A data backup strategy automatic generation method and system

Abstract

The application relates to the technical field of data backup, in particular to a data backup strategy automatic generation method and system. The method comprises the following steps: acquiring and quantifying strategy generation input information; evaluating the strategy related attributes of data assets, estimating the resource input cost of meeting different preset protection levels according to the strategy generation input information; constructing and real-time updating a dynamic model reflecting the strategy trade-off relationship based on the strategy generation input information, the strategy related attributes and the resource input cost; making a strategy decision based on the dynamic model, wherein the strategy decision selects a backup strategy based on compliance regulations combined with real-time resource conditions; deploying the backup strategy and continuously monitoring the execution effect of the backup strategy, and taking the actual operation data of the execution effect as feedback to adjust the parameters of the dynamic model. The method solves the limitations of the existing data backup strategy automatic generation method in complex situations and improves the ability to optimize the strategy based on the actual recovery result feedback.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of data backup, in particular to a data backup strategy automatic generation method and system. BACKGROUND

[0002] In modern enterprise operations, data has become a core asset, and its security and recoverability are directly related to the continuity of business and the survival of enterprises. In order to effectively protect the security of these massive data, the formulation and implementation of data backup strategy is particularly critical. Traditional backup strategies often rely on manual experience, which not only consumes time and effort, but also easily leads to improper strategy configuration due to negligence, thereby affecting the efficiency and success rate of data recovery. In order to overcome these challenges, enterprises generally introduce data backup strategy automatic generation methods, aiming to automatically recommend or generate an optimized backup strategy through intelligent analysis of data characteristics and business needs, in order to improve management efficiency and the reliability of data protection. However, in actual application, with the rapid development of business and the continuous evolution of IT environment, these automated methods also face a series of new and complex challenges.

[0003] Specifically, the existing data backup strategy automatic generation method has obvious limitations in the face of real-time changes in data characteristics, intensified competition for backup resources, conflicts between compliance and cost targets, discrepancies between actual recovery performance and expectations, and sudden security incidents. For example, when the data characteristics of an enterprise change in real time due to sudden business growth, the original backup strategy generated based on static data analysis may no longer be applicable, resulting in the inability to meet the new recovery point target requirements. At the same time, when multiple business systems simultaneously perform backup operations, shared resources within the data center, such as network data transmission capacity and storage device input / output performance, begin to show bottlenecks, affecting backup efficiency and even production system performance. In addition, the compliance requirements and cost control targets faced by enterprises are also constantly changing, and existing strategies may not fully meet new compliance requirements or generate unnecessary overhead. The deeper problem is that existing methods often only focus on the execution status of backup tasks, and lack the ability to iterate and optimize strategies based on actual recovery results, resulting in discrepancies between actual recovery performance and expected targets.

[0004] In view of the above problems, the existing technology needs to be improved. SUMMARY

[0005] The present application discloses a data backup strategy automatic generation method, which solves the limitations of existing data backup strategy automatic generation methods in complex situations and improves the ability to optimize strategies based on actual recovery results.

[0006] The technical solution of the present application is as follows:

[0007] In a first aspect, the application discloses a data backup strategy automatic generation method, comprising:

[0008] Obtaining and quantifying strategy generation input information, the strategy generation input information including compliance regulations of external data and cost control targets of the enterprise internally;

[0009] Evaluating the strategy-related attributes of the data assets, the strategy-related attributes including data sensitivity, importance, current backup strategy state, potential compliance risk and recovery performance requirement, and estimating the resource input cost for meeting different preset protection levels according to the strategy generation input information;

[0010] Based on the strategy generation input information, the strategy-related attributes and the resource input cost, a dynamic model reflecting the strategy trade-off relationship is constructed and updated in real time;

[0011] Based on the dynamic model, a strategy decision is made, the strategy decision being based on the compliance regulations combined with real-time resource conditions to select a backup strategy, the real-time resource conditions including available capacity of storage devices, network data transmission capability and input / output performance of storage devices;

[0012] Deploying the backup strategy and continuously monitoring the execution effect of the backup strategy, actual running data of the execution effect being used as feedback to adjust parameters of the dynamic model.

[0013] Further, in the process of making a strategy decision based on the dynamic model and deploying the backup strategy, it further comprises:

[0014] Collecting and storing micro-pattern characteristics of key network transmission paths, the micro-pattern characteristics including data packet delay, delay jitter, data packet interval time distribution, transmission control protocol window size change pattern and throughput distribution, forming a normal operation baseline;

[0015] When a large-scale data migration or backup task is started, micro-pattern characteristics of the task data stream are sampled and analyzed in real time to obtain a task flow characteristic sequence;

[0016] The task flow characteristic sequence is compared with the normal operation baseline to calculate a deviation degree index;

[0017] When the deviation degree index continuously exceeds a preset deviation threshold within a preset time window, it is determined that there is an implicit anomaly, an abnormality warning is triggered, and the abnormal event is provided as input to the strategy decision for strategy adjustment;

[0018] According to the recovery point target satisfaction degree, the cost increment and the resource availability, a preset emergency shunting plan is prioritized, the emergency shunting plan including a replacement transmission path and a backup scheduling scheme;

[0019] Based on the abnormal event, the highest priority alternative solution is recommended from the emergency diversion plan, the alternative solution is executed, and the execution result of the alternative solution is used as feedback to adjust the parameters of the dynamic model.

[0020] Further, in the process of making a strategy decision based on the dynamic model, it also includes:

[0021] A performance isolation threshold library of shared infrastructure components is established, which records the maximum number of concurrent data streams and / or input / output performance of each shared infrastructure component under different business loads, and the interference characteristics of a specific type of data stream on other services;

[0022] According to the task size, data stream type and scheduling time of the candidate large-scale data migration or backup task, the resource occupation mode of the candidate solution on the shared infrastructure component is simulated;

[0023] The resource occupation mode is matched with the interference characteristics in the performance isolation threshold library. When the matching result predicts that the candidate solution will cause the resource occupation of the shared infrastructure component to exceed the performance isolation threshold and / or overlap with the interference characteristics, the candidate solution is marked as a high-risk solution, and the interference penalty cost is calculated;

[0024] The interference penalty cost is included in the total cost input of the candidate solution in the dynamic model, and based on the adjusted total cost input, the backup strategy is selected from the candidate solutions that meet the compliance regulations in combination with the real-time resource status.

[0025] Further, in the process of deploying the backup strategy and continuously monitoring the execution effect of the backup strategy, it also includes:

[0026] A business scenario recovery path map is established, which is used to map the internal structure, mutual dependency relationship of data assets and the application environment supported thereby;

[0027] When the business scenario recovery verification process is performed, the complete path and corresponding time from data recovery to end-to-end function recovery of the business system are tracked and recorded based on the business scenario recovery path map;

[0028] The complete path and corresponding time are analyzed to identify key dependent links and bottlenecks in the actual recovery path, and to quantify the impact of each key dependent link and bottleneck on the overall business recovery time;

[0029] The overall business recovery time is compared with the recovery time target and / or recovery point target preset by the business department, and the deviation degree is calculated;

[0030] When the deviation degree exceeds the preset time deviation threshold, the actual deviation between the actual recovery performance and the business expectation caused by changes in the internal structure, mutual dependency relationship or application environment is identified;

[0031] The actual deviation is transformed into a business scenario recovery benefit correction factor, which is used to adjust the benefit function parameters in the dynamic model.

[0032] Furthermore, the process of assessing the strategy-related attributes of data assets also includes:

[0033] Continuously monitor and parse business process iteration information, which includes updates to business system functions, adjustments to data processing procedures, and changes in user access behavior.

[0034] Based on business process iteration information, identify the affected data assets and their associated business scenarios;

[0035] For the affected data assets, analyze the frequency of calls, data update frequency and data access paths of the data assets in the business system, and determine their current business value and access patterns by combining business system logs and user behavior data. Based on the impact of business process iteration information on data processing methods, reassess their compliance classification and obtain real-time detection results.

[0036] The real-time detection results are compared with the policy-related attributes stored in the data asset metadata to calculate the attribute deviation.

[0037] When the attribute deviation exceeds the preset attribute threshold, the strategy-related attributes of the data assets are automatically updated;

[0038] Based on the updated policy-related attributes and policy-generated input information, the system re-estimates the resource investment costs required to meet different preset protection levels and / or triggers real-time updates to the dynamic model.

[0039] Furthermore, continuously monitoring and parsing business process iteration information also includes:

[0040] Integrate with enterprise service bus and / or message queue to listen for service update messages published by business systems;

[0041] Based on the service update message, configure the log parsing adapter to convert heterogeneous update log information into a standardized data structure;

[0042] Construct an iterative information dependency graph based on a standardized data structure to identify update events with conflicting and / or sequential dependencies;

[0043] Natural language processing technology is used to parse unstructured business documents and change requests, extract keywords and semantic associations related to business system function updates, data processing flow adjustments, and changes in user access behavior, and obtain semantic association results.

[0044] By integrating standardized data structures, iterative information dependency graphs, and semantic association results, a comprehensive business process change report is generated for use when assessing the strategy-related attributes of data assets.

[0045] Furthermore, the real-time probing of affected data assets also includes:

[0046] Deploy infrastructure-level performance monitoring probes to collect real-time data on storage media read / write latency, IOPS, throughput, as well as network device bandwidth utilization, packet loss rate, and link jitter.

[0047] Deploy audit probes at the security policy enforcement level to record execution logs of data anonymization operations and / or trigger events of access control rules in real time;

[0048] By correlating and analyzing the infrastructure metrics collected by performance monitoring probes with the access patterns of data assets, changes in data asset access performance caused by infrastructure changes can be identified.

[0049] By correlating and analyzing the security policy execution logs recorded by the audit probes with the compliance classification of data assets, changes in data asset compliance risks caused by security policy changes can be identified.

[0050] Based on the results of the correlation analysis, the real-time detection results of the data asset's business value, compliance classification, and / or access patterns are revised.

[0051] Based on the corrected real-time detection results, combined with business system logs and user behavior data, the frequency of data asset calls, data update frequency and data access paths in the business system are re-analyzed to determine its current business value and access patterns, and compliance classification is reassessed.

[0052] Furthermore, when performing correlation analysis between infrastructure metrics collected by performance monitoring probes and access patterns of data assets, the following is also included:

[0053] The infrastructure indicator stream and data asset access pattern stream are transmitted in a streaming manner, with the accompanying identification information including at least the data asset identifier and the collection source identifier.

[0054] During data collection, timestamps are appended to each infrastructure metric sample and each data asset access event, and the infrastructure metric stream and data asset access pattern stream are aligned based on the timestamps.

[0055] A sliding time window is used to perform correlation analysis between the aligned infrastructure indicator stream and the data asset access pattern stream;

[0056] After infrastructure indicators fluctuate beyond a preset fluctuation threshold, the trend of data asset access performance changes within a subsequent preset time window. When there is a continuous and consistent trend of change within a preset time window that is no less than a preset window count threshold, a causal relationship is determined.

[0057] When a causal relationship is determined, the impact of infrastructure changes on data asset access performance is quantified, and the quantification results are used to correct the parameters of the dynamic model and / or for policy decisions based on the dynamic model.

[0058] Furthermore, when quantifying the impact of infrastructure changes on the performance of data asset access, the following also applies:

[0059] For data assets, the access patterns are identified as bursty access patterns and / or periodic access patterns;

[0060] When an abrupt access pattern is identified, after an infrastructure change occurs, access performance indicators of data assets are collected at a preset high-frequency sampling interval until the fluctuation of the access performance indicators within a preset time window that is no less than a preset window counting threshold does not exceed a preset stability threshold. The access performance indicators within the initial observation window after the change are compared with the access performance indicators within the reference window of the same length before the change to obtain the instantaneous difference, which is used as the instantaneous impact degree.

[0061] When a periodic access pattern is identified, after an infrastructure change occurs, the access performance indicators of the data asset are continuously collected for at least one complete cycle. The access performance indicators within one complete cycle after the change are compared with the access performance indicators of the corresponding cycle before the change, and the periodic average difference is calculated. The periodic average difference is used as the degree of continuous impact.

[0062] By combining the instantaneous impact and the sustained impact, a quantitative result is obtained on the impact of infrastructure changes on the performance of data asset access. The quantitative result is used to correct the parameters of the dynamic model and / or for policy decisions based on the dynamic model.

[0063] Secondly, this application also discloses an automatic data backup strategy generation system, comprising:

[0064] The information acquisition module is used to acquire and quantify the input information for strategy generation, which includes compliance regulations of external data and cost control targets of the enterprise.

[0065] The asset assessment module is used to assess the policy-related attributes of data assets. These attributes include data sensitivity, importance, current backup policy status, potential compliance risks, and recovery performance requirements. Based on the policy, it generates input information to estimate the resource investment costs required to meet different preset protection levels.

[0066] The model building module is used to build and update a dynamic model that reflects the trade-offs between strategies based on strategy-generated input information, strategy-related attributes, and resource input costs.

[0067] The strategy decision module is used to make strategy decisions based on a dynamic model. The strategy decision is based on compliance regulations and real-time resource status to select backup strategies. Real-time resource status includes available storage capacity, network data transmission capacity, and storage device input / output performance.

[0068] The strategy deployment and monitoring module is used to deploy backup strategies and continuously monitor the execution effect of backup strategies, using the actual running data of the execution effect as feedback to adjust the parameters of the dynamic model. Beneficial effects

[0069] This application discloses an automatic data backup strategy generation method. It acquires and quantifies input information for strategy generation, evaluates the strategy-related attributes of data assets, estimates the resource investment costs to meet different preset protection levels, and constructs and updates a dynamic model reflecting the strategy trade-offs in real time. Based on this, strategy decisions are made according to the dynamic model, selecting a backup strategy in conjunction with compliance regulations and real-time resource conditions, deploying the backup strategy, continuously monitoring its execution effect, and using actual operational data as feedback to adjust the parameters of the dynamic model. This method effectively solves the problems of reliance on manual experience, low efficiency, and error-proneness in data backup strategy formulation in existing technologies, significantly improving the automation and intelligence level of data backup strategies, enhancing the reliability of data protection, and improving business continuity for enterprises. Attached Figure Description

[0070] Figure 1 This is a flowchart illustrating a method for automatically generating a data backup strategy provided in this application.

[0071] Figure 2 This is a schematic diagram of the module structure of an automatic data backup strategy generation system provided in this application.

[0072] In the diagram: 1. Information acquisition module; 2. Asset valuation module; 3. Model building module; 4. Strategy decision-making module; 5. Strategy deployment and monitoring module. Detailed Implementation

[0073] The technical solutions of this application will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are merely some embodiments of this application, and not all embodiments. The components of this application described and shown in the accompanying drawings can generally be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of this application provided in the accompanying drawings is not intended to limit the scope of the claimed application, but merely to illustrate selected embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without inventive effort are within the scope of protection of this application.

[0074] Reference Figure 1 This application proposes a method for automatically generating data backup strategies, including:

[0075] S1000: Acquire and quantify strategy generation input information, which includes compliance regulations for external data and internal cost control objectives of the enterprise;

[0076] S2000: Assess the policy-related attributes of data assets, including data sensitivity, importance, current backup policy status, potential compliance risks, and recovery performance requirements, and estimate the resource investment costs to meet different preset protection levels based on policy-generated input information.

[0077] S3000: Based on policy generation input information, policy-related attributes, and resource input costs, it constructs and updates a dynamic model in real time that reflects the policy trade-offs.

[0078] S4000: Makes policy decisions based on a dynamic model. The policy decisions are based on compliance regulations and real-time resource conditions to select backup strategies. Real-time resource conditions include available storage capacity, network data transmission capacity, and storage device input / output performance.

[0079] S5000: Deploy backup strategies and continuously monitor their execution effectiveness, using actual operational data as feedback to adjust parameters of the dynamic model.

[0080] This application provides a method for automatically generating data backup strategies, aiming to achieve automated generation, optimization, and adjustment of data backup strategies through intelligent means. The core of this method lies in constructing a dynamic decision-making system capable of responding in real time to changes in the external environment and internal resource conditions.

[0081] The input information for strategy generation forms the foundation of the entire strategy generation process. This includes external regulatory compliance and internal cost control objectives. Regulatory compliance refers to legal requirements such as GDPR and HIPAA regarding data storage, processing, and transmission. These requirements directly impact the level and method of data backup protection. Cost control objectives represent the company's budget constraints and efficiency goals in data protection. For example, a company might set an annual cap on total backup costs or require backup storage costs to not exceed a certain percentage.

[0082] Policy-related attributes of data assets are crucial for assessing data value and risk. These include data sensitivity, importance, current backup policy status, potential compliance risks, and recovery performance requirements. Data sensitivity, for example, refers to personally identifiable information and financial data, which typically require a higher level of protection. Importance, for example, refers to data from core business systems, which have more stringent Recovery Time Objectives (RTO) and Recovery Point Objectives (RPO). The current backup policy status reflects the data's current protection level and backup cycle. Potential compliance risks refer to the compliance violations that the data may face under the existing policy. Recovery performance requirements refer to how long it takes for the data to be restored to an usable state in the event of a failure, and how much data loss can be tolerated.

[0083] A dynamic model is a mathematical or logical model that can be updated in real time to reflect the trade-offs between costs, risks, and benefits under different backup strategy choices. This model can be a multi-objective optimization model, for example, using linear or nonlinear programming to balance compliance, cost, and recovery performance.

[0084] Real-time resource status refers to the current physical and logical resource status of the system when making policy decisions, including available storage capacity, network data transfer capacity, and storage device I / O performance. For example, available storage capacity refers to the unused storage space in a current storage array or cloud storage service. Network data transfer capacity refers to the real-time bandwidth and latency of network links within or outside the data center. Storage device I / O performance refers to the number of read / write operations per second (IOPS) and data throughput that the storage system can handle.

[0085] Firstly, regarding the acquisition and quantification of input information for strategy generation, this input includes external compliance regulations and internal cost control objectives. For example, the latest compliance regulations, such as data retention periods and encryption requirements, can be manually entered or retrieved from the compliance management system via API. These regulations can be quantified as a series of constraints, such as requiring specific types of data to be retained for at least five years or to be encrypted using AES-256. Internal cost control objectives can be set through the financial system or management, for example, setting monthly backup storage costs to not exceed 100,000 yuan, or setting backup operation CPU and memory usage to not exceed 5% of the production system. These objectives can be quantified as cost items in the optimization function.

[0086] Secondly, regarding the assessment of policy-related attributes of data assets, these attributes include data sensitivity, importance, current backup policy status, potential compliance risks, and recovery performance requirements. For example, data sensitivity can be automatically identified and labeled using data classification tools; for instance, data containing ID numbers can be labeled as "highly sensitive." Importance can be assessed through business department priority settings or indicators such as data access frequency and modification frequency; for example, core database data is labeled as "highly important." The current backup policy status can be obtained by querying the backup management system; for example, it can show that a certain database currently uses a strategy of daily full backups plus hourly incremental backups. Potential compliance risks can be assessed using compliance scanning tools; for example, it can reveal that data in a storage bucket is not encrypted as required. Recovery performance requirements can be set based on the business department's RTO and RPO requirements; for example, a critical business system may be required to have an RTO of less than 4 hours and an RPO of less than 15 minutes. Based on this, the resource investment costs to meet different preset protection levels are estimated according to the policy-generated input information. For example, for highly sensitive data requiring an RPO of 1 hour and an RTO of 4 hours, the system can estimate the storage space, network bandwidth, and computing resource costs required to adopt an off-site replication accelerated photo backup solution.

[0087] Secondly, regarding the construction and real-time updating of dynamic models reflecting strategy trade-offs based on strategy generation input information, strategy-related attributes, and resource investment costs, a multi-objective optimization model can be constructed. The objective functions include minimizing total cost, minimizing compliance risk, and maximizing recovery performance. Constraints include compliance requirements and resource availability limitations. This model can be solved using linear programming, integer programming, or heuristic algorithms. When the strategy generation input information or strategy-related attributes change—for example, due to the release of new compliance regulations or increased sensitivity of a data asset—the dynamic model will update its parameters and constraints in real time to reflect the latest trade-offs.

[0088] Next, regarding strategy decision-making based on the dynamic model, the strategy selection is based on compliance regulations combined with real-time resource conditions. Real-time resource conditions include available storage device capacity, network data transfer capacity, and storage device I / O performance. For example, when making a decision, the dynamic model queries the available capacity of current storage devices. If a storage pool's capacity is insufficient, it prioritizes other storage pools or cloud storage. Simultaneously, the model considers current network data transfer capacity to avoid large-scale data transfers during peak network periods, preventing impact on the production system. Storage device I / O performance is also taken into consideration to ensure that backup tasks do not place excessive load on the storage system. By comprehensively considering these real-time resource conditions, the dynamic model can select a backup strategy that meets compliance requirements and can be executed efficiently under current resource conditions.

[0089] Finally, regarding the deployment of backup strategies and the continuous monitoring of their execution effectiveness, actual operational data is used as feedback to adjust parameters of the dynamic model. For example, once the strategy decision module selects a backup strategy, it is deployed to the backup management system. The system continuously monitors the execution of backup tasks, including backup success rate, backup duration, data transfer volume, and resource usage. This actual operational data is collected and fed back into the dynamic model. For instance, if the actual execution time of a backup strategy consistently exceeds expectations or resource usage is excessive, the dynamic model adjusts its internal parameters based on this feedback data, such as adjusting the weight of a resource cost, to avoid similar problems in future decisions, thereby achieving continuous optimization and adaptive adjustment of the strategy.

[0090] In another embodiment of this application, it is further proposed that the process of making strategy decisions and deploying backup strategies based on dynamic models also includes:

[0091] S6000: Collects and stores micro-pattern characteristics of critical network transmission paths, including packet delay, delay jitter, packet interval time distribution, transmission control protocol window size variation pattern, and throughput distribution, forming a normal operation baseline;

[0092] S7000: When a large-scale data migration or backup task is started, it samples and analyzes the micro-pattern characteristics of the task data stream in real time to obtain the task stream feature sequence.

[0093] S8000: Compares the task flow feature sequence with the normal operating baseline and calculates the deviation index;

[0094] S9000: When the deviation index continues to exceed the preset deviation threshold within the preset time window, it is determined that there is a non-obvious anomaly, triggers an anomaly warning, and provides the anomaly event as input to the strategy decision-making process for strategy adjustment.

[0095] S10000: Prioritize the preset emergency diversion plans based on the recovery point target satisfaction, cost increment, and resource availability. The emergency diversion plans include alternative transmission paths and backup scheduling schemes.

[0096] S11000: Based on abnormal events, recommend the highest priority alternative from the emergency triage plan, execute the alternative, and use the execution result of the alternative as feedback to adjust the parameters of the dynamic model.

[0097] Specifically, the micro-pattern characteristics of critical network transmission paths refer to the fine-grained behavioral patterns exhibited by the network and transport layers during data transmission. Packet latency refers to the time required for a data packet to travel from the source to the destination; latency jitter refers to the magnitude of variation in packet latency; packet interval distribution reflects the regularity of packet arrival; the transmission control protocol window size variation pattern reveals the behavior of transmission control protocol flow control; and throughput distribution represents the amount of data successfully transmitted per unit time and its fluctuations. These characteristics can be collected by deploying probes at critical network nodes or utilizing the monitoring functions of existing network devices (such as routers and switches). Through long-term, continuous collection and statistical analysis of these micro-pattern characteristics, a baseline can be established for the network under normal operating conditions. This baseline represents the typical performance of the network under anomaly-free conditions.

[0098] When a large-scale data migration or backup task is initiated, the system samples and analyzes the micro-pattern characteristics of the task's data stream in real time, generating a task stream feature sequence. This task stream feature sequence is then compared with a pre-established normal operating baseline to calculate a deviation index. The deviation index can be calculated using various statistical methods, such as Euclidean distance, Mahalanobis distance, or machine learning-based outlier scores, to quantify the difference between the current task stream characteristics and the normal baseline.

[0099] Non-obvious anomalies refer to network performance degradation or behavioral abnormalities that do not immediately cause service interruption or obvious failure, but can only be identified through continuous deviations in micro-pattern characteristics. When the deviation index continuously exceeds a preset deviation threshold within a preset time window, such as several minutes or hours, the system will determine the existence of such a non-obvious anomaly and immediately trigger an anomaly warning. This anomaly event is then provided as a key input to the policy decision module so that the system can adjust the current backup policy in a timely manner.

[0100] To address potential intangible anomalies, the system has pre-defined multiple emergency diversion plans. These plans include, but are not limited to, switching to backup network transmission paths, adjusting data transmission priorities, suspending some non-critical tasks, or adopting different backup scheduling schemes (e.g., switching from full backup to incremental backup). When an anomaly occurs, the system prioritizes these plans based on multiple dimensions, including recovery point objective satisfaction (RPO satisfaction based on a pre-defined evaluation function), cost increment, and resource availability. Recovery point objective satisfaction assesses the ability of different plans to control the risk of data loss, cost increment considers the additional overhead that may be incurred in implementing the plan, and resource availability ensures that the selected plan is feasible under current resource conditions. Based on the analysis of the anomaly and the prioritization of the plans, the system recommends and executes the highest-priority alternative. The execution results of the alternative, including its impact on network performance, backup progress, and resource consumption, will be used as feedback data to adjust the parameters of the dynamic model in real time. This allows the model to better adapt to changes in the actual operating environment, improving the accuracy and robustness of future strategy decisions.

[0101] In another embodiment of this application, the process of making strategy decisions based on the above dynamic model includes the following steps:

[0102] S12000: Establish a performance isolation threshold library for shared infrastructure components. The performance isolation threshold library records the maximum number of concurrent data streams and / or input / output performance of each shared infrastructure component under different business loads, as well as the interference characteristics of specific types of data streams that cause performance interference to other services.

[0103] S13000: Based on the task size, data stream type, and scheduling time of the candidate large-scale data migration or backup tasks, simulate the resource consumption pattern of the candidate solutions on the shared infrastructure components.

[0104] S14000: Match resource occupancy patterns with interference features in the performance isolation threshold library. When the matching result predicts that the candidate solution will cause the resource occupancy of the shared infrastructure component to exceed the performance isolation threshold and / or overlap with the interference features, the candidate solution is marked as a high-risk solution, and the interference penalty cost is calculated.

[0105] S15000: The interference penalty cost is included in the total cost of candidate solutions in the dynamic model, and a backup strategy is selected from candidate solutions that meet compliance regulations based on the adjusted total cost and real-time resource status.

[0106] Specifically, shared infrastructure components can refer to shared storage arrays, network switches, virtualization platforms, or computing clusters within an enterprise. The performance isolation threshold library aims to predefine the capacity boundaries of these shared resources under different business pressures, and the performance impact patterns that specific types of data flows (e.g., high-IOPS database backups, large file transfers) may have on other services (e.g., online transaction systems, real-time analytics platforms). For example, this threshold library can record the maximum IOPS or throughput that a storage array can use for backup tasks when carrying OLTP services, and how the packet size distribution and burstiness of backup tasks affect the latency of OLTP services.

[0107] Before making a strategy decision, the system simulates the resource consumption that a candidate large-scale data migration or backup task might generate when running on shared infrastructure components, based on its expected task size (e.g., data volume), data flow type (e.g., sequential read / write, random read / write), and planned scheduling time, thus forming a resource consumption pattern. This simulation process can be performed using historical data, predictive algorithms, or simulation techniques.

[0108] Subsequently, the resource usage pattern is matched against pre-defined interference features in the performance isolation threshold library. If the matching result predicts that the candidate solution will cause the resource usage of shared infrastructure components to exceed their performance isolation thresholds—for example, exceeding the bandwidth or IOPS limit reserved for other critical services—or if its data flow characteristics overlap with known interference features, such as the bursty read / write pattern of backup tasks conflicting with the sensitive read / write pattern of online services, then the candidate solution will be marked as a high-risk solution. At this point, the system will calculate an interference penalty cost (e.g., interference penalty cost = Σ(importance weight of affected system × performance degradation percentage × impact duration × unit business loss)), which quantifies the potential losses caused by the performance degradation, business interruption, or additional resource investment that the solution may cause.

[0109] Ultimately, the cost of interference penalties is factored into the total cost of the candidate solution in the dynamic model. A solution that may lead to shared resource conflicts, even if its direct resource consumption is low, will become "more expensive" due to the potential cost of interference penalties. Based on the adjusted total cost and combined with real-time resource conditions such as available storage capacity, network data transfer capabilities, and storage device I / O performance, the system will select the backup strategy with the lowest overall cost (including direct costs and potential interference costs) and lowest risk from all candidate solutions that meet compliance regulations.

[0110] As one specific implementation method:

[0111] Suppose a company has a shared storage area network (SAN) and a core database server that handles high-concurrency online transactions. The SAN is also used to store various backup data. When automatically generating data backup policies, the system needs to decide whether to execute a terabyte-scale database backup task during the operating hours.

[0112] First, the system queries a pre-established performance isolation threshold database. This database records the minimum IOPS and bandwidth reserved by the SAN for the database server during peak online transaction periods, as well as the latency jitter interference characteristics that large-scale backup tasks (e.g., sequential read / write mode) may cause to database IO in random read / write mode.

[0113] Next, the system will simulate the terabyte-level backup task's resource usage patterns on the SAN during the scheduled time (e.g., the working period), including its expected IOPS, throughput, and packet characteristics.

[0114] The system then matches the simulated resource usage pattern against performance isolation thresholds and interference characteristics in the threshold library. If the matching result predicts that the backup task, when executed during the working hours, will cause the SAN's IOPS or bandwidth usage to exceed the threshold reserved for online transaction services, and its sequential read / write pattern severely overlaps with the database's random read / write pattern, potentially leading to a significant increase in database response time, then the backup scheme will be marked as a high-risk scheme. At this point, the system calculates an interference penalty cost, for example, based on a quantified value of the potential loss of database services per minute or the degradation of user experience.

[0115] Ultimately, the cost of this interference penalty will be included in the total cost of the backup solution. For example, if the direct backup cost is 1000 yuan, but the interference penalty cost is 5000 yuan, the total cost becomes 6000 yuan. In contrast, a backup solution executed during off-peak hours might have a direct cost of 1200 yuan, but an interference penalty cost of 0 yuan, resulting in a total cost of 1200 yuan. Based on the adjusted total cost, the system will prioritize backup solutions executed during off-peak hours, thereby avoiding potential interference with core business operations and ensuring business continuity and stability.

[0116] In another embodiment of this application, the process of deploying a backup strategy and continuously monitoring its execution effect further includes:

[0117] S16000: Establish a business scenario recovery path diagram. The business scenario recovery path diagram is used to map the internal structure, interdependencies and supporting application environment of data assets.

[0118] S17000: When performing the business scenario recovery verification process, the complete path and corresponding time from data recovery to end-to-end functional recovery of the business system are tracked and recorded based on the business scenario recovery path diagram.

[0119] S18000: Analyze the complete path and corresponding time, identify key dependent links and bottlenecks in the actual recovery path, and quantify the impact of each key dependent link and bottleneck on the overall business recovery time.

[0120] S19000: Compare the overall business recovery time with the business department's preset recovery time target and / or recovery point target, and calculate the deviation.

[0121] S20000: When the deviation exceeds the preset time deviation threshold, identify the actual deviation between the actual recovery performance and business expectations caused by changes in internal structure, interdependencies, or application environment.

[0122] S21000: Converts actual deviations into business scenario recovery benefit correction factors, used to adjust the benefit function parameters in the dynamic model.

[0123] Specifically, the establishment of a business scenario recovery path diagram aims to provide a comprehensive view that clearly depicts all data assets upon which a specific business function or service depends, the logical and physical connections between these data assets, and the software and hardware environment in which they run. This path diagram can be represented using a graphical interface or a structured data model, with the purpose of revealing data flow, service call chains, and potential single points of failure or recovery bottlenecks. For example, for an online transaction system, its recovery path diagram might include the database, application server, message queue, caching service, and external interfaces, with detailed annotations of their dependency order and data synchronization mechanisms. The business scenario recovery verification process involves simulating or actually executing a recovery operation of the business system to verify the effectiveness of the backup strategy. During this process, the system, based on the pre-established business scenario recovery path diagram, accurately tracks and records every step, every stage, and the time consumed at each stage from the initial data recovery operation (e.g., restoring the database from backup storage) until the entire business system fully restores its end-to-end functionality (e.g., users can log in normally and complete transactions). This includes data loading time, application startup time, service registration time, and dependent service readiness time.

[0124] By analyzing the complete path and time from data recovery to business function recovery, the longest, most error-prone, and critical steps are identified. Sensitivity analysis or causal analysis is used to quantify the impact of each dependency and bottleneck on the overall recovery time. The actual total recovery time is compared with the recovery time and recovery point targets set in the business continuity plan or disaster recovery plan to determine the deviation. If the deviation exceeds a threshold, the cause is located (such as database schema changes, new service dependencies, operating system or hardware changes). Based on this, the deviation is transformed into a business scenario recovery efficiency adjustment factor, and the dynamic model efficiency function is adjusted: a large deviation increases the weight of fast recovery, while a small deviation decreases it to explore more cost-effective strategies.

[0125] In another embodiment of this application, it is further proposed that the process of evaluating the policy-related attributes of data assets also includes:

[0126] S2100: Continuously monitors and parses business process iteration information, including business system function updates, data processing process adjustments, and changes in user access behavior.

[0127] S2200: Identify the affected data assets and their associated business scenarios based on business process iteration information;

[0128] S2300: For affected data assets, analyze the frequency of data asset calls, data update frequency and data access path in business systems, and determine their current business value and access patterns by combining business system logs and user behavior data. Based on the impact of business process iteration information on data processing methods, reassess their compliance classification and obtain real-time detection results.

[0129] S2400: Compare the real-time detection results with the policy-related attributes stored in the data asset metadata, and calculate the attribute deviation.

[0130] S2500: When the attribute deviation exceeds the preset attribute threshold, automatically update the strategy-related attributes of the data assets;

[0131] S2600: Based on the updated policy-related attributes and policy-generated input information, re-estimate the resource input cost required to meet different preset protection levels, and / or trigger real-time updates of the dynamic model.

[0132] Specifically, continuously monitoring and parsing business process iteration information refers to the system's uninterrupted collection and analysis of changes in internal business processes. These changes may include updates to business system functions, such as adding modules, optimizing functions, or deprecating old functions; adjustments to data processing processes, such as modifications to data cleaning rules, changes in data flow paths, or changes in data storage formats; and changes in user access behavior, such as a significant increase or decrease in the access frequency of specific data assets, or a shift in access patterns from batch processing to real-time querying. This information can be obtained through various means, including integrating internal message queues, service buses, version control systems, or parsing business documents and change requests. Data processing methods include, but are not limited to, data collection, storage, replication, transmission, encryption, data masking, compression, archiving, and destruction—activities that impact the data lifecycle.

[0133] Identifying affected data assets and their associated business scenarios based on business process iteration information means that the system automatically identifies which specific data assets (such as database tables, files, API interface data, etc.) may be affected, and the business scenarios (such as order processing, customer management, financial statements, etc.) supported by these data assets may undergo changes based on the parsed business process changes. For example, if a new function involves the processing of sensitive customer information, the relevant customer data assets will be identified as affected.

[0134] For affected data assets, the frequency of data asset calls, data update frequency, and data access paths within business systems are analyzed. Combined with business system logs and user behavior data, their current business value and access patterns are determined. Furthermore, based on the impact of business process iteration information on data processing methods, their compliance classification is reassessed, yielding real-time detection results. This step aims to deeply analyze the actual usage of affected data assets. Call frequency and data update frequency reflect the activity and importance of data assets; data access paths reveal data flow and dependencies. Combining business system logs (such as operation logs and error logs) and user behavior data (such as access records and operation statistics) allows for a more accurate determination of the current business value of data assets (e.g., data accessed frequently and supporting core business has higher value) and access patterns (e.g., read-heavy, write-light, bursty access). Simultaneously, based on the impact of business process iteration information on data processing methods, such as adjustments to data anonymization rules or changes in data storage location, their compliance classification is reassessed (e.g., from general data to sensitive data, or from being subject to specific regulations to being subject to stricter regulations), thus obtaining a real-time detection result reflecting the latest state of the data assets.

[0135] The attribute deviation is calculated by comparing real-time detection results with policy-related attributes stored in the data asset's metadata. This involves comparing the newly detected data asset attributes (such as real-time business value, access patterns, and compliance classification) with the policy-related attributes (such as initially set data sensitivity, importance, and potential compliance risks) pre-stored in the metadata management system. Attribute deviation measures the difference between the actual situation and the initial assumptions. For example, if a data asset's compliance classification changes from "general" to "sensitive," the deviation is considered high.

[0136] When attribute deviation exceeds a preset attribute threshold, the policy-related attributes of the data asset are automatically updated. The preset attribute threshold is a configurable parameter used to define the degree of deviation that requires an attribute update. Once the deviation exceeds this threshold, the system will automatically correct the policy-related attributes of the data asset stored in the metadata to ensure that they are always consistent with the latest business status.

[0137] Based on the updated policy-related attributes and policy generation input information, the system re-estimates the resource investment costs required to meet different preset protection levels and / or triggers real-time updates to the dynamic model. After the attribute updates, the system uses these latest policy-related attributes, combined with the original policy generation input information (such as compliance regulations and cost control objectives), to recalculate the resource investment costs required to meet different protection levels (e.g., backup schemes with more stringent RPO / RTO requirements). Furthermore, these updates may also directly trigger real-time updates to the dynamic model to ensure that the model always makes policy trade-offs and decisions based on the latest and most accurate data asset status.

[0138] In some of the embodiments described above in this application, during the process of evaluating the policy-related attributes of data assets, it is necessary to continuously monitor and parse business process iteration information. Specifically, continuously monitoring and parsing business process iteration information may include:

[0139] S2110: Integrates enterprise service bus and / or message queue to listen for service update messages published by business systems;

[0140] S2120: Based on the service update message, configure the log parsing adapter to convert heterogeneous update log information into a standardized data structure;

[0141] S2130: Construct an iterative information dependency graph based on a standardized data structure to identify update events with conflicting and / or sequential dependencies;

[0142] S2140: Use natural language processing technology to parse unstructured business documents and change requests, extract keywords and semantic associations related to business system function updates, data processing flow adjustments, and changes in user access behavior, and obtain semantic association results;

[0143] S2150: Integrates standardized data structures, iterative information dependency graphs, and semantic association results to generate an overall business process change report for use when assessing the strategy-related attributes of data assets.

[0144] The integration of Enterprise Service Bus (ESB) and / or message queues aims to establish a unified communication backbone for capturing service update messages from different business systems. ESB acts as middleware, enabling message routing, transformation, and coordination between different systems; message queues provide asynchronous communication mechanisms, ensuring reliable message delivery and decoupling. By monitoring these service update messages, key information such as business system function updates, data processing flow adjustments, and changes in user access behavior can be obtained in real time. Based on the service update messages, a log parsing adapter is configured to convert update log information from different sources and with varying formats into a standardized data structure. For example, the adapter can parse logs in various formats such as XML, JSON, and CSV, mapping them to a predefined, structured data format for subsequent automated processing and analysis.

[0145] Based on standardized data structures, iterative information dependency graphs can be constructed. These graphs are used to visualize and identify conflicting and / or sequentially dependent update events in business process iteration information. For example, if an adjustment to a data processing workflow depends on an update to a business system function, the graph can clearly represent this dependency, helping to consider the order and scope of impact of updates in subsequent evaluations. Natural language processing (NLP) techniques are used to parse unstructured business documents and change requests, aiming to extract valuable business process iteration information from text descriptions. For example, NLP models can identify keywords and semantic associations such as "add user registration function," "modify order processing logic," and "optimize data query performance," thereby supplementing details that may be missed in structured messages and obtaining semantic association results. Finally, by integrating the standardized data structure, iterative information dependency graph, and semantic association results, a comprehensive business process change report can be generated. This report comprehensively summarizes all iteration information of the business process, providing comprehensive and accurate input for subsequent evaluation of the strategy-related attributes of data assets.

[0146] In another embodiment of this application, it is further proposed that the process of real-time detection of affected data assets also includes:

[0147] S2310: Deploy infrastructure-level performance monitoring probes to collect real-time data on storage media read / write latency, IOPS, throughput, as well as network device bandwidth utilization, packet loss rate, and link jitter.

[0148] S2320: Deploy audit probes at the security policy enforcement level to record execution logs of data anonymization operations and / or triggering events of access control rules in real time;

[0149] S2330: Correlate infrastructure metrics collected by performance monitoring probes with data asset access patterns to identify changes in data asset access performance caused by infrastructure changes.

[0150] S2340: Correlate the security policy execution logs recorded by the audit probe with the compliance classification of data assets to identify changes in data asset compliance risks caused by security policy changes;

[0151] S2350: Based on the results of correlation analysis, correct the real-time detection results of the business value, compliance classification and / or access patterns of data assets;

[0152] S2360: Based on the corrected real-time detection results, combined with business system logs and user behavior data, re-analyze the call frequency, data update frequency and data access path of data assets in the business system, determine their current business value and access patterns, and reassess compliance classification.

[0153] Specifically, infrastructure-level performance monitoring probes can be understood as software agents or hardware modules deployed on critical infrastructure components such as storage devices and network devices. Their purpose is to continuously and granularly collect the operational status and performance metrics of these components. For example, for storage media, metrics such as read / write latency, input / output operations per second (IOPS), and data throughput can be collected; for network devices, metrics such as bandwidth utilization, packet loss rate, and link jitter can be collected. These metrics directly reflect the performance status of the physical or virtual resources upon which data assets depend. Meanwhile, security policy enforcement-level audit probes refer to monitoring tools deployed on security components such as data masking systems and access control systems. Their purpose is to record events related to the execution of data security policies in real time. For example, when data is masked, the probe records the execution log of the masking operation, including the operation time, the operation object, and the operation result; when an access control rule is triggered, the probe records the triggering event, including the visitor, the accessed resource, the rule type, and the decision result. These logs and events directly reflect the actual operation of compliance assurance measures for data assets. In practical applications, correlation analysis between infrastructure metrics collected by performance monitoring probes and access patterns of data assets refers to using data analytics techniques to compare and integrate performance data of the underlying infrastructure with access behavior data of the upper-layer data assets to identify potential causal relationships between the two. For example, when the read / write latency of storage media suddenly increases, it can be analyzed whether it is related to changes in the access patterns of a specific data asset, thereby determining whether infrastructure changes have affected the access performance of that data asset.

[0154] Correlation analysis between security policy execution logs recorded by audit probes and compliance classifications of data assets refers to assessing the impact of security policy implementation on data asset compliance risks by analyzing the implementation status of security policies. For example, if the de-identification operation execution logs of a sensitive data asset show an increased de-identification failure rate, or if access control rules are frequently bypassed or triggered abnormally, it may indicate that the compliance risk of that data asset is increasing.

[0155] Based on the results of the aforementioned correlation analysis, the real-time detection results of the business value, compliance classification, and / or access patterns of data assets can be revised. For example, if infrastructure performance degradation leads to a significant deterioration in the access performance of a data asset, its business value may be reassessed as lower; if problems in security policy enforcement increase compliance risks, its compliance classification may be adjusted to a more stringent level. The revised real-time detection results will be combined with business system logs and user behavior data for a more comprehensive re-analysis to ultimately determine the current business value and access patterns of the data asset, and reassess its compliance classification, thereby ensuring a more accurate and comprehensive assessment of the data asset attributes.

[0156] In another embodiment of this application, when performing correlation analysis between infrastructure metrics collected by performance monitoring probes and access patterns of data assets, the method further includes:

[0157] S2331: Transmit infrastructure indicator streams and data asset access pattern streams in a streaming manner, with the accompanying identification information including at least data asset identifiers and collection source identifiers;

[0158] S2332: During data collection, timestamps are added to each infrastructure metric sample and each data asset access event, and the infrastructure metric stream and data asset access pattern stream are aligned based on the timestamps.

[0159] S2333: Use a sliding time window to perform correlation analysis on the aligned infrastructure indicator flow and data asset access pattern flow;

[0160] S2334: After infrastructure indicators fluctuate beyond a preset fluctuation threshold, track the trend of data asset access performance within a subsequent preset time window. When there is a continuous and consistent trend of change within a preset time window that is no less than a preset window count threshold, it is determined that there is a causal relationship.

[0161] S2335: When a causal relationship is determined, the impact of infrastructure changes on the performance of data asset access is quantified, and the quantification results are used to correct the parameters of the dynamic model and / or for policy decisions based on the dynamic model.

[0162] Specifically, streaming infrastructure metrics and data asset access pattern data refers to transmitting infrastructure-level performance monitoring probes' metric data (such as storage media read / write latency, IOPS, throughput, network device bandwidth utilization, packet loss rate, and link jitter) and data asset access pattern data (such as call frequency, data update frequency, and data access path) as a real-time, continuous data stream. Identification information, such as data asset identifiers, uniquely identifies specific data assets, while source identifiers indicate which monitoring probe or system the data originated from, ensuring data traceability. This identification information, along with timestamps, aims to provide precise context and alignment benchmarks for subsequent correlation analysis.

[0163] During data collection, timestamps are appended to each infrastructure metric sample and each data asset access event. The infrastructure metric stream and data asset access pattern stream are then aligned based on these timestamps. This aims to ensure accurate temporal matching of data from different sources and of different types. The accuracy of timestamps is crucial for identifying the chronological order and synchronization of events. Timestamp alignment eliminates data misalignment caused by data transmission delays or system clock differences, laying the foundation for subsequent causal relationship determination.

[0164] Using a sliding time window to perform correlation analysis on aligned infrastructure metric streams and data asset access pattern streams involves defining a fixed-length time window and allowing it to slide continuously along the time axis to perform real-time or near-real-time correlation calculations on the data within the window. This approach can dynamically capture short-term and long-term correlations between data streams, adapt to constantly changing system behavior, and avoid the lag or inaccuracies that may result from static analysis.

[0165] After infrastructure metrics fluctuate beyond a preset fluctuation threshold, the trend of data asset access performance is tracked within subsequent preset time windows. A causal relationship is determined when a consistent trend exists within a preset time window (at least a preset window count threshold). The preset fluctuation threshold defines a significant change in infrastructure metrics; for example, a drop in storage IOPS exceeding a preset percentage might be considered a fluctuation. The preset time window defines the time range for observing changes in data asset access performance, such as within a specific number of minutes after an infrastructure fluctuation occurs. The preset window count threshold requires a consistent trend across multiple consecutive observation windows; for example, a sustained increase in data asset access latency across multiple preset time windows is considered strong evidence of a causal relationship. This multi-condition judgment mechanism aims to improve the accuracy of causal relationship determination and avoid false positives.

[0166] When a causal relationship is determined, the impact of infrastructure changes on data asset access performance is quantified, and the quantification results are used to correct the parameters of the dynamic model and / or for policy decisions based on the dynamic model. The quantification of impact can be achieved by calculating the average change, maximum change, or rate of change of performance metrics (such as latency and throughput). These quantification results directly reflect the specific impact of infrastructure changes on data asset performance, providing precise input to the aforementioned dynamic model, enabling it to more accurately assess risks, costs, and benefits, thereby optimizing the generation and adjustment of backup strategies.

[0167] In some preferred embodiments, the following specific example illustrates the situation:

[0168] Suppose a company's data center is running a critical online transaction system, with its core data assets stored on a high-performance storage array. The IOPS (Input / Output Operations Per Second) of this storage array is one of the key infrastructure metrics.

[0169] At a certain point in time, the performance monitoring probe detects a sudden drop in the storage array's IOPS metric, which continues to exceed a preset fluctuation threshold (e.g., IOPS drops from 100,000 to 80,000, exceeding the 15% fluctuation threshold). At this point, the system immediately initiates tracking of the access performance of the data assets (e.g., transaction databases) associated with the storage array.

[0170] Specifically, the system streams the IOPS metrics of the storage array and the access latency of the transaction database. During data acquisition, each IOPS sample and each database access event is appended with a precise timestamp and aligned based on these timestamps. Subsequently, a 5-minute sliding time window is used to perform correlation analysis on the aligned data streams.

[0171] When a fluctuation in IOPS occurs, the system begins tracking the trend of transaction database access latency over subsequent preset time windows (e.g., each 5-minute window). If the system observes a sustained increase in the average access latency of the transaction database (e.g., from 10 milliseconds to 20 milliseconds, then to 30 milliseconds) over a consecutive period of at least a preset window count threshold (e.g., three consecutive 5-minute windows), and this increasing trend is consistent with the direction of IOPS decline, then the system will determine that there is a causal relationship between the decrease in storage array IOPS and the increase in transaction database access latency.

[0172] Once a causal relationship is determined, the system quantifies the degree of impact. For example, it calculates that for every 10% decrease in IOPS, the average access latency to the transaction database increases by 10 milliseconds. This quantification is then used to adjust the parameters of the aforementioned dynamic model. For instance, parameters regarding the relationship between storage performance and data recovery time in the dynamic model are adjusted to more accurately reflect recovery performance under the current infrastructure conditions. Based on this adjusted dynamic model, the strategy decision module may reassess the current backup strategy, for example, considering migrating some highly sensitive data to a more stable storage region, or adjusting the backup window to avoid peak periods, thereby ensuring that compliance and cost control objectives for data backup and recovery are still met even when infrastructure performance fluctuates.

[0173] In another embodiment of this application, when quantifying the impact of infrastructure changes on data asset access performance, the method further includes:

[0174] 23351: For the aforementioned data assets, the access patterns are identified as bursty access patterns and / or periodic access patterns;

[0175] 23352: When an abrupt access pattern is identified, after an infrastructure change occurs, access performance indicators of the aforementioned data assets are collected at a preset high-frequency sampling interval until the fluctuation of the aforementioned access performance indicators within a preset time window that is not less than a preset window counting threshold does not exceed a preset stability threshold; the access performance indicators within the initial observation window after the change are compared with the access performance indicators within the same-length reference window before the change to obtain the instantaneous difference, and the aforementioned instantaneous difference is taken as the instantaneous impact degree;

[0176] 23353: When a periodic access pattern is identified, after an infrastructure change occurs, access performance indicators of the above-mentioned data assets are continuously collected for at least one complete period, and the access performance indicators within one complete period after the change are compared with the access performance indicators of the corresponding period before the change. The periodic average difference is calculated, and the above-mentioned periodic average difference is used as the degree of continuous impact.

[0177] 23354: Combining the instantaneous impact level and the continuous impact level mentioned above, a quantitative result is obtained on the impact of infrastructure changes on the access performance of the data assets mentioned above. The quantitative result is used to correct the parameters of the dynamic model mentioned above, and / or for policy decisions based on the dynamic model mentioned above.

[0178] Specifically, identifying access patterns for data assets can be based on historical access logs, business system configuration information, or by analyzing real-time access data using machine learning algorithms. Bursting access patterns typically manifest as high-concurrency read / write requests occurring within a short period, such as generating reports at the end of the month or processing data during sudden marketing campaigns. Periodic access patterns, on the other hand, are recurring access behaviors at fixed time intervals, such as daily data synchronization, weekly data backups, and monthly bill generation.

[0179] A preset high-frequency sampling interval refers to continuously collecting performance metrics at short time intervals (e.g., seconds or milliseconds) after infrastructure changes occur to capture rapid changes in sudden access patterns. Preset window counting thresholds and preset stability thresholds are used to determine whether performance has stabilized, avoiding premature termination of sampling due to short-term fluctuations. The calculation of instantaneous differences aims to quickly assess the immediate impact of infrastructure changes on sudden access. For periodic access patterns, continuously collecting performance metrics for at least one complete cycle ensures comprehensive coverage of inherent access patterns, thereby calculating a more representative periodic average difference to reflect the sustained impact of changes on long-term, regular access. Finally, by combining the degree of instantaneous impact and the degree of sustained impact, a more comprehensive and accurate quantitative result can be obtained, which can more finely guide parameter correction and strategy decisions in dynamic models.

[0180] This application also discloses an automatic data backup strategy generation system, including:

[0181] Information acquisition module 1 is used to acquire and quantify the input information for strategy generation. The input information for strategy generation includes compliance regulations of external data and cost control objectives of the enterprise.

[0182] Asset assessment module 2 is used to assess the policy-related attributes of data assets. These attributes include data sensitivity, importance, current backup policy status, potential compliance risks, and recovery performance requirements. Based on the policy, it generates input information to estimate the resource investment costs required to meet different preset protection levels.

[0183] Model building module 3 is used to build and update a dynamic model reflecting the trade-offs between policies based on policy-generated input information, policy-related attributes, and resource input costs. This module can be implemented as a model management service, containing various optimization algorithms (e.g., multi-objective optimization algorithms, machine learning models) for building and adjusting the dynamic model based on input data.

[0184] The strategy decision module 4 is used to make strategy decisions based on a dynamic model. The strategy decision is based on compliance regulations and real-time resource status to select backup strategies. Real-time resource status includes the available capacity of storage devices, network data transmission capabilities, and storage device input / output performance.

[0185] The strategy deployment and monitoring module 5 is used to deploy backup strategies and continuously monitor the execution effect of backup strategies, using the actual running data of the execution effect as feedback to adjust the parameters of the dynamic model.

[0186] The above description is merely an embodiment of this application and is not intended to limit the scope of protection of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of protection of this application.

Claims

1. A method for automatically generating data backup strategies, characterized in that, include: Acquire and quantify strategy generation input information, which includes compliance regulations for external data and cost control targets within the enterprise; Assess the policy-related attributes of data assets, including data sensitivity, importance, current backup policy status, potential compliance risks, and recovery performance requirements, and estimate the resource investment costs to meet different preset protection levels based on the policy-generated input information. Based on the input information generated by the strategy, the relevant attributes of the strategy, and the resource investment cost, a dynamic model reflecting the trade-off relationship of the strategy is constructed and updated in real time. The strategy decision is made based on the dynamic model. The strategy decision is based on the compliance regulations and real-time resource status to select the backup strategy. The real-time resource status includes the available capacity of the storage device, the network data transmission capacity, and the input / output performance of the storage device. Deploy the backup strategy and continuously monitor its execution effect, using the actual running data of the execution effect as feedback to adjust the parameters of the dynamic model; The process of making strategy decisions based on the dynamic model and deploying the backup strategy also includes: Collect and store the micro-pattern characteristics of key network transmission paths, including packet delay, delay jitter, packet interval time distribution, transmission control protocol window size change pattern, and throughput distribution, to form a normal operation baseline; When a large-scale data migration or backup task is started, the micro-pattern characteristics of the task data stream are sampled and analyzed in real time to obtain the task stream feature sequence. The task flow feature sequence is compared with the normal operating baseline to calculate the deviation index; When the deviation index continues to exceed the preset deviation threshold within a preset time window, it is determined that there is a non-obvious anomaly, triggering an anomaly warning, and the anomaly event is provided as input to the strategy decision for strategy adjustment. The preset emergency diversion plans are prioritized based on the satisfaction of the recovery point target, the cost increment, and the resource availability. The emergency diversion plans include alternative transmission paths and backup scheduling schemes. Based on the abnormal event, the highest priority alternative is recommended from the emergency diversion plan, the alternative is executed, and the execution result of the alternative is used as feedback to adjust the parameters of the dynamic model.

2. The automatic data backup strategy generation method according to claim 1, characterized in that, The process of making strategy decisions based on the dynamic model also includes: Establish a performance isolation threshold library for shared infrastructure components. The performance isolation threshold library records the maximum number of concurrent data streams and / or input / output performance of each shared infrastructure component under different business loads, as well as the interference characteristics of specific types of data streams that cause performance interference to other services. Based on the task size, data stream type, and scheduling time of the candidate large-scale data migration or backup tasks, simulate the resource consumption pattern of the candidate schemes on the shared infrastructure components. The resource occupancy pattern is matched with the interference features in the performance isolation threshold library. When the matching result predicts that the candidate scheme will cause the resource occupancy of the shared infrastructure component to exceed the performance isolation threshold and / or overlap with the interference features, the candidate scheme is marked as a high-risk scheme, and the interference penalty cost is calculated. The interference penalty cost is included in the total cost of the candidate solution in the dynamic model, and based on the adjusted total cost and the real-time resource status, a backup strategy is selected from the candidate solutions that meet the compliance regulations.

3. The automatic data backup strategy generation method according to claim 1, characterized in that, The process of deploying the backup strategy and continuously monitoring its effectiveness also includes: Establish a business scenario recovery path diagram, which is used to map the internal structure, interdependencies and supporting application environment of data assets; During the business scenario recovery verification process, the complete path and corresponding time from data recovery to end-to-end functional recovery of the business system are tracked and recorded based on the business scenario recovery path diagram. Analyze the complete path and corresponding time, identify the key dependent links and bottlenecks in the actual recovery path, and quantify the impact of each key dependent link and bottleneck on the overall business recovery time. The overall business recovery time is compared with the recovery time target and / or recovery point target preset by the business department, and the deviation is calculated. When the deviation exceeds a preset time deviation threshold, the actual deviation between the actual recovery performance and the business expectation is identified, caused by changes in the internal structure, interdependencies, or application environment. The actual deviation is converted into a business scenario recovery benefit correction factor, which is used to adjust the benefit function parameters in the dynamic model.

4. The automatic data backup strategy generation method according to claim 1, characterized in that, The process of assessing the strategy-related attributes of data assets also includes: Continuously monitor and parse business process iteration information, which includes business system function updates, data processing process adjustments, and changes in user access behavior. Based on the business process iteration information, identify the affected data assets and their associated business scenarios; For the affected data assets, analyze the call frequency, data update frequency and data access path of the data assets in the business system, and determine their current business value and access mode by combining business system logs and user behavior data. Based on the impact of the business process iteration information on the data processing method, re-evaluate their compliance classification and obtain real-time detection results. The real-time detection results are compared with the policy-related attributes stored in the data asset metadata to calculate the attribute deviation. When the deviation of the attribute exceeds a preset attribute threshold, the strategy-related attributes of the data asset are automatically updated; Based on the updated policy-related attributes and the policy, input information is generated to re-estimate the resource investment cost required to meet different preset protection levels, and / or trigger real-time updates of the dynamic model.

5. The automatic data backup strategy generation method according to claim 4, characterized in that, The continuous monitoring and parsing of business process iteration information also includes: Integrate with enterprise service bus and / or message queue to listen for service update messages published by business systems; Based on the service update message, configure the log parsing adapter to convert heterogeneous update log information into a standardized data structure; Based on the standardized data structure, an iterative information dependency graph is constructed to identify update events with conflicting and / or sequential dependencies; Natural language processing technology is used to parse unstructured business documents and change requests, extract keywords and semantic associations related to business system function updates, data processing flow adjustments, and changes in user access behavior, and obtain semantic association results. By integrating the standardized data structure, the iterative information dependency graph, and the semantic association results, a comprehensive business process change report is generated for use when assessing the strategy-related attributes of data assets.

6. The automatic data backup strategy generation method according to claim 4, characterized in that, The process of real-time probing of the affected data assets also includes: Deploy infrastructure-level performance monitoring probes to collect real-time data on storage media read / write latency, IOPS, throughput, as well as network device bandwidth utilization, packet loss rate, and link jitter. Deploy audit probes at the security policy enforcement level to record execution logs of data anonymization operations and / or trigger events of access control rules in real time; The infrastructure metrics collected by the performance monitoring probe are correlated with the access patterns of the data assets to identify changes in data asset access performance caused by infrastructure changes. The security policy execution logs recorded by the audit probe are correlated with the compliance classification of the data assets to identify changes in data asset compliance risks caused by security policy changes. Based on the results of the correlation analysis, the real-time detection results of the business value, compliance classification, and / or access patterns of the data assets are revised. Based on the revised real-time detection results, combined with business system logs and user behavior data, the frequency of data asset calls, data update frequency and data access path in the business system are re-analyzed to determine its current business value and access pattern, and compliance classification is reassessed.

7. The automatic data backup strategy generation method according to claim 6, characterized in that, When performing correlation analysis between the infrastructure metrics collected by the performance monitoring probe and the access patterns of the data assets, the method also includes: The infrastructure indicator stream and data asset access pattern stream are transmitted in a streaming manner, with attached identification information, which includes at least a data asset identifier and a collection source identifier. During data collection, a timestamp is appended to each infrastructure metric sample and each data asset access event, and the infrastructure metric stream is aligned with the data asset access pattern stream based on the timestamp. A sliding time window is used to perform correlation analysis between the aligned infrastructure indicator stream and the data asset access pattern stream; After infrastructure indicators fluctuate beyond a preset fluctuation threshold, the trend of data asset access performance changes within a subsequent preset time window. When there is a continuous and consistent trend of change within a preset time window that is no less than a preset window count threshold, a causal relationship is determined. When the causal relationship is determined to exist, the impact of infrastructure changes on the access performance of the data assets is quantified, and the quantification results are used to correct the parameters of the dynamic model and / or for policy decisions based on the dynamic model.

8. The automatic data backup strategy generation method according to claim 7, characterized in that, When quantifying the impact of infrastructure changes on the access performance of the data assets, the following is also included: For the data assets, the access patterns are identified as bursty access patterns and / or periodic access patterns; When a sudden access pattern is identified, after an infrastructure change occurs, access performance indicators of the data assets are collected at a preset high-frequency sampling interval until the fluctuation range of the access performance indicators within a preset time window that is not less than a preset window counting threshold does not exceed a preset stability threshold; the access performance indicators within the initial observation window after the change are then... The instantaneous difference is obtained by comparing the access performance index within the reference window of the same length before the change, and the instantaneous difference is used as the instantaneous degree of impact. When a periodic access pattern is identified, after an infrastructure change occurs, the access performance indicators of the data assets are continuously collected for at least one complete cycle, and the access performance indicators in the complete cycle after the change are compared with the access performance indicators in the corresponding cycle before the change to calculate the periodic average difference, which is used as the degree of continuous impact. By combining the instantaneous impact level and the sustained impact level, a quantitative result is obtained of the impact of infrastructure changes on the access performance of the data assets. The quantitative result is used to correct the parameters of the dynamic model and / or for policy decisions based on the dynamic model.

9. A data backup strategy automatic generation system, characterized in that, include: The information acquisition module is used to acquire and quantify the strategy generation input information, which includes compliance regulations of external data and cost control targets of the enterprise. The asset assessment module is used to assess the policy-related attributes of data assets, including data sensitivity, importance, current backup policy status, potential compliance risks, and recovery performance requirements. Based on the policy, it generates input information to estimate the resource investment cost to meet different preset protection levels. The model building module is used to generate input information based on the strategy, the strategy-related attributes, and the resource investment cost, and to build and update a dynamic model that reflects the strategy trade-offs in real time. The strategy decision module is used to make strategy decisions based on the dynamic model. The strategy decision is based on the compliance regulations and real-time resource status to select a backup strategy. The real-time resource status includes the available capacity of the storage device, the network data transmission capacity, and the input / output performance of the storage device. The strategy deployment and monitoring module is used to deploy the backup strategy and continuously monitor the execution effect of the backup strategy, and use the actual running data of the execution effect as feedback to adjust the parameters of the dynamic model.

Images