Data conflict processing method and device, equipment and medium

Abstract

The invention discloses a data conflict processing method and device, equipment and a medium, and relates to the technical field of object storage. According to the scheme, a conflict detection model used for conflict detection and a conflict processing model used for conflict processing are generated in advance by using historical object storage service operation records; when an object storage service operation record of the data center is collected, conflict detection can be realized through the conflict detection model; when it is confirmed that conflicts exist, the optimal conflict processing strategy can be automatically recommended through the conflict processing model, and therefore the requirement for manual intervention is reduced while the detection accuracy is improved; compared with a traditional rule, the automatic conflict detection and recommendation conflict processing scheme is especially suitable for a complex conflict situation, and the data conflict processing efficiency is improved.

Description

Technical Field

[0001] The present application relates to the field of object storage technology, and in particular to a data conflict processing method, device, equipment and medium. Background Art

[0002] When the Object Storage Service (OSS) is deployed in multiple data centers, conflicts may arise when the same file is updated simultaneously. Although there are some conflict resolution solutions, in the scenario of dual-write in different locations, problems such as concurrent writing of data objects and inconsistent order may still lead to data conflicts. Traditional conflict detection and resolution methods mostly rely on timestamps and rule matching. However, in complex conflict scenarios, the effectiveness of these methods is limited and may cause high false positives or false negatives.

[0003] In addition, traditional conflict resolution strategies are often based on static rules and lack the ability to dynamically adjust and optimize according to actual business scenarios. At the same time, these methods are unable to analyze and predict conflicts and automatically adjust resolution strategies, resulting in the need to improve the level of system intelligence. These shortcomings have jointly restricted the efficiency and accuracy of OSS in handling data conflicts.

[0004] In view of the above, how to solve the low efficiency and lack of accuracy of the current OSS in handling data conflicts is an urgent problem to be solved by technicians in this field. Summary of the invention

[0005] The purpose of this application is to provide a data conflict processing method, device, equipment and medium to solve the problem of low efficiency and insufficient accuracy in processing data conflicts in the current OSS.

[0006] In order to solve the above technical problems, the present application provides a data conflict processing method, comprising:

[0007] Monitor the object storage service operation records of each data center according to the preset period;

[0008] Extracting input features corresponding to each of the object storage service operation records; wherein the input features at least include time series features, operation type features, user behavior features, and object features;

[0009] Input each of the input features into a conflict detection model to output a data conflict probability value; wherein the conflict detection model is a model pre-trained according to the input features and data conflict probability values ​​corresponding to the historical object storage service operation records, and constructed based on the LSTM algorithm and the DNN algorithm;

[0010] When the data conflict probability value is greater than a threshold, the input feature is input into a conflict handling model to output a corresponding conflict handling strategy; wherein the conflict handling model is a neural network model pre-trained according to the input features and handling strategies corresponding to the historical object storage service operation records;

[0011] A target conflict handling strategy is selected from the conflict handling strategies, and the target conflict handling strategy is executed.

[0012] On the one hand, the generation process of the conflict detection model includes:

[0013] Acquire multiple historical object storage service operation records, and convert each of the historical object storage service operation records into a corresponding historical input feature;

[0014] Each of the historical input features is passed into the LSTM layer, and a conflict label is added to the final time step of each of the historical input features in the LSTM layer to output a labeled time series feature corresponding to the historical input feature; wherein each of the historical input features includes multiple time steps;

[0015] The labeled time series features are input into the DNN layer, and feature combination is performed in the DNN layer in combination with the operation type features, user behavior features and object features corresponding to the historical input features to generate the conflict detection model.

[0016] On the other hand, the generation process of the conflict handling model includes:

[0017] Obtain multiple historical object storage service operation records and corresponding historical processing strategies;

[0018] Each of the historical processing strategies is used as label data corresponding to the historical object storage service operation record, and the conflict processing model is generated through supervised learning training.

[0019] On the other hand, the target conflict handling strategy is selected from the conflict handling strategies, including:

[0020] When the data conflict probability value is lower than a first preset value, selecting an automatic processing strategy from the conflict processing strategies as the target conflict processing strategy;

[0021] When the complexity of the current data conflict is higher than a second preset value, a manual review strategy is selected from the conflict handling strategies as the target conflict handling strategy.

[0022] On the other hand, executing the target conflict handling strategy includes:

[0023] When the target conflict handling strategy is the automatic handling strategy, directly executing the automatic handling strategy;

[0024] When the target conflict handling strategy is the manual review strategy, the object storage service operation record is sent to an administrator through a management interface for manual review.

[0025] On the other hand, after executing the target conflict handling strategy, the method further includes:

[0026] The object storage service operation record of the current data conflict is stored in a database; wherein the database contains a plurality of historical object storage service operation records;

[0027] The conflict detection model is iteratively trained according to the historical object storage service operation records in the database.

[0028] On the other hand, after executing the target conflict handling strategy, the method further includes:

[0029] Record the processing operation and processing effect data of this data conflict and store them in the database;

[0030] The conflict handling model is iteratively trained according to the processing operation and the processing effect data.

[0031] In order to solve the above technical problems, the present application also provides a data conflict processing device, including:

[0032] A monitoring module, used to monitor the object storage service operation records of each data center according to a preset period;

[0033] An extraction module, configured to extract input features corresponding to each of the object storage service operation records; wherein the input features at least include time series features, operation type features, user behavior features, and object features;

[0034] A first prediction module, used to input each of the input features into a conflict detection model to output a data conflict probability value; wherein the conflict detection model is a model pre-trained according to the input features and data conflict probability values ​​corresponding to the historical object storage service operation records and constructed based on the LSTM algorithm and the DNN algorithm;

[0035] A second prediction module is used to input the input feature into a conflict processing model to output a corresponding conflict processing strategy when the data conflict probability value is greater than a threshold value; wherein the conflict processing model is a neural network model pre-trained according to the input features and processing strategies corresponding to the historical object storage service operation records;

[0036] The execution module is used to select a target conflict handling strategy from the conflict handling strategies and execute the target conflict handling strategy.

[0037] In order to solve the above technical problems, the present application also provides a data conflict processing device, including:

[0038] Memory for storing computer programs;

[0039] A processor is used to implement the steps of the above-mentioned data conflict handling method when executing the computer program.

[0040] In order to solve the above technical problem, the present application also provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps of the above data conflict handling method are implemented.

[0041] The data conflict handling method provided in the present application specifically monitors the object storage service operation records of each data center according to a preset period; extracts the input features corresponding to each object storage service operation record; wherein the input features at least include time series features, operation type features, user behavior features and object features; inputs each input feature into a conflict detection model to output a data conflict probability value; wherein the conflict detection model is a model pre-trained according to the input features and data conflict probability values ​​corresponding to the historical object storage service operation records, and constructed based on the LSTM algorithm and the DNN algorithm; when the data conflict probability value is greater than a threshold value, inputs the input features into a conflict handling model to output a corresponding conflict handling strategy; wherein the conflict handling model is a neural network model pre-trained according to the input features and handling strategies corresponding to the historical object storage service operation records; selects a target conflict handling strategy from the conflict handling strategies, and executes the target conflict handling strategy.

[0042] The beneficial effects of the present application are that a conflict detection model for conflict detection and a conflict handling model for conflict handling are pre-generated using historical object storage service operation records; when object storage service operation records are collected from a data center, conflict detection can be implemented through the conflict detection model; when a conflict is confirmed to exist, the conflict handling model can automatically recommend the best conflict handling strategy, thereby improving detection accuracy while reducing the need for manual intervention; compared with traditional rules, the automatic conflict detection and recommended conflict handling solutions of this solution are particularly suitable for complex conflict scenarios, thereby improving the efficiency of data conflict handling.

[0043] In addition, the present application also provides a data conflict processing device, equipment and medium, with the same effect as above. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] In order to more clearly illustrate the embodiments of the present application, the following is a brief introduction to the drawings required for use in the embodiments. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0045] Figure 1 A flowchart of a data conflict handling method provided in an embodiment of the present application;

[0046] Figure 2 A schematic diagram of a data conflict processing device provided in an embodiment of the present application;

[0047] Figure 3 A schematic diagram of a data conflict processing device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0048] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0049] The core of this application is to provide a data conflict processing method, device, equipment and medium to solve the problem of low efficiency and insufficient accuracy in processing data conflicts in the current OSS.

[0050] In order to enable those skilled in the art to better understand the present application, the present application is further described in detail below in conjunction with the accompanying drawings and specific implementation methods.

[0051] Currently, in remote dual-write scenarios, problems such as concurrent writing and inconsistent order of data objects may still lead to data conflicts. Traditional conflict detection and processing methods mostly rely on timestamps and rule matching. However, in complex conflict scenarios, the effectiveness of these methods is limited and may cause high false positives or false negatives. In addition, traditional conflict resolution strategies are often based on static rules and lack the ability to dynamically adjust and optimize according to actual business scenarios. At the same time, these methods cannot analyze and predict conflicts, so as to automatically adjust the resolution strategy, resulting in the need to improve the level of system intelligence. In view of the above problems, the present application provides a data conflict processing method.

[0052] Figure 1 Flow chart of a data conflict handling method provided in an embodiment of the present application. Figure 1 As shown, the method includes:

[0053] S10: Monitor the object storage service operation records of each data center according to a preset period.

[0054] This solution detects the OSS operation records of the data center through a time window sliding, and specifically monitors the OSS operation records of each data center according to a preset period. It should be noted that in this embodiment, there is no restriction on the preset period, for example, it can be 1 minute.

[0055] On the other hand, OSS operation records specifically refer to data change operations performed by the data center, such as data write operations or data update operations. By collecting multi-dimensional operation data in OSS, including user behavior, object state changes, operation intervals, etc., and storing these data in a highly scalable distributed database, it supports subsequent feature extraction and real-time detection needs.

[0056] S11: Extract input features corresponding to each object storage service operation record.

[0057] The input features include at least time series features, operation type features, user behavior features and object features.

[0058] Furthermore, the input features corresponding to each object storage service operation record are extracted. It should be noted that the input features at least include time series features, operation type features, user behavior features and object features. The following is a detailed description of each input feature:

[0059] Timing features include the interval between operations and the frequency of operations on each object. It is understandable that frequent operations in a short period of time usually mean potential conflict risks, and calculating the operation interval can help the model identify these risks. For dual writes in different locations, the operation interval can reflect the latency difference between write operations in two data centers. By monitoring the operation intervals across centers, the system can determine which operations are prone to conflicts.

[0060] The operation type feature is specifically the operation type recorded in chronological order, such as read, write, delete, etc. Different combinations of operation types may cause conflicts. In dual-write scenarios, the operation type is particularly important because when writing to the same object concurrently, the conflict risk of different write types is different, which may lead to serious data inconsistency.

[0061] User behavior characteristics represent the user's operation mode on an object, such as multiple consecutive write operations. If the dual-center environment has different user groups or applications generating operations, understanding the write frequency and behavior patterns of these users or applications will help better detect and predict conflicts. For example, you can preset user groups with high conflict risks for priority detection.

[0062] Finally, object characteristics are properties associated with the object, such as file type, data priority, business criticality, etc.

[0063] S12: Input each input feature into the conflict detection model to output a data conflict probability value.

[0064] Among them, the conflict detection model is a model that is pre-trained according to the input features and data conflict probability values ​​corresponding to the historical object storage service operation records, and is built based on the LSTM algorithm and DNN algorithm.

[0065] Furthermore, each input feature is input into the conflict detection model to output a data conflict probability value. It should be noted that the conflict detection model is a model that is pre-trained based on the input features and data conflict probability values ​​corresponding to the historical object storage service operation records, and is constructed based on the Long Short-Term Memory (LSTM) algorithm and the Deep Neural Network (DNN) algorithm.

[0066] The LSTM algorithm is an important time series algorithm and a special recurrent neural network (RNN) that can learn long-term dependencies. DNN is a machine learning technology that can complete tasks that are difficult to complete with traditional programming technology by training computers. In this solution, in order to achieve conflict prediction, the conflict detection model is mainly trained and generated through LSTM and DNN algorithms. During the detection process, features are input into the model so that the model outputs the corresponding detection results; specifically, the output result of the LSTM combined with the DNN layer is the probability value of the conflict risk. In this embodiment, there is no restriction on the specific training process of the conflict detection model, which depends on the specific implementation situation.

[0067] S13: When the data conflict probability value is greater than a threshold, the input feature is input into a conflict handling model to output a corresponding conflict handling strategy.

[0068] Among them, the conflict handling model is a neural network model that is pre-trained based on the input features and processing strategies corresponding to the historical object storage service operation records.

[0069] After obtaining the data conflict probability value output by the conflict detection model, determine whether the data conflict probability value is greater than the threshold. If not, it is considered that there will be no data conflict at present, and the current process ends. If so, it is considered that there is a data conflict, and the input features need to be input into the conflict handling model to output the corresponding conflict handling strategy. In this embodiment, there is no restriction on the threshold value.

[0070] It should be noted that the conflict handling model is a neural network model pre-trained according to the input features and handling strategies corresponding to the historical object storage service operation records. In this embodiment, there is no restriction on the generation process of the conflict handling model, which depends on the specific implementation situation.

[0071] S14: Select a target conflict handling strategy from the conflict handling strategies, and execute the target conflict handling strategy.

[0072] In this embodiment, the number of conflict handling strategies output by the conflict handling model can be one or more. In order to better handle data conflicts, after obtaining the conflict handling strategies, a target conflict handling strategy that is more suitable for this data conflict can be selected. The target conflict handling strategy is executed to achieve data conflict handling. It should be noted that in this embodiment, there is no restriction on the selection method of the target conflict handling strategy, which depends on the specific implementation situation.

[0073] In this embodiment, a conflict detection model for conflict detection and a conflict handling model for conflict handling are pre-generated using historical object storage service operation records; when object storage service operation records of the data center are collected, conflict detection can be implemented through the conflict detection model; when a conflict is confirmed to exist, the conflict handling model can automatically recommend the best conflict handling strategy, thereby improving detection accuracy while reducing the need for manual intervention; compared with traditional rules, the automatic conflict detection and recommended conflict handling solutions of this solution are particularly suitable for complex conflict scenarios, thereby improving the efficiency of data conflict handling.

[0074] Based on the above embodiments, in some embodiments, the generation process of the conflict detection model includes:

[0075] S101: Acquire multiple historical object storage service operation records, and convert each historical object storage service operation record into a corresponding historical input feature;

[0076] S102: passing each historical input feature into the LSTM layer, and adding a conflict label to the final time step of each historical input feature in the LSTM layer, so as to output a labeled time series feature corresponding to the historical input feature; wherein each historical input feature includes multiple time steps;

[0077] S103: Input the labeled time series features into the DNN layer, and perform feature combination in the DNN layer in combination with the operation type features, user behavior features, and object features corresponding to the historical input features to generate a conflict detection model.

[0078] In order to generate a conflict detection model, in this embodiment, multiple historical OSS operation records are specifically obtained, and each historical OSS operation record is converted into a corresponding historical input feature. It can be understood that the historical input feature is a conflict-related feature, including historical time series features, operation type features, user behavior features, and object features.

[0079] Each historical input feature is further passed to the LSTM layer. It should be noted that the LSTM layer is responsible for capturing the temporal dependency characteristics of the operation. LSTM can capture the temporal dependency of the operation, and DNN is used for further modeling of multi-dimensional nonlinear features. This combination can better handle complex temporal and multi-dimensional relationships in conflict detection. The LSTM network can remember the previous and next dependencies, enabling the model to capture cross-time conflict risks. Each historical input feature includes an operation record within a time window, and LSTM transforms the input operation sequence into a temporal feature. Specifically, in the LSTM layer, a conflict label is added to the final time step of each historical input feature, with conflicts as positive samples and non-conflicts as negative samples. The definition of conflict can mark known conflicts based on historical data, helping the model to distinguish conflict risks in different situations. Finally, the labeled temporal features corresponding to the historical input features are output. It should also be noted that each historical input feature contains multiple time steps.

[0080] Furthermore, the DNN layer is used to enhance the nonlinear modeling of features. After the time series features are output by LSTM, the labeled time series features are input to the DNN layer, and the operation type features, user behavior features, and object features corresponding to the historical input features are combined in the DNN layer. Based on the LSTM output, the DNN learns deeper patterns and nonlinear relationships, and finally generates a conflict detection model.

[0081] Based on the above embodiments, in some embodiments, the generation process of the conflict handling model includes:

[0082] S111: Acquire multiple historical object storage service operation records and corresponding historical processing strategies;

[0083] S112: Using each historical processing strategy as label data of the corresponding historical object storage service operation record, and generating a conflict processing model through supervised learning training.

[0084] In order to generate a conflict handling model, we first obtain multiple historical OSS operation records and corresponding historical handling strategies. It should be noted that the historical OSS operation records contain conflict-related features, including historical time series features, operation type features, user behavior features, and object features; the historical handling strategies contain historical conflict handling records and effect data, which are used for model training and recommendation strategy optimization.

[0085] Furthermore, each historical processing strategy is used as the label data of the corresponding historical OSS operation record, and the model is trained through supervised learning, so that it can accurately match the conflict type and solution, and finally generate a conflict processing model.

[0086] Based on the above embodiments, in some embodiments, selecting a target conflict handling strategy from the conflict handling strategies includes:

[0087] S121: When the data conflict probability value is lower than a first preset value, selecting an automatic processing strategy from the conflict processing strategies as a target conflict processing strategy;

[0088] S122: When the complexity of the current data conflict is higher than a second preset value, a manual review strategy is selected as a target conflict handling strategy from the conflict handling strategies.

[0089] After obtaining the conflict handling strategy output by the model, the data conflict risk and conflict complexity of this time can be further determined. Specifically, when the data conflict risk is lower than the first preset value, or when a mature solution exists, the automated processing strategy can be selected as the target conflict handling strategy in the conflict handling strategy. In this embodiment, there is no restriction on the specific content of the automated processing strategy, such as priority writing, version retention, etc. Correspondingly, when the target conflict handling strategy is the automated processing strategy, the automated processing strategy can be directly executed. In this embodiment, there is no restriction on the first preset value.

[0090] On the other hand, when the complexity of this data conflict is higher than the second preset value, the manual review strategy is selected as the target conflict handling strategy in the conflict handling strategy. In this embodiment, there is no restriction on the second preset value. Correspondingly, when the target conflict handling strategy is the manual review strategy, a manual review notification is triggered, and the OSS operation record is sent to the administrator through the management interface, and the administrator selects a processing solution.

[0091] In addition, in order to improve the accuracy of model prediction, based on the above embodiments, in some embodiments, after executing the target conflict handling strategy, the following is further included:

[0092] S131: storing the object storage service operation record of the current data conflict in a database; wherein the database includes a plurality of historical object storage service operation records;

[0093] S132: Iteratively train the conflict detection model according to the historical object storage service operation records in the database.

[0094] Specifically, the OSS operation record of this data conflict is stored in the database. It should be noted that the database contains multiple historical OSS operation records. The conflict detection model is further iteratively trained based on the historical OSS operation records in the database, and the adaptability is improved by periodically fine-tuning the model, and the model is retrained regularly to gradually optimize the conflict detection effect. It can be understood that the model training process is the same as the model generation training process in the above embodiment.

[0095] Correspondingly, after executing the target conflict handling strategy, it also includes:

[0096] S133: Record the processing operation and processing effect data of this data conflict and store them in the database;

[0097] S134: Iteratively train the conflict processing model according to the processing operation and processing effect data.

[0098] Specifically, the processing operations and processing effect data of this data conflict are recorded and stored in the database, and the conflict processing model is iteratively trained based on the processing operations and processing effect data. By dynamically adjusting the decision weights of the model, the accuracy of conflict resolution and system consistency are continuously optimized.

[0099] In the above embodiments, the data conflict processing method is described in detail, and the present application also provides a corresponding embodiment of the data conflict processing device.

[0100] Figure 2 Schematic diagram of a data conflict processing device provided in an embodiment of the present application. Figure 2 As shown, the device comprises:

[0101] The monitoring module 10 is used to monitor the object storage service operation records of each data center according to a preset period.

[0102] The extraction module 11 is used to extract input features corresponding to each object storage service operation record; wherein the input features at least include time series features, operation type features, user behavior features and object features.

[0103] The first prediction module 12 is used to input each input feature into the conflict detection model to output a data conflict probability value; wherein the conflict detection model is a model pre-trained according to the input features and data conflict probability values ​​corresponding to the historical object storage service operation records, and constructed based on the LSTM algorithm and the DNN algorithm.

[0104] The second prediction module 13 is used to input the input features into the conflict handling model to output the corresponding conflict handling strategy when the data conflict probability value is greater than the threshold; wherein the conflict handling model is a neural network model pre-trained according to the input features and handling strategies corresponding to the historical object storage service operation records.

[0105] The execution module 14 is used to select a target conflict handling strategy from the conflict handling strategies and execute the target conflict handling strategy.

[0106] Since the embodiments of the apparatus part correspond to the embodiments of the method part, please refer to the description of the embodiments of the method part for the embodiments of the apparatus part, which will not be repeated here.

[0107] Figure 3 A schematic diagram of a data conflict processing device provided in an embodiment of the present application. Figure 3 As shown, the data conflict processing device includes:

[0108] A memory 20, used for storing computer programs;

[0109] The processor 21 is used to implement the steps of the data conflict handling method mentioned in the above embodiment when executing a computer program.

[0110] The data conflict processing device provided in this embodiment may include but is not limited to a smart phone, a tablet computer, a laptop computer, or a desktop computer.

[0111] Among them, the processor 21 may include one or more processing cores, such as a 4-core processor, an 8-core processor, etc. The processor 21 can be implemented in at least one hardware form of a digital signal processor (DSP), a field programmable gate array (FPGA), and a programmable logic array (PLA). The processor 21 may also include a main processor and a coprocessor. The main processor is a processor for processing data in an awake state, also known as a central processing unit (CPU); the coprocessor is a low-power processor for processing data in a standby state. In some embodiments, the processor 21 may be integrated with a graphics processing unit (GPU), and the GPU is responsible for rendering and drawing the content to be displayed on the display screen. In some embodiments, the processor 21 may also include an artificial intelligence (AI) processor, which is used to process computing operations related to machine learning.

[0112] The memory 20 may include one or more computer-readable storage media, which may be non-transitory. The memory 20 may also include a high-speed random access memory, and a non-volatile memory, such as one or more disk storage devices, flash memory storage devices. In this embodiment, the memory 20 is at least used to store the following computer program 201, wherein, after the computer program is loaded and executed by the processor 21, it can implement the relevant steps of the data conflict handling method disclosed in any of the aforementioned embodiments. In addition, the resources stored in the memory 20 may also include an operating system 202 and data 203, etc., and the storage method may be temporary storage or permanent storage. Among them, the operating system 202 may include Windows, Unix, Linux, etc. Data 203 may include but is not limited to data involved in the data conflict handling method.

[0113] In some embodiments, the data conflict processing device may further include a display screen 22 , an input / output interface 23 , a communication interface 24 , a power supply 25 , and a communication bus 26 .

[0114] Those skilled in the art will understand that Figure 3 The structure shown in the figure does not constitute a limitation on the data conflict processing device, and may include more or less components than those shown in the figure.

[0115] Finally, the present application also provides an embodiment corresponding to a computer-readable storage medium. The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps recorded in the above method embodiment are implemented.

[0116] It is understandable that if the method in the above embodiment is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application is essentially or the part that contributes to the prior art or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium to execute all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM), random access memory (RAM), disk or optical disk and other media that can store program code.

[0117] The above is a detailed introduction to a data conflict processing method, device, equipment and medium provided by the present application. The various embodiments in the specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same and similar parts between the various embodiments can be referenced to each other. For the device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the method part description. It should be pointed out that for ordinary technicians in this technical field, without departing from the principles of the present application, several improvements and modifications can be made to the present application, and these improvements and modifications also fall within the scope of protection of the present application.

[0118] It should also be noted that, in this specification, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "comprise", "include" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, an element defined by the statement "comprises a ..." does not exclude the presence of other identical elements in the process, method, article or device including the element.

Claims

1. A data conflict processing method, characterized in that: include: Monitor the object storage service operation records of each data center according to the preset period; Extracting input features corresponding to each of the object storage service operation records; wherein the input features at least include time series features, operation type features, user behavior features, and object features; Input each of the input features into a conflict detection model to output a data conflict probability value; wherein the conflict detection model is a model pre-trained according to the input features and data conflict probability values ​​corresponding to the historical object storage service operation records, and constructed based on the LSTM algorithm and the DNN algorithm; When the data conflict probability value is greater than a threshold, the input feature is input into a conflict handling model to output a corresponding conflict handling strategy; wherein the conflict handling model is a neural network model pre-trained according to the input features and handling strategies corresponding to the historical object storage service operation records; A target conflict handling strategy is selected from the conflict handling strategies, and the target conflict handling strategy is executed.

2. The data conflict handling method according to claim 1, characterized in that: The generation process of the conflict detection model includes: Acquire multiple historical object storage service operation records, and convert each of the historical object storage service operation records into a corresponding historical input feature; Each of the historical input features is passed into the LSTM layer, and a conflict label is added to the final time step of each of the historical input features in the LSTM layer to output a labeled time series feature corresponding to the historical input feature; wherein each of the historical input features includes multiple time steps; The labeled time series features are input into the DNN layer, and feature combination is performed in the DNN layer in combination with the operation type features, user behavior features and object features corresponding to the historical input features to generate the conflict detection model.

3. The data conflict handling method according to claim 1, characterized in that: The generation process of the conflict handling model includes: Obtain multiple historical object storage service operation records and corresponding historical processing strategies; Each of the historical processing strategies is used as label data corresponding to the historical object storage service operation record, and the conflict processing model is generated through supervised learning training.

4. The data conflict handling method according to claim 1, characterized in that: Select the target conflict handling strategy in the conflict handling strategy, including: When the data conflict probability value is lower than a first preset value, selecting an automatic processing strategy from the conflict processing strategies as the target conflict processing strategy; When the complexity of the current data conflict is higher than a second preset value, a manual review strategy is selected from the conflict handling strategies as the target conflict handling strategy.

5. The data conflict handling method according to claim 4, characterized in that: Executing the target conflict handling strategy includes: When the target conflict handling strategy is the automatic handling strategy, directly executing the automatic handling strategy; When the target conflict handling strategy is the manual review strategy, the object storage service operation record is sent to an administrator through a management interface for manual review.

6. The data conflict handling method according to any one of claims 1 to 5, characterized in that: After executing the target conflict handling strategy, the method further includes: The object storage service operation record of the current data conflict is stored in a database; wherein the database contains a plurality of historical object storage service operation records; The conflict detection model is iteratively trained according to the historical object storage service operation records in the database.

7. The data conflict handling method according to claim 6, characterized in that: After executing the target conflict handling strategy, the method further includes: Record the processing operation and processing effect data of this data conflict and store them in the database; The conflict handling model is iteratively trained according to the processing operation and the processing effect data.

8. A data conflict processing device, characterized in that: include: A monitoring module, used to monitor the object storage service operation records of each data center according to a preset period; An extraction module, configured to extract input features corresponding to each of the object storage service operation records; wherein the input features at least include time series features, operation type features, user behavior features, and object features; A first prediction module, used to input each of the input features into a conflict detection model to output a data conflict probability value; wherein the conflict detection model is a model pre-trained according to the input features and data conflict probability values ​​corresponding to the historical object storage service operation records and constructed based on the LSTM algorithm and the DNN algorithm; A second prediction module is used to input the input feature into a conflict processing model to output a corresponding conflict processing strategy when the data conflict probability value is greater than a threshold value; wherein the conflict processing model is a neural network model pre-trained according to the input features and processing strategies corresponding to the historical object storage service operation records; The execution module is used to select a target conflict handling strategy from the conflict handling strategies and execute the target conflict handling strategy.

9. A data conflict processing device, characterized in that: include: Memory for storing computer programs; A processor, configured to implement the steps of the data conflict handling method according to any one of claims 1 to 7 when executing the computer program.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the data conflict handling method according to any one of claims 1 to 7 are implemented.

Images