. Cloud platform A is the first cloud platform, and the preset coupling threshold is 0.6. The coupling degree between cloud platform B and cloud platform A is calculated to be 0.5, the coupling degree between cloud platform C and cloud platform A is 0.7, and the coupling degree between cloud platform D and cloud platform A is 0.4. Since the coupling degrees between cloud platform B and cloud platform D and cloud platform A are less than or equal to 0.6, the initial clustering result for cloud platform A includes cloud platforms A, B, and D.

[0065] Step 1042: Use the unclustered cloud platform in the third candidate cloud platform as a new cluster center.

[0066] Furthermore, after completing the initial clustering of the first cloud platform, the data management system selects a cloud platform from the third candidate cloud platform that was not included in the first cloud platform clustering and uses it as the new cluster center. Therefore, it can be understood that the unclustered cloud platform is the cloud platform in the third candidate cloud platform excluding the first and second cloud platforms. Furthermore, the data management system uses the unclustered cloud platform as a new clustering starting point to construct another set of clusters, further classifying the cloud platforms and searching for a set of cloud platforms that are different from the first cloud platform cluster but have similar characteristics (as measured by coupling). Continuing with the above example, cloud platform C was not included in the initial clustering of cloud platform A, so cloud platform C was selected as the new cluster center.

[0067] Step 1043: classify the third cloud platform in the second cloud platform into the cluster corresponding to the new cluster center to obtain the first original clustering result of the unclustered cloud platform, and remove the fourth cloud platform in the third cloud platform from the initial clustering result to obtain the second original clustering result of the first cloud platform.

[0068] Furthermore, the data management system traverses the second cloud platform again (i.e., the cloud platforms other than the first cloud platform in the initial clustering of the first cloud platform), and for each cloud platform in the second cloud platform, calculates the coupling degree of each cloud platform with the new cluster center (unclustered cloud platform).

[0069] Furthermore, if there is a third cloud platform in the second cloud platform, where the coupling degree between the third cloud platform and the unclustered cloud platform is less than or equal to a preset coupling threshold, the data management system classifies the third cloud platform into the cluster corresponding to the new clustering center, and obtains the first original clustering result of the unclustered cloud platform.

[0070] Furthermore, for each cloud platform in the third cloud platform, a coupling deviation value between each cloud platform's coupling degree with the first cloud platform and its coupling degree with the unclustered cloud platforms is calculated. If a fourth cloud platform exists in the third cloud platform, and the coupling deviation value between the fourth cloud platform's coupling degree with the first cloud platform and its coupling degree with the unclustered cloud platforms is greater than a preset deviation threshold, the data management system removes the fourth cloud platform from the initial clustering result of the first cloud platform, thereby obtaining a second original clustering result for the first cloud platform. The preset deviation threshold is used to determine whether a cloud platform's tendency to belong between two cluster centers is sufficiently obvious to determine whether its cluster affiliation should be adjusted.

[0071] Continuing with the above example, the preset deviation threshold is 0.2. The coupling degree between cloud platform B and cloud platform C is calculated to be 0.55, and the coupling degree between cloud platform D and cloud platform C is 0.5. Since the coupling degrees of cloud platform B and cloud platform D with cloud platform C are less than or equal to 0.6, the first original clustering result for cloud platform C includes cloud platforms C, B, and D. The coupling degree deviation of cloud platform B's coupling degree with cloud platform A and its coupling degree with cloud platform C is calculated to be |0.5-0.55|=0.05, which is less than 0.2. Cloud platform B remains in the cluster with cloud platform A. The coupling degree deviation of cloud platform D's coupling degree with cloud platform A and its coupling degree with cloud platform C is |0.4-0.5|=0.1, which is less than 0.2. Cloud platform D also remains in the cluster with cloud platform A. Therefore, the second original clustering result for cloud platform A still includes cloud platforms A, B, and D.

[0072] Step 1044: Determine at least two target cloud platforms for each type of business data based on the first original clustering result and the second original clustering result.

[0073] Furthermore, the data management system determines at least two target cloud platforms for each type of business data based on the first original clustering result and the second original clustering result, as specifically described in steps 10441 to 10445.

[0074] The embodiment of the present invention can select at least two cloud platforms that have different coupling characteristics but can meet business needs for each type of business data, so that during the disaster recovery process, business data can be stored and restored based on cloud platforms with different characteristics, thereby improving the flexibility and reliability of the disaster recovery solution, reducing data risks caused by failures or performance problems of a single cloud platform, thereby better ensuring business continuity and data security, and thus improving the stability and reliability of enterprise business operations.

[0075] In one embodiment, steps 10441 to 10445 are described as follows:

[0076] Step 10441: Determine structural characteristic constraints based on the data storage structural characteristics of the first cloud platform and the data storage structural characteristics of the non-clustered cloud platform.

[0077] Optionally, the data storage structure characteristics in the embodiments of the present invention include a storage hierarchy structure, a data access mode, and a data redundancy strategy, including a storage hierarchy structure (such as whether it is multi-level storage, the type and capacity allocation of each level of storage), a data access mode (random access, sequential access, etc.), and a data redundancy strategy (full redundancy, incremental redundancy, etc.).

[0078] Therefore, the data management system compares the differences and commonalities of the data storage structure characteristics of the first cloud platform and the unclustered cloud platform, determines the structural characteristic constraints applicable to the entire data disaster recovery scenario, and ensures that the selected cloud platforms can cooperate with each other in terms of data storage structure to meet the storage and recovery needs of business data.

[0079] In one embodiment, the first cloud platform utilizes a three-tier storage hierarchy: the upper tier comprises high-speed solid-state storage for frequently accessed data, the middle tier comprises mechanical hard disk storage for general data, and the lower tier comprises tape library storage for long-term backup. The data access mode is primarily random access to meet real-time business requirements, and the data redundancy strategy is full redundancy to ensure high data reliability. The unclustered cloud platform utilizes a two-tier storage hierarchy: the upper tier comprises hybrid storage (part solid-state and part mechanical hard disk) and the lower tier comprises cloud storage. The data access mode includes both random and sequential access, and the data redundancy strategy is incremental redundancy. Therefore, after analyzing the differences and commonalities in data storage structural characteristics, the following structural characteristic constraints can be determined: the storage hierarchy must have at least two tiers, one of which must have high read and write speeds to enable fast access to a portion of data; the data access mode must support random access; and the data redundancy strategy must ensure a certain level of data reliability, allowing for either full or incremental redundancy.

[0080] Step 10442: For each original clustering result in the first original clustering result and the second original clustering result, determine the characteristic similarity between any two cloud platforms based on the data storage structure characteristics of any two cloud platforms in the original clustering results.

[0081] Furthermore, for each of the first and second original clustering results, the data management system traverses any two cloud platforms in the original clustering results. Based on the data storage structure characteristics of these two cloud platforms, the similarity between the two cloud platforms is calculated from three aspects: storage hierarchy structure, data access pattern, and data redundancy strategy. The similarity of the storage hierarchy structure can be calculated using a tree edit distance algorithm, a hierarchical clustering algorithm, etc., the similarity of the data access pattern can be calculated using a sequence alignment algorithm, a hidden Markov model (HMM), etc., and the similarity of the data redundancy strategy can be calculated using a set similarity algorithm, an information entropy algorithm, etc.

[0082] Furthermore, for each original clustering result, the data management system calculates the similarity between any two cloud platforms by comprehensively considering the similarity of their storage hierarchical structures, data access patterns, and data redundancy strategies. This gives us a better understanding of the degree of correlation between cloud platforms in terms of data storage structures. The specific formula for feature similarity is as follows:

[0083] .

[0084] in, Indicates feature similarity, Indicates the similarity of the storage hierarchy structure, Indicates the similarity of data access patterns, Indicates the similarity of data redundancy strategies.

[0085] In one embodiment, for any two cloud platforms in the first original clustering result, cloud platform A and cloud platform B. The storage hierarchy structure of cloud platform A is two-layer, with solid state storage on the upper layer and mechanical hard disk on the lower layer; the data access mode is random access; and the data redundancy strategy is full redundancy. The storage hierarchy structure of cloud platform B is two-layer, with hybrid storage on the upper layer and cloud storage on the lower layer; the data access mode is random and sequential access; and the data redundancy strategy is incremental redundancy. According to the corresponding algorithm, for calculating the similarity of the storage hierarchy structure: both are two-layer, and both have high-speed storage layers (the solid state layer of cloud platform A and the solid state part of the hybrid storage of cloud platform B), the similarity is 0.8. Data access mode similarity: both support random access, and the similarity is 0.7. Data redundancy strategy similarity: one is full redundancy and the other is incremental redundancy, and the similarity is 0.5. Therefore, the feature similarity between cloud platform A and cloud platform B is .

[0086] Optionally, embodiments of the present invention may further perform external correlation analysis based on the internal correlation of the data storage structure characteristics of the same cloud platform in different characteristic dimensions, combined with the internal correlation between the data storage structure characteristics of different cloud platforms, to determine the characteristic similarity between any two cloud platforms, specifically:

[0087] The data management system decomposes the data storage structure characteristics into different characteristic dimensions, such as storage hierarchy structure, data access mode, data redundancy strategy, etc. A cloud platform, which is The characteristic value on the characteristic dimension is ,in, , , represents the total number of cloud platforms in each original clustering result, The total number of dimensions that represent the structural characteristics of the data storage.

[0088] Furthermore, the data management system analyzes the internal correlation of the data storage structure characteristics of the same cloud platform in different characteristic dimensions. For any two characteristic dimensions, and , any two feature dimensions and The internal correlation between them is:

[0089] .

[0090] Among them, the internal correlation Characterizes two characteristic dimensions and The degree of mutual influence in the data storage structure of the cloud platform.

[0091] Furthermore, the data management system performs external correlation analysis based on the internal correlation between the data storage structure characteristics of different cloud platforms to determine the characteristic similarity between any two cloud platforms. Therefore, for each original clustering result, A cloud platform and Cloud platform, A cloud platform and The specific formula for the characteristic similarity between cloud platforms is: .

[0092] Step 10443 : Based on the feature association network constructed based on the feature similarity between any two cloud platforms, the cloud platforms in the original clustering result are divided into communities to obtain multiple clustering sub-results.

[0093] Furthermore, for each original clustering result, the data management system constructs a feature association network with the cloud platform as the node and the feature similarity between any two cloud platforms as the edge between the nodes.

[0094] Furthermore, the data management system uses a complex network community partitioning algorithm (such as an improved version of the Louvain algorithm) combined with the feature similarity between any two cloud platforms to perform community partitioning on the feature association network, and divides cloud platforms with higher feature similarity into the same clustering sub-result. This makes the cloud platforms within each clustering sub-result more similar in data storage structure characteristics, while the cloud platform features between different clustering sub-results are relatively different, resulting in multiple clustering sub-results. Therefore, the cloud platforms can be classified more clearly, which facilitates subsequent screening based on structural feature constraints.

[0095] In one embodiment, the first original clustering result includes cloud platforms A, B, C, and D. In the constructed feature association network, the feature similarity between cloud platforms A and B is 0.7, between A and C is 0.5, between A and D is 0.4, between B and C is 0.6, between B and D is 0.5, and between C and D is 0.6. After performing community partitioning using the improved Louvain algorithm, two sub-clustering results are obtained: one containing cloud platforms A and B, and the other containing C and D. This is because the feature similarity between A and B, and between C and D, is relatively high, while the similarity between A and B, and between C and D, is relatively low.

[0096] Step 10444, based on the structural characteristic constraint condition, the cloud platforms in each clustering sub-result are screened to eliminate the cloud platforms whose data storage structural characteristics do not meet the structural characteristic constraint condition in each clustering sub-result, and obtain the optimized clustering result of the original clustering result.

[0097] Furthermore, for each clustering sub-result, the data management system screens each cloud platform one by one according to the structural characteristic constraints determined in step 10441, and checks whether the data storage structural characteristics of each cloud platform meet the structural characteristic constraints. If not, the cloud platform is removed from the clustering sub-result to obtain the optimized result of each clustering sub-result.

[0098] Furthermore, for each clustering sub-result in each original clustering result, the data management system integrates the optimized results of all clustering sub-results to obtain the optimized clustering result of each original clustering result. Therefore, the cloud platform in the optimized clustering result meets the overall requirements in terms of data storage structure characteristics, thereby improving the quality and availability of the clustering results.

[0099] In one embodiment, the structural characteristic constraint condition requires at least two layers, and in a cluster sub-result, there are cloud platforms E and F. Cloud platform E has only one storage hierarchy layer, which does not meet the requirement of at least two layers in the structural characteristic constraint condition. Therefore, cloud platform E is removed from the cluster sub-result. Cloud platform F meets all structural characteristic constraints and is retained in the cluster sub-result.

[0100] Step 10445: Determine the optimized clustering result containing the largest number of cloud platforms as the target clustering result, and determine the cloud platform in the target clustering result as the target cloud platform.

[0101] Furthermore, the data management system compares the number of cloud platforms included in all optimized clustering results, and determines the optimized clustering result including the largest number of cloud platforms as the target clustering result.

[0102] Furthermore, the data management system determines the cloud platforms in the target clustering results as target cloud platforms, wherein the clustering results with the largest number are selected so that the cloud platforms in the clustering results have relatively high commonality and stability while satisfying structural characteristic constraints, and can provide more extensive and reliable disaster recovery support for business data. In one embodiment, there are three optimized clustering results, the first of which includes three cloud platforms, the second includes five cloud platforms, and the third includes two cloud platforms. Therefore, the second optimized clustering result is determined as the target clustering result, and the five cloud platforms in the clustering results are the target cloud platforms.

[0103] The embodiment of the present invention eliminates cloud platforms that do not meet the requirements based on the screening of constraint conditions, thereby improving the quality of clustering results, and then determines the most representative and reliable target cloud platform by selecting the optimized clustering result containing the largest number of cloud platforms as the target clustering result. Overall, it can accurately screen out the cloud platform that is most suitable for business data disaster recovery from the original clustering results, thereby improving the effectiveness and stability of the disaster recovery solution, ensuring the safe storage and efficient recovery of business data in different scenarios, and thus improving the stability and reliability of enterprise business operations.

[0104] In one embodiment, steps 301 to 304 are described as follows:

[0105] Step 301 : Perform adaptability analysis based on the failure type of the master device and the data backup integrity status of each cloud platform in the target cloud platform to determine the adaptability index of each cloud platform in the target cloud platform to the failure type.

[0106] Optionally, the data management system determines the failure type of the main device, where the failure type may include hardware failure, software failure, network failure, etc. Different failure types have different impacts on the integrity of data backup. Therefore, for each cloud platform in the target cloud platform corresponding to each type of business data, the data management system analyzes the adaptability of each cloud platform to various types of failures based on the data backup integrity status of each cloud platform, and determines the adaptability index of each cloud platform to the failure type. The adaptability index characterizes the potential ability of the cloud platform to recover data under a specific failure type. The data backup integrity status can be measured by the difference rate between the backup data and the latest business data. If the difference rate is lower, it means that the backup data is closer to the latest business data, and the adaptability in dealing with failures may be stronger. The specific formula of the adaptability index is as follows:

[0107] .

[0108] in, Indicates the target cloud platform Cloud platform for fault types The adaptability index, Indicates the The data backup integrity status of each cloud platform, represents the number of factors that affect adaptability, Indicates the The weight of the factors, Indicates the fault type and data backup integrity status No. The influence coefficient of each influencing factor, for example, for software failure, It may be the impact coefficient of the integrity of key business data, It may be the influence coefficient of the consistency of data status before and after the failure.

[0109] In one embodiment, a software failure on the master device may cause some data to be incorrectly modified or lost. The target cloud platforms include Cloud Platform A, Cloud Platform B, and Cloud Platform C. The data backup integrity of Cloud Platform A is 99% (i.e., the difference between the backup data and the latest business data is 1%), the data backup integrity of Cloud Platform B is 95%, and the data backup integrity of Cloud Platform C is 90%.

[0110] For software failures, the data management system analysis believes that the higher the data backup integrity, the greater the advantage in recovering data problems caused by software failures. After adaptability analysis, the adaptability index of cloud platform A to software failures is 0.9, the adaptability index of cloud platform B to software failures is 0.7, and the adaptability index of cloud platform C to software failures is 0.5.

[0111] Step 302 : Perform a matching analysis based on the fault severity of the master device and the server load status and network delay status of each cloud platform in the target cloud platform to obtain a matching index of each cloud platform in the target cloud platform to the fault severity.

[0112] Furthermore, the data management system evaluates the severity of the failure of the main device, which can be divided into levels such as minor, moderate, and severe. Therefore, for each cloud platform in the target cloud platform corresponding to each type of business data, the data management system performs a matching analysis based on the severity of the failure combined with the server load status and network delay status of each cloud platform in the target cloud platform, analyzes the support capabilities of each cloud platform in the target cloud platform for business data recovery under different failure severities, and obtains the matching index of each cloud platform in the target cloud platform for the severity of the failure. Among them, the matching index represents the actual processing capacity of the cloud platform under a specific failure severity, and the server load status represents the proportion of the current server's used resources to the total resources. The lower the load, the more capable it may be to cope with additional loads when processing data recovery requests. The network delay status represents the average network delay time between the cloud platform and the failed main device. The lower the delay, the faster the data is transmitted back to the backup device. The specific formula for the matching index is as follows:

[0113] .

[0114] in, Indicates the target cloud platform The matching index of each cloud platform to the severity of the fault, Indicates the severity of the fault. If the fault severity is minor, ; When the fault severity is moderate, ; When the fault severity is serious, ; Indicates the The server load status of each cloud platform, Indicates the The network latency status of each cloud platform.

[0115] In one embodiment, the severity of the master device failure is critical. For critical failures, low server load and network latency are required to ensure rapid data recovery. Cloud platform A's server load is 60%, or 0.6, and its network latency is 50ms; Cloud platform B's server load is 40%, and its network latency is 80ms; and Cloud platform C's server load is 20%, and its network latency is 100ms.

[0116] After matching analysis, it is found that the matching index of cloud platform A for severe faults is 0.6, the matching index of cloud platform B for severe faults is 0.4, and the matching index of cloud platform C for severe faults is 0.2.

[0117] Step 303 : Perform capability evaluation based on the adaptability index and matching index of each cloud platform in the target cloud platform to determine the data recovery capability of each cloud platform in the target cloud platform for the business data of the master device.

[0118] Furthermore, for each target cloud platform corresponding to each type of business data, the management system comprehensively evaluates each cloud platform's data recovery capability for the primary device's business data based on each cloud platform's adaptability index to the fault type and its matching index to the fault severity. The specific formula for data recovery capability is as follows:

[0119] .in, Indicates the target cloud platform The cloud platform's ability to recover the business data of the master device.

[0120] In one embodiment, for cloud platform A, its adaptability index is 0.9 and its matching index is 0.6. Cloud platform B's adaptability index is 0.7 and its matching index is 0.4. Cloud platform C's adaptability index is 0.5 and its matching index is 0.2. Therefore, the calculated data recovery capability of cloud platform A for the primary device's business data is 2.22, cloud platform B's data recovery capability for the primary device's business data is 1.61, and cloud platform C's data recovery capability for the primary device's business data is 1.01.

[0121] Step 304 , traverse the data recovery capability of each cloud platform in the target cloud platform, and determine the cloud platform with the greatest data recovery capability in the target cloud platform as the optimal data recovery cloud platform for each type of business data.

[0122] Furthermore, for each target cloud platform corresponding to each type of business data, the data management system traverses the data recovery capability value of each target cloud platform and determines the cloud platform with the largest data recovery capability value as the optimal data recovery cloud platform for each type of business data. Continuing with the above example, cloud platform A has a data recovery capability of 2.22 for the business data of the master device, cloud platform B has a data recovery capability of 1.61 for the business data of the master device, and cloud platform C has a data recovery capability of 1.01 for the business data of the master device. Therefore, cloud platform A is determined to be the optimal data recovery cloud platform for each type of business data.

[0123] When a main device fails, the embodiment of the present invention can accurately find the most suitable cloud platform for data recovery for each type of business data, minimize the impact of the failure on the business, ensure the integrity of the business data and the continuity of the business, and thus improve the stability and reliability of the enterprise's business operations.

[0124] In one embodiment, steps 21 to 23 are described as follows:

[0125] In step 21, for each type of business data, initial key mapping is performed according to the platform ID information of the target cloud platform to generate the initial encryption key of the target cloud platform, and the initial encryption key is expanded according to the character distribution characteristics of the platform ID information to obtain the expanded encryption key of the target cloud platform.

[0126] Optionally, the platform ID information in the embodiment of the present invention is in the form of a string. Therefore, the platform ID information can be decomposed into multiple characters. Therefore, the platform ID information of the target cloud platform ID can be expressed as ,in, The number of characters representing the platform ID information, Indicates the platform ID information characters, so the ASCII code value of each character in the platform ID information is obtained Therefore, for each target cloud platform corresponding to each type of business data, the data management system performs initial key mapping based on the ASCII code value of each character in the platform ID information to generate the initial encryption key of the target cloud platform. , the specific formula is as follows:

[0127] .

[0128] Furthermore, the data management system obtains the character length of the platform ID information and determines the number of key iterations based on the character length. , where the number of key iterations is , Indicates the length of characters, Indicates a round-down operation. Further, the data management system uses the number of key iterations to Combined with the character distribution characteristics of the platform ID information, key iteration is performed to expand the initial encryption key to obtain the expanded encryption key of the target cloud platform. is a 32-bit binary number, so the initial encryption key It can be expressed as , for each iteration ,Will The key obtained by the iteration Divided into two 16-bit parts and , according to the platform ID information ASCII code value of the character ,calculate , , then and Merge , the final expanded key is The result of the iteration .

[0129] Step 22, divide the business data into multiple data blocks according to the parity of the number of characters in the platform ID information, and perform a permutation operation on the bytes in each data block according to the character information of the characters in the platform ID information combined with the data block index value of each data block to obtain a permuted data block.

[0130] Furthermore, the data management system obtains the parity of the number of characters in the platform ID information, that is, whether the number of characters is odd or even, that is, determines Is it true?

[0131] Furthermore, the data management system determines the data block size according to the parity of the number of characters in the platform ID information, and divides the service data into multiple data blocks according to the data block size, wherein each data block is established with a data block index value related to the platform ID information. In one embodiment, if Established, data block size Bytes, if Not true, data block size Bytes, business data Divide into multiple data blocks ,in, ,in, Represents business data For each data block , its data block index value for .

[0132] Furthermore, the data management system performs a permutation operation on the bytes in each data block according to the character information of the characters in the platform ID information combined with the data block index value of each data block to obtain a permuted data block, wherein the permutation rule combines the character sequence and the value of the platform ID information, so that the byte order of each data block is disrupted and closely associated with the platform ID. In one embodiment, for a data block , which contains the byte sequence , according to the permutation table, the data block Perform a permutation operation on each byte in the data block The permuted data block, where the permutation table For the permutation table Each position in , and its corresponding value The formula is:

[0133] .

[0134] Step 23: Encrypt the replaced data block based on the expanded encryption key to obtain the final encrypted data of the target cloud platform.

[0135] Furthermore, the data management system encrypts the replaced data block based on the extended encryption key to obtain the final encrypted data of the target cloud platform, as specifically described in steps 2301 to 2304.

[0136] The embodiment of the present invention encrypts business data according to the platform ID information of the target cloud platform, thereby enhancing the security of the data transmission process, avoiding intrusion operations such as interception and virus implantation of business data during transmission, and ensuring the security of enterprise business.

[0137] In one embodiment, steps 2301 to 2304 are described as follows:

[0138] Step 2301, summing the code values ​​of the previously preset characters in the platform ID information to obtain a cyclic shift amount, multiplying the code values ​​of all characters in the platform ID information to obtain a mask value, and determining the number of cycles based on the number of characters in the platform ID information.

[0139] Optionally, the data management system sums the code values ​​of the pre-set characters in the platform ID information and determines the cyclic shift amount according to the sum value. In the embodiment of the present invention, the pre-set characters are, for example, 5 bits. Therefore, the sum value can be expressed as , the cyclic shift amount is expressed as .

[0140] Furthermore, the data management system multiplies the code values ​​of all characters in the platform ID information and takes the lower 8 bits of the product value as the mask value.

[0141] Furthermore, the data management system determines the number of cycles based on the number of characters in the platform ID information, where the number of cycles can be expressed as .

[0142] Step 2302: perform a cyclic left shift operation on each byte in the replaced data block according to the cyclic shift amount to obtain a first encrypted data block, and perform a bitwise XOR operation on each byte in the first encrypted data block and the mask value to obtain a second encrypted data block.

[0143] Furthermore, the data management system performs a cyclic left shift operation on each byte in the replaced data block according to the cyclic shift amount to obtain a first encrypted data block.

[0144] Furthermore, the data management system performs a bitwise exclusive OR operation on each byte in the first encrypted data block and the mask value to obtain a second encrypted data block.

[0145] Step 2303: Perform a loop operation on the second encrypted data block according to the number of loops, and shift the bytes in the second encrypted data block backward by one position in each loop to obtain a third encrypted data block.

[0146] Furthermore, the data management system performs a loop operation on the second encrypted data block according to the number of loops, and in each loop, the bytes in the second encrypted data block are moved backward by one position (the last byte is moved to the first position) to obtain a third encrypted data block.

[0147] Step 2304: The extended encryption key and the hash value of the platform ID information are merged to obtain a final encryption key, and the final encryption key is used as a prefix and the third encrypted data block is used as a suffix to obtain final encrypted data.

[0148] Furthermore, the data management system obtains the key information of the platform ID information (such as the combination of the last 4 characters, the combination of the first 4 characters, the combination of the middle 4 characters, etc.), and performs a hash operation on the key information to obtain the hash value of the platform ID information. , where the hash operation algorithm here includes MD5 algorithm, SHA-1 algorithm, SHA-256 algorithm, etc.

[0149] Furthermore, the data management system merges the expanded encryption key and the hash value of the platform ID information to obtain the final encryption key. Therefore, the final encryption key It can be expressed as

[0150] .

[0151] Furthermore, the data management system uses the final encryption key as a prefix and the third encrypted data block as a suffix to perform fusion to obtain the final encrypted data.

[0152] The embodiment of the present invention encrypts business data, thereby enhancing the security of the data transmission process, preventing the business data from being intercepted and infected with viruses during transmission, and ensuring the security of enterprise business.

[0153] Furthermore, the cross-cloud platform disaster recovery data migration and recovery system provided by the present invention is described below. The cross-cloud platform disaster recovery data migration and recovery system described below and the cross-cloud platform disaster recovery data migration and recovery method described above can be referenced to each other.

[0154] Optional, see Figure 2 , Figure 2 This is a structural diagram of the cross-cloud platform disaster recovery data migration and recovery system provided by the present invention. The cross-cloud platform disaster recovery data migration and recovery system includes:

[0155] A cross-cloud platform determination module 210 is configured to determine, for each type of business data to be restored in the primary device, at least two target cloud platforms for each type of business data based on data characteristic information of each type of business data and platform characteristic information of each cloud platform;

[0156] Disaster recovery data migration module 220, used to transfer each type of business data to its corresponding at least two target cloud platforms for storage;

[0157] The source cloud platform positioning module 230 is used to determine the optimal data recovery cloud platform for each type of business data based on the platform status information of each of the at least two corresponding target cloud platforms when a primary device fails;

[0158] The disaster recovery data recovery module 240 is used to restore each type of business data to a backup device of the primary device based on each optimal data recovery cloud platform.

[0159] The embodiment of the present invention selects at least two target cloud platforms as disaster recovery storage locations for each type of business data, avoiding the risk of business data being centrally stored on a single cloud platform. Even if a cloud platform fails in the subsequent process, the business data can still be recovered from other cloud platforms, avoiding the problem of business data loss and improving the reliability of the enterprise's business operations. On the other hand, when the main device fails, the optimal data recovery cloud platform is selected from multiple target cloud platforms based on platform status information to recover each type of business data, avoiding the risk of business interruption, while reducing the time required for business data recovery and improving the stability of the enterprise's business operations.

[0160] See also Figure 3 , Figure 3 This is a diagram of an embodiment of an electronic device provided by an embodiment of the present invention. Figure 3 As shown, an embodiment of the present invention provides an electronic device 300, including a memory 310, a processor 320, and a computer program 311 stored in the memory 310 and executable on the processor 320. When the processor 320 executes the computer program 311, the following steps are implemented:

[0161] For each type of business data to be restored in the primary device, determine at least two target cloud platforms for each type of business data based on the data characteristics of each type of business data and the platform characteristics of each cloud platform.

[0162] Transmit each type of business data to at least two corresponding target cloud platforms for storage;

[0163] When a master device fails, for each type of business data, the optimal data recovery cloud platform is determined based on the platform status information of each of the at least two corresponding target cloud platforms;

[0164] Restore each type of business data to the backup device of the primary device based on each optimal data recovery cloud platform.

[0165] See also Figure 4 , Figure 4 Detailed description of an embodiment of a computer-readable storage medium provided by an embodiment of the present invention. Figure 4 As shown, this embodiment provides a computer-readable storage medium 400 on which a computer program 311 is stored. When the computer program 311 is executed by a processor, the following steps are implemented:

[0166] For each type of business data to be restored in the primary device, determine at least two target cloud platforms for each type of business data based on the data characteristics of each type of business data and the platform characteristics of each cloud platform.

[0167] Transmit each type of business data to at least two corresponding target cloud platforms for storage;

[0168] When a master device fails, for each type of business data, the optimal data recovery cloud platform is determined based on the platform status information of each of the at least two corresponding target cloud platforms;

[0169] Restore each type of business data to the backup device of the primary device based on each optimal data recovery cloud platform.

[0170] In another aspect, the present invention further provides a computer program product, which includes a computer program. The computer program may be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can perform the cross-cloud platform disaster recovery data migration and recovery method provided by the above methods, which includes:

[0171] For each type of business data to be restored in the primary device, determine at least two target cloud platforms for each type of business data based on the data characteristics of each type of business data and the platform characteristics of each cloud platform.

[0172] Transmit each type of business data to at least two corresponding target cloud platforms for storage;

[0173] When a master device fails, for each type of business data, the optimal data recovery cloud platform is determined based on the platform status information of each of the at least two corresponding target cloud platforms;

[0174] Restore each type of business data to the backup device of the primary device based on each optimal data recovery cloud platform.

[0175] The system embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units. They may be located in one place or distributed across multiple network units. Some or all of the modules may be selected based on actual needs to achieve the objectives of this embodiment. Persons of ordinary skill in the art will be able to understand and implement the present invention without inventive effort.

[0176] Through the description of the above embodiments, those skilled in the art will clearly understand that each embodiment can be implemented using software plus the necessary general-purpose hardware cloud platform, or of course, hardware. Based on this understanding, the essence of the above technical solution, or the portion that contributes to the existing technology, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium such as ROM / RAM, a magnetic disk, or an optical disk, and includes a number of instructions for causing a computer device (such as a personal computer, server, or network device) to execute the methods described in each embodiment or certain portions of the embodiments.

[0177] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the various embodiments of the present invention.

Claims

1. A cross-cloud platform disaster recovery data migration and recovery method, characterized in that: include: For each type of business data to be restored in the primary device, determine at least two target cloud platforms for each type of business data based on the data characteristics of each type of business data and the platform characteristics of each cloud platform. Transmit each type of business data to at least two corresponding target cloud platforms for storage; When a master device fails, for each type of business data, the optimal data recovery cloud platform is determined based on the platform status information of each of the at least two corresponding target cloud platforms; Restore each type of business data to the backup device of the primary device based on each optimal data recovery cloud platform; The data characteristic information includes business type, importance and change frequency; the platform characteristic information includes storage capacity, read and write speed and security level; The step of determining at least two target cloud platforms for each type of business data based on the data characteristic information of each type of business data and the platform characteristic information of each cloud platform includes: Based on the security level of each cloud platform and the business type and importance of each type of business data, the first candidate cloud platform in the cloud platform library is determined; Determining a second candidate cloud platform from the first candidate cloud platform based on matching the read and write speed of each cloud platform with the change frequency of each type of business data; Selecting a third candidate cloud platform from the second candidate cloud platforms based on the storage capacity of each cloud platform; Determining at least two target cloud platforms for each type of business data based on the coupling degree of each cloud platform pair in the third candidate cloud platform; the coupling degree of the cloud platform pair is determined based on the business association information and resource sharing interaction information of the cloud platform pair; The determining, based on the coupling degree of each cloud platform pair in the third candidate cloud platforms, at least two target cloud platforms for each type of business data includes: Taking any first cloud platform among the third candidate cloud platforms as an initial clustering center, classifying the second cloud platform among the third candidate cloud platforms into the cluster corresponding to the initial clustering center, and obtaining an initial clustering result of the first cloud platform; Taking the non-clustered cloud platform in the third candidate cloud platform as a new cluster center; Classifying the third cloud platform in the second cloud platform into the cluster corresponding to the new cluster center to obtain a first original clustering result of the unclustered cloud platform, and removing the fourth cloud platform in the third cloud platform from the initial clustering result to obtain a second original clustering result of the first cloud platform; Determining at least two target cloud platforms for each type of business data based on the first original clustering result and the second original clustering result; Among them, the coupling degree between the second cloud platform and the first cloud platform is less than or equal to a preset coupling threshold; the unclustered cloud platform is the cloud platform other than the first cloud platform and the second cloud platform in the third candidate cloud platform; the coupling degree between the third cloud platform and the unclustered cloud platform is less than or equal to the preset coupling threshold; the coupling degree deviation value between the coupling degree of the fourth cloud platform and the first cloud platform and the coupling degree with the unclustered cloud platform is greater than the preset deviation threshold.

2. The cross-cloud platform disaster recovery data migration and recovery method according to claim 1, characterized in that: The determining at least two target cloud platforms for each type of business data based on the first original clustering result and the second original clustering result includes: Determining structural characteristic constraints based on the data storage structural characteristics of the first cloud platform and the data storage structural characteristics of the non-clustered cloud platform; For each original clustering result in the first original clustering result and the second original clustering result, determining characteristic similarity between any two cloud platforms based on data storage structure characteristics of any two cloud platforms in the original clustering results; A feature association network constructed based on the feature similarity between any two cloud platforms divides the cloud platforms in the original clustering results into communities to obtain multiple clustering sub-results; the feature association network is a topological network with cloud platforms as nodes and feature similarities between cloud platforms as edges between nodes; The cloud platforms in each clustering sub-result are screened based on the structural characteristic constraint condition to eliminate cloud platforms whose data storage structural characteristics do not satisfy the structural characteristic constraint condition in each clustering sub-result, thereby obtaining an optimized clustering result of the original clustering result; The optimized clustering result containing the largest number of cloud platforms is determined as the target clustering result, and the cloud platform in the target clustering result is determined as the target cloud platform.

3. The cross-cloud platform disaster recovery data migration and recovery method according to claim 1, characterized in that: Determining the optimal data recovery cloud platform for each type of business data based on the platform status information of each of the at least two target cloud platforms corresponding to each type of business data includes: Based on the failure type of the primary device and the data backup integrity status of each cloud platform in the target cloud platform, an adaptability analysis is performed to determine the adaptability index of each cloud platform in the target cloud platform to the failure type; Based on the fault severity of the main device, a matching analysis is performed on the server load status and network delay status of each cloud platform in the target cloud platform to obtain the matching index of each cloud platform in the target cloud platform for the fault severity. Perform capability assessment based on the adaptability and matching indicators of each cloud platform in the target cloud platform to determine the data recovery capability of each cloud platform in the target cloud platform for the business data of the primary device; The data recovery capability of each cloud platform in the target cloud platform is traversed, and the cloud platform with the largest data recovery capability in the target cloud platform is determined as the optimal data recovery cloud platform for each type of business data.

4. The cross-cloud platform disaster recovery data migration and recovery method according to any one of claims 1 to 3, characterized in that: After determining at least two target cloud platforms for each type of business data, the method further includes: For each type of business data, perform initial key mapping based on the platform ID information of the target cloud platform to generate the initial encryption key of the target cloud platform, and perform key expansion on the initial encryption key based on the character distribution characteristics of the platform ID information to obtain the expanded encryption key of the target cloud platform; Dividing the service data into a plurality of data blocks according to the parity of the number of characters in the platform ID information, and performing a permutation operation on the bytes in each data block according to the character information of the characters in the platform ID information combined with the data block index value of each data block to obtain a permuted data block; The replaced data block is encrypted based on the extended encryption key to obtain the final encrypted data of the target cloud platform.

5. The cross-cloud platform disaster recovery data migration and recovery method according to claim 4, characterized in that: The encrypting the replaced data block based on the extended encryption key to obtain the final encrypted data of the target cloud platform includes: Summing the code values ​​of the pre-set characters in the platform ID information to obtain a cyclic shift amount, multiplying the code values ​​of all characters in the platform ID information to obtain a mask value, and determining the number of cycles based on the number of characters in the platform ID information; Performing a cyclic left shift operation on each byte in the replaced data block according to the cyclic shift amount to obtain a first encrypted data block, and performing a bitwise exclusive OR operation on each byte in the first encrypted data block and the mask value to obtain a second encrypted data block; Performing a loop operation on the second encrypted data block according to the number of loops, shifting bytes in the second encrypted data block backward by one position in each loop, to obtain a third encrypted data block; The extended encryption key and the hash value of the platform ID information are merged to obtain a final encryption key, and the final encryption key is used as a prefix and the third encrypted data block is used as a suffix to obtain the final encrypted data.

6. A cross-cloud platform disaster recovery data migration and recovery system, characterized by: A cross-cloud platform disaster recovery data migration and recovery method applied to any one of claims 1 to 5; The cross-cloud platform disaster recovery data migration and recovery system includes: A cross-cloud platform determination module is used to determine at least two target cloud platforms for each type of business data to be restored in the primary device based on the data characteristics of each type of business data and the platform characteristics of each cloud platform; Disaster recovery data migration module, used to transfer each type of business data to at least two corresponding target cloud platforms for storage; A source cloud platform positioning module is used to determine the optimal data recovery cloud platform for each type of business data based on the platform status information of each of the at least two corresponding target cloud platforms when a primary device fails; Disaster recovery data recovery module, used to restore each type of business data to the backup device of the primary device based on each optimal data recovery cloud platform; The data characteristic information includes business type, importance and change frequency; the platform characteristic information includes storage capacity, read and write speed and security level; The step of determining at least two target cloud platforms for each type of business data based on the data characteristic information of each type of business data and the platform characteristic information of each cloud platform includes: Based on the security level of each cloud platform and the business type and importance of each type of business data, the first candidate cloud platform in the cloud platform library is determined; Determining a second candidate cloud platform from the first candidate cloud platform based on matching the read and write speed of each cloud platform with the change frequency of each type of business data; Selecting a third candidate cloud platform from the second candidate cloud platforms based on the storage capacity of each cloud platform; Determining at least two target cloud platforms for each type of business data based on the coupling degree of each cloud platform pair in the third candidate cloud platform; the coupling degree of the cloud platform pair is determined based on the business association information and resource sharing interaction information of the cloud platform pair; The determining, based on the coupling degree of each cloud platform pair in the third candidate cloud platforms, at least two target cloud platforms for each type of business data includes: Taking any first cloud platform among the third candidate cloud platforms as an initial clustering center, classifying the second cloud platform among the third candidate cloud platforms into the cluster corresponding to the initial clustering center, and obtaining an initial clustering result of the first cloud platform; Taking the non-clustered cloud platform in the third candidate cloud platform as a new cluster center; Classifying the third cloud platform in the second cloud platform into the cluster corresponding to the new cluster center to obtain a first original clustering result of the unclustered cloud platform, and removing the fourth cloud platform in the third cloud platform from the initial clustering result to obtain a second original clustering result of the first cloud platform; Determining at least two target cloud platforms for each type of business data based on the first original clustering result and the second original clustering result; Among them, the coupling degree between the second cloud platform and the first cloud platform is less than or equal to a preset coupling threshold; the unclustered cloud platform is the cloud platform other than the first cloud platform and the second cloud platform in the third candidate cloud platform; the coupling degree between the third cloud platform and the unclustered cloud platform is less than or equal to the preset coupling threshold; the coupling degree deviation value between the coupling degree of the fourth cloud platform and the first cloud platform and the coupling degree with the unclustered cloud platform is greater than the preset deviation threshold.

7. An electronic device comprising: Memory for storing computer software programs; A processor for reading and executing the computer software program, wherein when the processor executes the computer software program, it implements the cross-cloud platform disaster recovery data migration and recovery method as described in any one of claims 1 to 5.

8. A non-transitory computer-readable storage medium storing a computer software program, characterized in that: When the computer software program is executed by a processor, the cross-cloud platform disaster recovery data migration and recovery method as described in any one of claims 1 to 5 is implemented.

Images