Data disaster recovery processing method and device, storage medium and electronic equipment

Through the target model, the detection data of multiple data types is processed, faulty components are identified and isolated, and the problems of slow response speed and insufficient recovery capabilities in the prior art are solved, and fast response and efficient recovery are achieved.

CN118916205APending Publication Date: 2024-11-08STATE GRID BEIJING ELECTRIC POWER CO +1
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411066473.1
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-05
Publication Date
2024-11-08

AI Technical Summary

Technical Problem

The prior art responds slowly when dealing with burst hardware failures, software configuration errors, network problems or security attacks, making it difficult to quickly and accurately identify faults and effectively recover, resulting in business interruptions and data loss.

Method used

The target model is used to process detection data of multiple data types, extract features through the encoder and identify faults by the decoder, and transfer them to an isolated environment based on the dependence between the faulty component and other components to restore, ensuring the continuity of the data and Integrity.

Benefits of technology

It improves the system's rapid response ability and recovery efficiency to faults, realizes accurate identification and isolation of fault components when disaster recovery identification occurs, and quickly recovers data, solving the problem of unsatisfactory data response efficiency.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118916205A_ABST
    Figure CN118916205A_ABST
Patent Text Reader

Abstract

The invention discloses a data disaster recovery processing method and device, a storage medium and electronic equipment. The method comprises the steps that a target model is adopted, detection data of multiple data types of a target system is processed, disaster recovery recognition results corresponding to the multiple data types are determined, and an encoder and a decoder are correspondingly arranged in the target model according to the detection data of the multiple data types; the encoder is used for extracting features of the detection data of the corresponding data type, and the decoder is used for identifying whether the detection data of the corresponding data type has a fault or not; under the condition that the disaster recovery identification result indicates that abnormity exists, the fault component is transferred to an isolation environment based on the dependency relationship between the fault component existing in the target system and other components; and recovering the fault component in the isolation environment to obtain a recovered target system. The technical problem that the data fault response efficiency is not ideal in the prior art is solved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of network security and artificial intelligence, and in particular to a data disaster recovery processing method, device, storage medium and electronic equipment. Background Art

[0002] At present, enterprise operations are highly dependent on the stability and security of information systems. Traditional disaster recovery management methods usually rely on preset static strategies, such as regular data backup and mirror servers, which can provide basic data protection in a static environment. However, with the continuous expansion of system scale and the increasing complexity of the technical environment, related technologies have shown obvious deficiencies in dealing with sudden hardware failures, software configuration errors, network problems or security attacks. There are problems such as slow response speed, lack of real-time performance, and difficulty in quickly and accurately identifying faults and performing effective recovery, which may lead to business interruption and data loss, causing losses to the enterprise.

[0003] Related technologies often show slow response speed and insufficient recovery capabilities when dealing with dynamic and complex system failures. Especially when faced with sudden hardware failures, software configuration errors, network problems or security attacks, it is difficult to identify and recover quickly in real time, which may lead to business interruption and data loss.

[0004] To address the above-mentioned problems, no effective solution has been proposed yet. Summary of the invention

[0005] The embodiments of the present invention provide a data disaster recovery processing method, device, storage medium and electronic device to at least solve the technical problem of unsatisfactory data failure response efficiency in the related art.

[0006] According to one aspect of an embodiment of the present invention, a data disaster recovery processing method is provided, comprising: using a target model to process detection data of multiple data types of a target system, and determining disaster recovery identification results corresponding to the multiple data types, wherein encoders and decoders are respectively set in the target model according to the detection data of the multiple data types, the encoder is used to extract features of the detection data of the corresponding data type, and the decoder is used to identify whether there is a fault in the detection data of the corresponding data type; when the disaster recovery identification result indicates that there is an abnormality, based on the dependency relationship between the faulty component and other components in the target system, the faulty component is transferred to an isolated environment, wherein the faulty component corresponds to the data type whose disaster recovery identification result is abnormal among the multiple data types; and the faulty component is restored in the isolated environment to obtain a restored target system.

[0007] Optionally, the target model is used to process detection data of multiple data types of the target system to determine disaster recovery identification results corresponding to the multiple data types, including: based on the detection data corresponding to the multiple data types, inputting into multiple encoders included in the target model for processing to obtain encoding features corresponding to the multiple data types; based on the encoding features corresponding to the multiple data types, using the implicit feature layer included in the target model for processing to obtain association features between the multiple data types; based on the association features, using multiple decoders included in the target model for processing to obtain reconstructed data corresponding to the multiple data types; based on the reconstructed data corresponding to the multiple data types, determining disaster recovery identification results corresponding to the multiple data types.

[0008] Optionally, the detection data corresponding to the multiple data types are input into a plurality of encoders included in the target model for processing to obtain encoding features corresponding to the multiple data types, including: based on the detection data corresponding to the multiple data types, the detection data are input into a plurality of input sublayers included in the target model for processing to obtain input features corresponding to the multiple data types, wherein the multiple input sublayers correspond one-to-one to the multiple data types; based on the input features corresponding to the multiple data types, the detection data are input into a plurality of encoders included in the target model for processing to obtain encoding features corresponding to the multiple data types.

[0009] Optionally, the method further includes: determining, based on the dependency relationship, an impacted component among the other components that is impacted by the faulty component; and restarting the impacted component using a predetermined configuration strategy.

[0010] Optionally, recovering the faulty component in the isolated environment to obtain a recovered target system includes: in the isolated environment, using backup data before the exception occurs to recover the faulty component to obtain a recovery component; remapping data connections, updating system configurations, and restoring user access, performing integration processing based on the recovery component to obtain the recovered target system.

[0011] Optionally, the method of remapping data connections, updating system configurations, and restoring user access is used to perform integration processing based on the recovery component to obtain the restored target system, including: performing a data consistency check on the recovery component based on a predetermined foreign key relationship and a predetermined data format, wherein the predetermined foreign key relationship is used to constrain key references to be in a valid state; when the recovery component passes the data consistency check, the method of remapping data connections, updating system configurations, and restoring user access is used to perform integration processing based on the recovery component to obtain the restored target system.

[0012] Optionally, the method further includes: determining the fault property of the faulty component and the fault impact caused by the faulty component; and reconfiguring resource distribution of the recovered target system based on the fault property and the fault impact.

[0013] According to another aspect of an embodiment of the present invention, a data disaster recovery processing device is provided, comprising: a detection and identification module, which uses a target model to process detection data of multiple data types of a target system, and determines disaster recovery identification results corresponding to the multiple data types, wherein encoders and decoders are respectively set in the target model according to the detection data of the multiple data types, the encoder is used to extract features of the detection data of the corresponding data type, and the decoder is used to identify whether there is a fault in the detection data of the corresponding data type; a fault transfer module, which is used to transfer the faulty component to an isolated environment based on the dependency relationship between the faulty component and other components in the target system when the disaster recovery identification result indicates that there is an abnormality, wherein the faulty component corresponds to the data type whose disaster recovery identification result is abnormal among the multiple data types; a recovery module, which is used to recover the faulty component in the isolated environment to obtain a recovered target system.

[0014] According to another aspect of an embodiment of the present invention, a non-volatile storage medium is provided, wherein the non-volatile storage medium stores a plurality of instructions, wherein the instructions are suitable for being loaded by a processor and executing any one of the data disaster recovery processing methods.

[0015] According to another aspect of an embodiment of the present invention, there is provided an electronic device, comprising: one or more processors and a memory, wherein the memory is used to store one or more programs, wherein when the one or more programs are executed by the one or more processors, the one or more processors implement any one of the data disaster recovery processing methods described.

[0016] In an embodiment of the present invention, by adopting a target model, the detection data of multiple data types of the target system are processed to determine the disaster recovery identification results corresponding to the multiple data types, wherein the target model is provided with encoders and decoders corresponding to the detection data of the multiple data types, the encoder is used to extract the features of the detection data of the corresponding data type, and the decoder is used to identify whether the detection data of the corresponding data type has a fault; when the disaster recovery identification result indicates that an abnormality exists, based on the dependency relationship between the faulty component and other components in the target system, the faulty component is transferred to an isolated environment, wherein the faulty component corresponds to the data type whose disaster recovery identification result is abnormal in the multiple data types; the faulty component is restored in the isolated environment to obtain the restored target system. The purpose of improving the rapid response capability and recovery efficiency of the target system to faults and ensuring the continuity and integrity of data is achieved, and the technical effect of accurately identifying and isolating the faulty component and quickly recovering the data when an abnormality occurs in the disaster recovery identification is achieved, thereby solving the technical problem of unsatisfactory data fault response efficiency in the related technology. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] The drawings described herein are used to provide a further understanding of the present invention and constitute a part of this application. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation of the present invention. In the drawings:

[0018] Figure 1 is a flowchart of an optional data disaster recovery processing method provided according to an embodiment of the present invention;

[0019] Figure 2 is a target model flow chart of an optional data disaster recovery processing method provided according to an embodiment of the present invention;

[0020] Figure 3 is a schematic diagram of an optional data disaster recovery processing method provided according to an embodiment of the present invention;

[0021] Figure 4 It is a schematic diagram of an optional data disaster recovery processing device provided according to an embodiment of the present invention. DETAILED DESCRIPTION

[0022] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. 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 ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

[0023] It should be noted that the terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0024] For the convenience of description, some nouns or terms involved in the embodiments of the present application are explained below:

[0025] MSE (Mean Square Error) is a commonly used loss function that measures the difference between predicted values ​​and actual values. The mean square error is the average of the squares of the prediction errors and can be used to evaluate the accuracy of the model in regression problems.

[0026] Gradient descent is an optimization algorithm used to minimize the loss function. It gradually approaches the minimum value of the loss function by calculating the gradient of the loss function with respect to the parameters and updating the parameters in the opposite direction of the gradient.

[0027] Containerization is a lightweight virtualization technology that packages an application and its dependent environment into a portable container to enable rapid deployment and isolated operation of the application.

[0028] RAID (Redundant Array of Independent Disks) is a data storage virtualization technology that combines multiple physical hard disks into a logical unit to provide data redundancy and performance improvement. RAID is used for hardware failure processing and reduces the risk of data loss by adding redundant hardware devices and multi-hard disk systems.

[0029] RTO (Recovery Time Objective) is the maximum acceptable time required for a system or application to return to normal operating state after a failure occurs.

[0030] RPO (Recovery Point Objective) is the point in time when a system or application can be restored to its most recent state before a failure occurs.

[0031] SVM (Support Vector Machine) is a machine learning classification algorithm that distinguishes data points of different categories by finding an optimal hyperplane in the feature space.

[0032] According to an embodiment of the present invention, a method embodiment of data disaster recovery processing is provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and although a logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0033] Figure 1 is a flow chart of an optional data disaster recovery processing method according to an embodiment of the present invention. Figure 1 As shown, the method comprises the following steps:

[0034] Step S102, using the target model to process the detection data of multiple data types of the target system, and determine the disaster recovery identification results corresponding to the multiple data types, wherein the target model is provided with encoders and decoders corresponding to the detection data of the multiple data types, respectively, the encoder is used to extract features of the detection data of the corresponding data type, and the decoder is used to identify whether the detection data of the corresponding data type has a fault;

[0035] It can be understood that the target model is used to process and analyze different data types and test data from the target system. The test data includes operation data, user behavior data, application performance data, and external environmental factor data. The target model is equipped with a dedicated encoder and decoder for each data type. The encoder's task is to extract features from the test data, and the decoder is responsible for identifying whether the data has faults based on these features, providing data support for subsequent fault processing.

[0036] Optionally, the above-mentioned operation data (such as CPU utilization, memory usage, network bandwidth usage), user behavior data (such as login frequency, session duration, page access pattern), application performance data (such as response time, transaction processing speed) and external environmental factor data (such as data center temperature and humidity, power supply status) can be obtained by the detection module. The detection module supports detection tools for multi-dimensional data collection, adopts open source system detection software, integrates open source system detection software into IT infrastructure, and supports customized data collection requirements. All collected data are pre-processed, including data normalization to unify the scale of different data sources, smoothing to reduce noise in the data, and feature engineering to extract features.

[0037] The detection module may include, for example, installing detection agents or sensors on system components, including servers, databases, application servers, and network devices. The detection agents or sensors are responsible for collecting real-time operating data and system status. Integrate user behavior analysis tools (Google Analytics) to collect and analyze user interaction behavior data to understand user usage patterns and application performance issues. Establish a centralized data integration platform to aggregate data collected by various detection tools and sensors. Among them, the target system includes multiple system components, and fault components, recovery components, and reset components belong to multiple system components.

[0038] Optionally, Figure 2 It is a target model flow chart of an optional data disaster recovery processing method provided according to an embodiment of the present invention. Among them, the design of the target model, that is, the autoencoder model, includes: an input layer, an input sublayer, a dense connection layer, an encoder layer, a hidden feature representation layer, a decoder layer, an output sublayer, and an output layer. Among them, multiple input sublayers are designed in the input layer, and each sublayer corresponds to a data type (system operation data, user behavior, application performance, external environmental factors, etc.) to process different types of data characteristics, where N is the total number of multiple data types. Each input sublayer is followed by an encoder, which is composed of multiple dense connection layers and is used to extract the features of each data type. The outputs of each encoder are integrated into a unified hidden feature representation layer by splicing or weighted summation. The hidden feature representation layer is the central layer that integrates all detection data features, allowing the capture of internal correlations across data types. Starting from the hidden feature representation layer, data flows to multiple decoder layers, each decoder layer is responsible for reconstructing the original data of a data type, and each decoder is symmetrical in structure with the corresponding encoder to reconstruct the detection data. The output layer is similar to the input layer structure, and each output sublayer corresponds to reconstructing the reconstructed data of a data type.

[0039] Optionally, in the target model, there are two parameters, the weight matrix W and the bias vector b of the layer. The present application continuously iterates through the gradient descent method to reduce the reconstruction error and optimize the parameters of the model. The steps of the gradient descent method include: initializing parameters, forward propagation, calculating losses, back propagation, updating parameters, and repeated iterations. Before the start of training, all parameters of the target model, namely weights W and bias b, are randomly initialized. The detection data is forward propagated through the encoder and decoder to finally generate reconstructed data. During the forward propagation process of the target model, the output of each layer is a linear combination of input, weight and bias, and then a nonlinear activation function is applied. For each densely connected layer, the calculation is expressed as: output = f (W × input + b), where W is the weight matrix of the layer, the input is the data from the previous layer, b is the bias vector, and f is the activation function. Calculate the loss function value, that is, the mean square error between the detection data and the reconstructed data, and calculate the gradient of the loss function with respect to each parameter through the back propagation algorithm. This involves the application of the chain rule, from the output layer back to the input layer, and calculating the gradient of each parameter layer by layer. The model parameters are updated using the gradient and learning rate. The update is performed in the negative direction of the gradient to reduce the value of the loss function. Repeat the forward propagation, loss calculation, back propagation, and parameter update steps until the performance of the target model no longer improves significantly or the preset number of iterations is reached.

[0040] In an optional embodiment, a target model is used to process detection data of multiple data types of a target system to determine disaster recovery identification results corresponding to the multiple data types, including: based on the detection data corresponding to the multiple data types, inputting the detection data into multiple encoders included in the target model for processing to obtain encoding features corresponding to the multiple data types; based on the encoding features corresponding to the multiple data types, using an implicit feature layer included in the target model for processing to obtain association features between the multiple data types; based on the association features, using multiple decoders included in the target model for processing to obtain reconstructed data corresponding to the multiple data types; based on the reconstructed data corresponding to the multiple data types, determining disaster recovery identification results corresponding to the multiple data types.

[0041] It can be understood that the target model is used to process the detection data of different data types in the target system, and the detection data of various data types are input into the encoder of the target model, and the encoding features of each data type are extracted by the encoder. The encoding features analyze and integrate the correlation features between different data types through the implicit feature layer, and the decoder is used to obtain the reconstructed data of the correlation features, and the reconstruction error is obtained, that is, the difference between the reconstructed data and the original detection data. When the reconstruction error exceeds the set threshold, the target system automatically triggers fault detection and emergency response, and obtains disaster recovery identification results corresponding to multiple data types. Through the above processing, according to different data types, the disaster recovery identification results of the corresponding system components can be targetedly identified, providing data support for the subsequent execution of fault transfer and fault isolation of individual fault components.

[0042] Optionally, the reconstruction error uses mean square error as a metric, and the calculation formula is: Among them, x i is the test data, is the reconstructed data of the target model, and n is the total number of detection data.

[0043] Optionally, the patterns and characteristics of the reconstruction errors are used to classify the identified abnormal events, including distinguishing system failures (e.g., hardware failures, software failures, network failures) and network security attacks; first, extract key features from the reconstruction error of the target model, which can be achieved by analyzing the statistical characteristics of the error (e.g., mean, variance, skewness); associate the known abnormal event types in the historical data (e.g., system failures, network security attacks, natural disaster impacts) with the corresponding reconstruction error patterns, and establish a labeled classification training set through event log analysis; use the extracted features and label data to train a classification model, and use the classification network of the support vector machine (SVM) to learn how to predict the type of event based on the pattern and characteristics of the reconstruction error; during the operation of the system, calculate the reconstruction error of the inbound data in real time and extract the corresponding features. These features are input into the trained classification model to identify and classify the current abnormal events in real time.

[0044] It should be noted that the use of support vector machines (SVM) to classify abnormal events includes the following steps: extracting feature data (such as mean, variance, skewness) from the reconstruction error of the target model, and standardizing the feature data to eliminate the dimension. Through event log analysis, known abnormal event types and their corresponding reconstruction error patterns are collected. Each sample consists of a feature vector and a corresponding label (abnormal type). The reconstruction error is calculated in real time, and the corresponding features are extracted. These features are input into the trained support vector machine model to identify and classify abnormal events.

[0045] In an optional embodiment, based on the detection data corresponding to the multiple data types, multiple encoders included in the input target model are processed to obtain encoding features corresponding to the multiple data types, including: based on the detection data corresponding to the multiple data types, multiple input sublayers included in the input target model are processed to obtain input features corresponding to the multiple data types, wherein the multiple input sublayers correspond one-to-one to the multiple data types; based on the input features corresponding to the multiple data types, multiple encoders included in the input target model are processed to obtain encoding features corresponding to the multiple data types.

[0046] It can be understood that each data type is preprocessed by a corresponding input sublayer to extract input features. The input sublayer is the front-end part of the target model, which is used to identify and process specific data. The input features are sent to the encoder, which uses these input features to generate corresponding encoding features as compression and abstract identification of the data. The encoding features capture the key information of the detection data and provide a basis for subsequent fault detection and correlation analysis.

[0047] Step S104, when the disaster recovery identification result indicates that there is an abnormality, based on the dependency relationship between the faulty component and other components in the target system, the faulty component corresponds to the data type whose disaster recovery identification result is abnormal among multiple data types.

[0048] It can be understood that when the disaster recovery identification result indicates an abnormal situation in the target system, the dependency relationship between the faulty component and other components is analyzed to ensure that the status of related system components and services is taken into account during the isolation process to minimize the impact on the business, where the faulty component corresponds to the data type for which the disaster recovery identification result is abnormal. The faulty component is transferred to an isolated environment to prevent the fault from spreading to other parts of the system.

[0049] In an optional embodiment, the method further includes: determining, based on the dependency relationship, an impacted component among other components that is impacted by the faulty component; and restarting the impacted component using a predetermined configuration strategy.

[0050] It can be understood that according to the dependency relationship between the faulty component and other components, the affected component is located, and a predetermined configuration strategy is adopted to perform a restart process on the affected component to avoid the spread of the fault. The above process is conducive to reducing the abnormality of other components caused by the faulty component.

[0051] Optionally, the preset configuration strategy can be multiple, for example: for simple configuration errors detected, the preset repair strategy is automatically applied to correct them, including modifying the wrong configuration file or updating the wrong system settings. For services that need to be restarted, the restart procedure will be automatically executed, including stopping the current service, ensuring that all data is correctly saved, and then restarting the service.

[0052] Step S106, recovering the failed component in the isolated environment to obtain a recovered target system.

[0053] It can be understood that after the failed component is transferred to the isolated environment, the previously backed-up data is used to restore the failed component, and the restored component is reintegrated into the target system to obtain a restored target system.

[0054] In an optional embodiment, the failed component is recovered in an isolated environment to obtain a recovered target system, including: in the isolated environment, using the backup data before the exception occurs to recover the failed component to obtain a recovery component; remapping data connections, updating system configurations, and restoring user access, performing integration processing based on the recovery component to obtain the recovered target system.

[0055] It can be understood that the most recent data backup point before the exception occurs is automatically identified and selected in the isolated environment to perform recovery operations on the faulty component. The selected backup data is restored to an isolated environment to generate a recovery component, and then verified, and the verified data is reintegrated into the target system, including remapping data connections, updating system configurations, and restoring user access, to obtain a restored target system.

[0056] It should be noted that the most recent data backup point is an incremental backup or full backup based on a timestamp or event trigger. The isolation environment can be of various types, such as using a pre-configured backup system or cloud service environment to automatically perform failover operations in the isolation environment to ensure that critical services can continue to run on the backup system or cloud service.

[0057] In an optional embodiment, the method of remapping data connections, updating system configurations, and restoring user access is adopted, and integration processing is performed based on recovery components to obtain a restored target system, including: based on a predetermined foreign key relationship and a predetermined data format, a data consistency check is performed on the recovery component, wherein the predetermined foreign key relationship is used to constrain key references to be in a valid state; when the recovery component passes the data consistency check, the method of remapping data connections, updating system configurations, and restoring user access is adopted, and integration processing is performed based on the recovery component to obtain a restored target system.

[0058] It is understood that before the recovery component is reintegrated into the target system, the target system will perform a data consistency check on the recovery component based on the predetermined foreign key relationship and data format to ensure the validity of the data. After passing the consistency check, the verified data will be reintegrated into the target system, including remapping data connections, updating system configurations, and restoring user access. After the recovery component is restored to the original target system, performance testing can be performed to ensure that it is restored to the operating state before the failure.

[0059] It should be noted that the checking of foreign key relationships and data formats is completed through SQL queries or data validation scripts to ensure that all data relationships and logical constraints comply with business rules and database integrity requirements.

[0060] Optionally, performance testing is used to verify whether the target system can reach the level before the failure after data recovery. The performance testing may include, for example: response time testing, resource utilization testing, throughput testing, data integrity and consistency testing, etc.

[0061] Optionally, remapping data connections, updating system configurations, and restoring user access are key steps in the data disaster recovery process, ensuring the consistency of the target system after recovery. Remapping data connections involves identifying and re-establishing data flows and communication paths between system components, including reconfiguring database connection strings, network interfaces, and data synchronization mechanisms to ensure that data can be correctly transmitted in the system; updating system configurations involves adjusting system settings to adapt to the recovered environment, including reallocation of hardware resources, adjustment of software parameters, and updating of security policies to ensure that the system configuration is consistent with the current business needs and technical environment; restoring user access requires restarting the user authentication mechanism, synchronizing access rights, and reconfiguring the user interface to ensure that users can access system data as unimpeded as before the failure.

[0062] In an optional embodiment, the method further includes: determining the fault property of the faulty component and the fault impact caused by the faulty component; and reconfiguring the resource distribution of the recovered target system based on the fault property and the fault impact.

[0063] It is understandable that after the target system is restored, it is necessary to reconstruct and optimize the target system according to the nature and impact of the faulty components. Identify the specific fault nature of the faulty components, and the target system evaluates the impact of the fault on business operations and system performance. Perform resource distribution redistribution based on the fault nature and fault impact assessment results. It is also necessary to regularly evaluate the target system performance and capacity requirements, and adjust resource allocation according to business growth and technological changes, including increasing server processing power, storage capacity, and network bandwidth.

[0064] Optionally, the nature of the failure and the aspects affected by the failed component may be as follows: Hardware failures may be of various types, including server hard disk failures, memory failures, and power supply problems, with impacts including data loss, system instability, or complete downtime. Software failures may be of various types, such as operating system crashes, application software errors, and configuration errors, with impacts including service interruptions, data processing errors, or performance degradation. Network failures may be of various types, such as network hardware failures, configuration errors, or network interruptions caused by external attacks, with impacts including data transmission delays, inaccessible services, or data security issues. Network security attack failures may be of various types, such as security vulnerabilities being exploited and malware infections, with impacts including data leakage, system tampering, or service interruptions.

[0065] Optionally, resource reconfiguration based on the nature of the failure includes the following two aspects: For hardware failures, add redundant hardware devices, such as multi-disk systems using RAID configuration, multi-power supply systems, and hot spares. For software failures, update and patch management strategy optimization, ensure that all system and application software are up to date, and configure management tools to prevent unauthorized changes.

[0066] The optimization of settings based on the impact of failures includes the following two aspects: for network failures, design a multi-path network architecture, introduce high-availability network devices, including multi-link switches and routers, and configure network performance detection monitoring and automatic fault switching mechanisms. For security incidents, implement end-to-end data encryption, and regularly perform security vulnerability scanning and penetration testing. Through the above step S102, the target model is used to process the detection data of multiple data types of the target system, and the disaster recovery identification results corresponding to the multiple data types are determined, wherein the target model is respectively set according to the detection data of multiple data types. The encoder is used to extract the features of the detection data of the corresponding data type, and the decoder is used to identify whether the detection data of the corresponding data type has a fault; step S104, when the disaster recovery identification result indicates that there is an abnormality, based on the dependency relationship between the faulty component and other components in the target system, the faulty component is transferred to an isolated environment, wherein the faulty component corresponds to the data type whose disaster recovery identification result is abnormal in the multiple data types; step S106, the faulty component is restored in the isolated environment to obtain the restored target system. It can improve the system's ability to respond quickly to failures and its recovery efficiency, ensure the continuity and integrity of data, and achieve the technical effect of accurately identifying and isolating faulty components and quickly recovering data when disaster recovery identifies abnormalities, thereby solving the technical problem of unsatisfactory data failure response efficiency in related technologies.

[0067] Based on the above embodiments and optional embodiments, the present invention proposes an optional implementation manner.

[0068] Step S1, real-time data and system status detection: deploy a detection module to detect the system's operating data, user behavior, and application performance in real time.

[0069] Optionally, the detection module supports detection tools for multi-dimensional data collection, uses open source system detection software, integrates the open source system detection software into the IT infrastructure, and supports customized data collection requirements. The detection module also includes installing detection agents or sensors on system components, including servers, databases, application servers, and network devices. The detection agents or sensors are responsible for collecting real-time operating data and system status; integrating user behavior analysis tools (Google Analytics) to collect and analyze user interaction behavior data, understand user usage patterns and application performance issues; and establishing a centralized data integration platform to aggregate data collected by various detection tools and sensors.

[0070] Step S2, intelligent fault detection: using the target model, the detection data is analyzed in real time to identify and classify system failures or disaster events. When a failure is detected, the management personnel are notified immediately and the emergency response procedure is automatically triggered.

[0071] The target model specifically includes: obtaining system operation data (CPU utilization, memory usage, network bandwidth usage), user behavior data (login frequency, session duration, page access pattern), application performance data (response time, transaction processing speed) and external environmental factors (data center temperature and humidity, power supply status), preprocessing all collected data, including data normalization to unify the scale of different data sources, smoothing to reduce noise in the data, and feature engineering to extract features; target model architecture design: design the target model to process various types of detection data, including different input layers, one for each data type, and configure the architecture of the target model, including the number of hidden layers and the number of nodes in each layer, to adapt to the complexity and diversity of the detection data; model training and optimization: use the multidimensional data collected during normal operation as the training set to train the target model, and optimize the model by minimizing the reconstruction error; real-time anomaly detection and classification: deploy the trained model for real-time data detection and analysis, and identify it as a potential failure or disaster event when the detected data reconstruction error exceeds the predetermined threshold; use the pattern and characteristics of the reconstruction error to classify the identified abnormal events, including distinguishing system failures and network security attacks. The gradient descent method is used to minimize the reconstruction error. The target model is trained by calculating the gradient of the loss function with respect to each weight and updating the weight according to this gradient to gradually reduce the error.

[0072] Step S3, automated fault isolation: After a fault occurs, the fault isolation procedure is automatically executed to prevent the fault from spreading and affecting other system resources. At the same time, simple configuration errors are automatically corrected or the affected services are restarted to stabilize the system status. After a fault occurs, it also includes performing emergency data backup, especially fast backup of data that has changed since the last backup.

[0073] Step S31, the fault isolation program is started: the connection of the affected system or service is cut off from the network layer to prevent the fault from spreading to other unaffected parts, and the containerization strategy is used to isolate the affected instance to prevent the fault from affecting the overall system.

[0074] Step S32, automatic correction and service restart: For simple configuration errors detected, the preset repair strategy is automatically applied to correct them, including modifying the erroneous configuration file or updating the erroneous system settings. For services that need to be restarted, the restart procedure will be automatically executed, including stopping the current service, ensuring that all data is correctly saved, and then restarting the service.

[0075] Step S33, dependency and impact analysis: During the fault isolation and repair process, the dependencies between the faulty component and other system components are analyzed to ensure that the status of related system components and services are taken into account during the isolation and restart process to minimize the impact of the operation on the business.

[0076] Step S4, failover and load balancing: Utilize pre-configured backup systems or cloud services to automatically perform failover operations to ensure that critical services can continue to run on the backup systems or cloud services.

[0077] Step S5, initial data recovery: restore the data from the most recent backup to a stable state, and automatically perform data integrity and consistency checks to ensure that the recovered data is accurate and available.

[0078] Step S51, automatically identifying and selecting the latest data backup point, where the latest data backup point is an incremental backup or a full backup based on a timestamp or event trigger.

[0079] Step S52: restore the selected backup data to an isolated environment.

[0080] Step S53, using SQL query or data verification script to check foreign key relationships and data formats, perform data consistency check, and ensure that all data relationships and logical constraints comply with business rules and database integrity requirements.

[0081] Step S54, reintegrate the verified data into the main system, including remapping data connections, updating system configurations and restoring user access. After the data is restored to the production environment, perform performance testing to ensure that it is restored to the operating state before the failure.

[0082] Step S6, system reconstruction and optimization: After fault response and initial recovery, reconfigure and optimize system settings and resource distribution according to the nature and impact of the fault.

[0083] Step S61, reconfigure resources based on the nature of the failure: for hardware failures, add redundant hardware devices, such as a multi-hard disk system configured with a redundant array of independent disks (RAID), a multi-power supply system, and hot spares; for software failures, optimize update and patch management strategies to ensure that all detection data and application software are up to date, and configure management tools to prevent unauthorized changes.

[0084] Step S62, impact-based setting optimization: for network failures, design a multi-path network architecture, introduce high-availability network devices, including multi-link switches and routers, configure network performance detection and automatic fault switching mechanism; for security incidents, implement end-to-end data encryption, and conduct security vulnerability scanning and penetration testing regularly.

[0085] Step S63, regularly evaluate system performance and capacity requirements, and adjust resource allocation according to business growth and technology changes, including increasing server processing power, storage capacity, and network bandwidth.

[0086] Figure 3 is a schematic diagram of an optional microgrid stability determination method provided according to an embodiment of the present invention. Figure 3 The processing method shown in the figure is explained.

[0087] The real-time monitoring module is deployed in the enterprise's IT infrastructure, including servers, databases, application servers and network devices. It can be used to detect various system components and support multi-dimensional data collection, including system operation data, user behavior and application performance.

[0088] The intelligent fault detection module is used to adopt the target model to analyze the data collected by the real-time detection module, identify abnormal behavior, and train and optimize by analyzing the reconstruction error to achieve anomaly detection. When the anomaly detected exceeds the predetermined threshold, the module automatically notifies the management personnel and triggers the emergency response procedure.

[0089] The automated fault isolation and response module is used to automatically perform fault isolation after a fault is detected, using network layer measures and containerization strategies to prevent the fault from affecting other parts of the system. At the same time, it automatically corrects configuration errors or restarts affected services and performs emergency data backup.

[0090] The data recovery module is used to automatically restore data from the most recent backup into an isolated environment, perform data integrity and consistency checks, and ensure data accuracy and availability.

[0091] The system reconstruction and optimization module is used to conduct in-depth diagnosis based on the nature and impact of the fault, reconfigure and optimize system settings and resource distribution, and implement adjustments to resource configuration and system settings for different types of faults.

[0092] The above optional implementation method achieves at least the following effects: through real-time data and system status detection, intelligent fault detection and automated fault isolation, the disaster recovery management method significantly improves the system's ability to respond quickly to faults. Real-time detection ensures that the problem is identified when it first occurs, and the target model allows the system to automatically diagnose and isolate the problem area at the beginning of the fault to prevent the spread of the fault. Rapid response and preventive measures reduce system downtime and ensure that key businesses can continue to run seamlessly on backup systems or cloud services, thereby ensuring business continuity and data integrity. Through resource reconfiguration and system setting optimization based on the nature of the fault, resource allocation is made more scientific and efficient. The redundant system and automatic update management configured in advance for hardware and software faults not only reduce the waste of physical and software resources, but also reduce the risk of unexpected failures through preventive maintenance.

[0093] It should be noted that the steps shown in the flowcharts of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and that, although a logical order is shown in the flowcharts, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0094] In this embodiment, a data disaster recovery processing device is also provided, which is used to implement the above-mentioned embodiments and preferred implementation modes, and the descriptions that have been made will not be repeated. As used below, the terms "module" and "device" can implement a combination of software and / or hardware for a predetermined function. Although the devices described in the following embodiments are preferably implemented in software, the implementation of hardware, or a combination of software and hardware, is also possible and conceivable.

[0095] According to an embodiment of the present invention, there is also provided an apparatus embodiment for implementing a data disaster recovery processing method. Figure 4 is a schematic diagram of a data disaster recovery processing device according to an embodiment of the present invention. Figure 4 As shown, the above-mentioned data disaster recovery processing device includes: a detection and identification module 402, a fault transfer module 404, and a recovery module 406. The device is described below.

[0096] The detection and identification module 402 is used to use the target model to process the detection data of various data types of the target system, and determine the disaster recovery identification results corresponding to the various data types. The target model is provided with encoders and decoders corresponding to the detection data of various data types, respectively. The encoder is used to extract the features of the detection data of the corresponding data type, and the decoder is used to identify whether there is a fault in the detection data of the corresponding data type.

[0097] The fault transfer module 404 is connected to the detection and identification module 402, and is used to transfer the faulty component to an isolated environment based on the dependency relationship between the faulty component and other components in the target system when the disaster recovery identification result indicates an abnormality, wherein the faulty component corresponds to the data type whose disaster recovery identification result is abnormal among the multiple data types.

[0098] The recovery module 406 is connected to the failover module 404 and is used to recover the failed component in the isolated environment to obtain a recovered target system.

[0099] In a data disaster recovery processing device provided by an embodiment of the present invention, a detection and identification module 402 is set to use a target model to process the detection data of multiple data types of the target system, and determine the disaster recovery identification results corresponding to the multiple data types, wherein the target model is respectively provided with an encoder and a decoder according to the detection data of the multiple data types, the encoder is used to extract the characteristics of the detection data of the corresponding data type, and the decoder is used to identify whether the detection data of the corresponding data type has a fault; a fault transfer module 404 is connected to the detection and identification module 402, and is used to transfer the fault component to an isolated environment based on the dependency relationship between the fault component and other components in the target system when the disaster recovery identification result indicates that there is an abnormality, wherein the fault component corresponds to the data type whose disaster recovery identification result is abnormal in the multiple data types; a recovery module 406 is connected to the fault transfer module 404, and is used to recover the fault component in the isolated environment to obtain the recovered target system. The purpose of improving the system's rapid response capability and recovery efficiency to faults, ensuring the continuity and integrity of data, and achieving the technical effect of accurately identifying and isolating the fault component and quickly recovering data when the disaster recovery identification is abnormal, thereby solving the technical problem of unsatisfactory data fault response efficiency in the related technology.

[0100] It should be noted that the above modules can be implemented by software or hardware. For example, for the latter, it can be implemented in the following ways: the above modules can be located in the same processor; or the above modules can be located in different processors in any combination.

[0101] It should be noted that the detection and identification module 402, the fault transfer module 404, and the recovery module 406 correspond to steps S102 to S106 in the embodiment, and the examples and application scenarios implemented by the modules and the corresponding steps are the same, but are not limited to the contents disclosed in the above embodiment. It should be noted that the modules as part of the device can be run in a computer terminal.

[0102] It should be noted that the optional or preferred implementation of this embodiment can refer to the relevant description in the embodiment, which will not be repeated here.

[0103] The above-mentioned data disaster recovery processing device may also include a processor and a memory. The detection and identification module 402, the fault transfer module 404, and the recovery module 406 correspond to steps S102 to S106 in the embodiment, which are all stored in the memory as program units, and the processor executes the above-mentioned program units stored in the memory to realize the corresponding functions.

[0104] The processor includes a kernel, which retrieves the corresponding program unit from the memory. There can be one or more kernels. The memory may include non-permanent memory in a computer-readable medium, random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM, and the memory includes at least one storage chip.

[0105] An embodiment of the present invention provides a non-volatile storage medium on which a program is stored. When the program is executed by a processor, a data disaster recovery processing method is implemented.

[0106] An embodiment of the present invention provides an electronic device, which includes a processor, a memory, and a program stored in the memory and can be run on the processor. When the processor executes the program, the following steps are implemented: using a target model, processing detection data of multiple data types of a target system, and determining disaster recovery identification results corresponding to the multiple data types, wherein encoders and decoders are respectively set according to the detection data of the multiple data types in the target model, the encoder is used to extract the features of the detection data of the corresponding data type, and the decoder is used to identify whether the detection data of the corresponding data type has a fault; when the disaster recovery identification result indicates that an abnormality exists, based on the dependency relationship between the faulty component and other components in the target system, the faulty component is transferred to an isolated environment, wherein the faulty component corresponds to the data type whose disaster recovery identification result is abnormal in the multiple data types; the faulty component is restored in the isolated environment to obtain a restored target system. The device in this article can be a server, a PC, etc.

[0107] The present invention also provides a computer program product, which, when executed on a data processing device, is suitable for executing a program that is initialized with the following method steps: using a target model to process detection data of multiple data types of a target system, and determining disaster recovery identification results corresponding to the multiple data types, wherein encoders and decoders are respectively set in the target model according to the detection data of the multiple data types, the encoder is used to extract features of the detection data of the corresponding data type, and the decoder is used to identify whether there is a fault in the detection data of the corresponding data type; when the disaster recovery identification result indicates that an abnormality exists, based on the dependency relationship between the faulty component and other components in the target system, the faulty component is transferred to an isolated environment, wherein the faulty component corresponds to the data type whose disaster recovery identification result is an abnormality among the multiple data types; and the faulty component is restored in the isolated environment to obtain a restored target system.

[0108] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0109] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0110] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0111] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0112] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0113] The memory may include non-permanent memory in a computer-readable medium, random access memory (RAM) and / or non-volatile memory in the form of read-only memory (ROM) or flash RAM. The memory is an example of a computer-readable medium.

[0114] Computer readable media include permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology to store information. Information can be computer readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic tape disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer readable media does not include temporary computer readable media (transitory media), such as modulated data signals and carrier waves.

[0115] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity 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, commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.

[0116] It will be appreciated by those skilled in the art that embodiments of the present invention may be provided as methods, systems or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment or an embodiment combining software and hardware. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program codes.

[0117] The above are only embodiments of the present invention and are not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent substitution, improvement, etc. made within the spirit and principle of the present invention should be included in the scope of the claims of the present invention.

Claims

1. A data disaster recovery processing method, characterized in that: include: Using a target model, processing detection data of multiple data types of a target system, and determining disaster recovery identification results corresponding to the multiple data types, wherein an encoder and a decoder are respectively set in the target model according to the detection data of the multiple data types, the encoder is used to extract features of the detection data of the corresponding data type, and the decoder is used to identify whether the detection data of the corresponding data type has a fault; When the disaster recovery identification result indicates that an abnormality exists, based on the dependency relationship between the faulty component and other components in the target system, the faulty component is transferred to an isolated environment, wherein the faulty component corresponds to the data type of which the disaster recovery identification result indicates an abnormality among the multiple data types; The failed component is restored in the isolated environment to obtain a restored target system.

2. The method according to claim 1, characterized in that The target model is used to process the detection data of multiple data types of the target system to determine the disaster recovery identification results corresponding to the multiple data types, including: Based on the detection data corresponding to the multiple data types, the detection data are input into multiple encoders included in the target model for processing to obtain encoding features corresponding to the multiple data types; Based on the encoding features respectively corresponding to the multiple data types, the implicit feature layer included in the target model is used for processing to obtain correlation features between the multiple data types; Based on the associated features, a plurality of decoders included in the target model are used to perform processing respectively to obtain reconstructed data corresponding to the plurality of data types respectively; Based on the reconstructed data respectively corresponding to the multiple data types, disaster recovery identification results respectively corresponding to the multiple data types are determined.

3. The method according to claim 2, characterized in that The detection data corresponding to the multiple data types are input into multiple encoders included in the target model for processing to obtain encoding features corresponding to the multiple data types, including: Based on the detection data corresponding to the multiple data types, the detection data are input into multiple input sub-layers included in the target model for processing to obtain input features corresponding to the multiple data types, wherein the multiple input sub-layers correspond to the multiple data types one by one; Based on the input features respectively corresponding to the multiple data types, multiple encoders included in the target model are input for processing to obtain the encoding features respectively corresponding to the multiple data types.

4. The method according to claim 1, characterized in that The method further comprises: Based on the dependency relationship, determine an affected component among the other components that is affected by the faulty component; A predetermined configuration strategy is adopted to restart the affected component.

5. The method according to claim 1, characterized in that The recovering the failed component in the isolated environment to obtain a recovered target system includes: In the isolated environment, the failed component is restored using the backup data before the abnormality occurs to obtain a restored component; The restored target system is obtained by performing integration processing based on the recovery component in a manner of remapping data connection, updating system configuration, and restoring user access.

6. The method according to claim 5, characterized in that The method of remapping data connection, updating system configuration, and restoring user access is used to perform integration processing based on the recovery component to obtain the restored target system, including: Based on a predetermined foreign key relationship and a predetermined data format, a data consistency check is performed on the recovery component, wherein the predetermined foreign key relationship is used to constrain the reference of the key to be in a valid state; In the case where the recovery component passes the data consistency check, the data connection is remapped, the system configuration is updated, and the user access is restored, and integration processing is performed based on the recovery component to obtain the restored target system.

7. The method according to any one of claims 1 to 6, characterized in that: The method further comprises: Determining the fault nature of the faulty component and the fault impact caused by the faulty component; Based on the nature of the fault and the impact of the fault, resource distribution of the recovered target system is reconfigured.

8. A data disaster recovery processing device, characterized in that: include: A detection and identification module, used to process detection data of multiple data types of a target system using a target model, and determine disaster recovery identification results corresponding to the multiple data types, wherein an encoder and a decoder are respectively set in the target model according to the detection data of the multiple data types, the encoder is used to extract features of the detection data of the corresponding data type, and the decoder is used to identify whether the detection data of the corresponding data type has a fault; a fault transfer module, configured to transfer the faulty component to an isolated environment based on a dependency relationship between the faulty component and other components in the target system when the disaster recovery identification result indicates that an abnormality exists, wherein the faulty component corresponds to a data type of which the disaster recovery identification result indicates that an abnormality exists among the multiple data types; The recovery module is used to recover the failed component in the isolated environment to obtain a recovered target system.

9. A non-volatile storage medium, characterized in that: The non-volatile storage medium stores a plurality of instructions, and the instructions are suitable for being loaded by a processor and executing the data disaster recovery processing method described in any one of claims 1 to 7.

10. An electronic device, characterized in that: include: One or more processors and a memory, wherein the memory is used to store one or more programs, wherein when the one or more programs are executed by the one or more processors, the one or more processors implement the data disaster recovery processing method described in any one of claims 1 to 7.