A method and system for automatically managing disaster recovery processes in an intelligent computing center

Abstract

The present invention relates to the field of automatic management technology, specifically to an automatic management method and system for the disaster recovery process of an intelligent computing center, including the following steps: based on the software architecture of the intelligent computing center, through graph network analysis, identifying the key dependencies and data transmission paths between the architectures, comparing the difference between the current software running state and the fault standard, determining the fault source and isolating it, and generating a fault isolation record. The present invention improves the accuracy and speed of fault source location by applying graph network analysis, which not only makes fault isolation faster, but also reduces the downtime of the system and avoids the loss of large-scale data. Through the decision tree, automation and intelligence are introduced in the judgment of fault type and severity, optimizing the adaptability and accuracy of the repair process, and the real-time dynamic adjustment of resource allocation ensures the optimal performance of the system under various loads, improves resource utilization efficiency, and reduces operation and maintenance costs, thereby strengthening the disaster recovery process of the intelligent computing center.

Description

Technical Field

[0001] The present invention relates to the field of automatic management technology, and in particular to an automatic management method and system for a disaster recovery process of an intelligent computing center. Background Art

[0002] The field of automated management technology involves performing management tasks in an automated manner, thereby reducing human intervention and improving efficiency and reliability. This field is widely used in various IT environments, including but not limited to network management, system configuration, fault detection and recovery, and resource allocation. Automated management technology can automatically perform tasks through preset strategies, such as monitoring system performance, performing routine maintenance, and responding to faults in a timely manner.

[0003] Among them, the automatic management method of the disaster recovery process of the intelligent computing center refers to a systematic disaster recovery process implemented in the intelligent computing center. This process can automatically manage and respond to potential system crashes or data loss events. Its main purpose is to ensure that the data center can quickly and effectively resume operations when encountering natural disasters, technical failures or other emergencies, protect critical data from loss, and ensure service continuity.

[0004] Existing technologies cannot accurately locate faults under rapidly changing fault conditions, resulting in prolonged isolation and repair time. Traditional fault handling lacks sufficient intelligence and cannot make efficient fault type judgments and repair selections, increasing the risk of data loss. Resource management relies on static configurations and lacks the ability to adjust resource allocation based on real-time needs, which can easily lead to performance bottlenecks and service quality issues when resource demands change frequently. The consequences include increased operational risks, higher operation and maintenance costs, and service interruptions. Summary of the invention

[0005] The purpose of the present invention is to solve the shortcomings existing in the prior art and to propose an automatic management method and system for the disaster recovery process of an intelligent computing center.

[0006] In order to achieve the above object, the present invention adopts the following technical solution: a method for automatically managing the disaster recovery process of an intelligent computing center, comprising the following steps:

[0007] S1: Based on the software architecture of the intelligent computing center, through graph network analysis, identify the key dependencies and data transmission paths between architectures, compare the differences between the current software running status and the fault standard, determine the fault source and isolate it, and generate fault isolation records;

[0008] S2: Based on the fault isolation record, a decision tree is used to determine the fault type and severity, and a matching repair process is determined, including restarting the affected service or deploying data backup, and automatically scheduling resources for fault repair to obtain a repair confirmation configuration;

[0009] S3: Based on the repair confirmation configuration, real-time monitoring of software resource usage, dynamic adjustment of computing resources and network bandwidth allocation, minimization of operation interruption, and updating of resource configuration of the intelligent computing center according to current needs, generating resource allocation optimization information;

[0010] S4: Based on the resource allocation optimization information, check the status of the intelligent computing center after repair, verify the effectiveness of the resource configuration, determine whether there are new fault points, and update the operation and maintenance plan according to the verification results to generate a service stability evaluation result.

[0011] The present invention is improved in that the step of identifying the key dependency and the data transmission path is specifically as follows:

[0012] S111: Based on the software architecture of the intelligent computing center, collect software operation status and interaction data, including the call frequency and data transmission volume between software, and obtain an overview of software interaction;

[0013] S112: Based on the software interaction overview, the formula is used:

[0014] ;

[0015] Calculate the connection strength between software , and obtain the connection strength graph, where It's software To Software The calling frequency, is the amount of data transferred, and It is the maximum value of call frequency and data volume;

[0016] S113: Based on the connection strength graph, extract the software with the best connection strength as the key dependency, and determine the transmission path with the maximum data flow as the key path, so as to obtain the key dependency and transmission path information.

[0017] The present invention is improved in that the steps of determining the fault source and isolating it are specifically as follows:

[0018] S121: collecting operation performance data, including CPU usage, memory usage, network traffic, and disk I / O operations, comparing real-time data with fault standards, determining whether there are abnormalities, and generating preliminary fault detection records;

[0019] S122: Based on the preliminary fault detection record, use the formula:

[0020] ;

[0021] Calculate the fault source index ,in, Representative The deviation between the monitoring index and the fault standard, It is The weight coefficient of each monitoring indicator is is the number of monitoring indicators;

[0022] S123: According to the fault source index, an isolation process is automatically executed to cut off or limit the network access rights of the fault source, optimize the diffusion of the fault impact, and obtain a fault isolation record.

[0023] The present invention is improved in that the steps of determining the fault type and severity are specifically as follows:

[0024] S211: collecting key information of the fault isolation record, including the timestamp of the fault occurrence, the affected service type, the duration of the fault and the error code, to obtain a preliminary fault data summary;

[0025] S212: Based on the preliminary fault data summary, a decision tree is used to perform fault analysis, using the formula:

[0026] ;

[0027] Calculating the Fault Severity Index , get the fault assessment data, where is the sensitivity weight for each fault code, is the deviation value currently monitored, is the adjustment factor for the rate of change of deviation, Indicates the number of fault indicators, is the baseline sensitivity, is the current operating load factor;

[0028] S213: Based on the fault assessment data, by comparing and analyzing with known fault modes, the accuracy of the fault type and severity is determined to obtain a fault verification result.

[0029] The present invention is improved in that the step of automatically scheduling resources to repair the fault is specifically as follows:

[0030] S221: Retrieve a maintenance process that matches the fault type and severity, select a service restart process for a service interruption fault, and select a data backup process for a data corruption fault, to obtain a repair process plan;

[0031] S222: Based on the repair process plan, perform resource scheduling optimization using the formula:

[0032] ;

[0033] Optimized resource scheduling plan ,in, is the resource demand weight defined according to the fault type, is the basic amount of computing resources required for the repair operation, is an acceleration factor based on the severity of the fault, Indicates the amount of computing resources currently available. is the amount of spare resources available;

[0034] S223: Execute the optimized resource scheduling plan, monitor the implementation of service restart or data backup, verify whether the operation is successfully executed, and record the repair activities and status changes to obtain the repair confirmation configuration.

[0035] The present invention is improved in that the steps of dynamically adjusting the computing resources and network bandwidth allocation are specifically as follows:

[0036] S311: Implement real-time monitoring according to the repair confirmation configuration, collect software resource usage data, including CPU usage, memory usage, and network bandwidth consumption, and generate real-time resource usage information;

[0037] S312: Based on the real-time information on resource usage, analyze data and calculate resource allocation using the formula:

[0038] ;

[0039] Generate resource adjustment configuration ,in, Representative resources The priority weight of Represents the current utilization of resources. is the resource usage sensitivity adjustment factor, Represents the total available resources;

[0040] S313: Based on the resource adjustment configuration, dynamically adjust computing resources and network bandwidth allocation, optimize the performance bottleneck area, and continuously monitor the resource adjustment effect to generate resource adjustment execution results.

[0041] The present invention is improved in that the step of updating the resource configuration of the intelligent computing center is specifically:

[0042] S321: Monitor the current resource usage of the intelligent computing center, collect data, including resource consumption and identified performance bottlenecks and resource waste areas, and generate a resource usage table;

[0043] S322: Based on the resource usage table, balance resource usage and waste, using the formula:

[0044] ;

[0045] Get resource allocation optimization information, where: and They represent the utilization rate and waste rate of resource i respectively, and is the weight of the utilization rate and waste rate of resource i, It is the result of resource optimization. Is the total number of resource types.

[0046] The present invention is improved in that the steps of obtaining the service stability evaluation result are specifically as follows:

[0047] S411: Based on the resource allocation optimization information, monitor multiple key performance indicators of the intelligent computing center, including CPU usage, memory usage, disk IO and network bandwidth usage, and obtain current operation status information;

[0048] S412: Analyze the current running status information using the formula:

[0049] ;

[0050] Evaluate resource configuration effectiveness and performance bottlenecks, and obtain performance and configuration evaluation results, where: is the performance efficiency index, is the total number of resource types monitored by the intelligent computing center. Indicates The utilization rate of the resource, It is The weight factor of the resource type, Representative The maximum ideal utilization rate of a resource;

[0051] S413: Verify the performance and configuration evaluation results, compare historical maintenance records and fault data, determine whether new performance bottlenecks or fault points are known, update the operation and maintenance plan based on the comparison results, and generate service stability evaluation results.

[0052] An automatic management system for disaster recovery process of an intelligent computing center, the system comprising:

[0053] The fault isolation module is based on the intelligent computing center software architecture. Through graph network analysis, it determines the key dependencies and data transmission paths, compares the difference between the current software running status and the fault standard, determines the fault source and isolates it, and generates fault isolation records.

[0054] The repair scheduling module uses a decision tree to determine the fault type and severity based on the fault isolation record, determines a matching repair process, including restarting or performing data backup, and automatically schedules resources for fault repair and generates a repair confirmation configuration;

[0055] The resource adjustment module confirms the configuration based on the repair, monitors the usage of software resources, dynamically adjusts computing resources and network bandwidth, minimizes operation interruptions, and updates the resource configuration of the intelligent computing center according to current needs, generating resource allocation optimization information;

[0056] The status verification module checks the status of the intelligent computing center after repair based on the resource allocation optimization information, verifies the validity of resource configuration, identifies potential new fault points, and generates status verification results;

[0057] The operation and maintenance process optimization module updates the operation and maintenance plan based on the status verification result, and continuously verifies the continuity and stability of the service to obtain a service stability evaluation result.

[0058] Compared with the prior art, the advantages and positive effects of the present invention are:

[0059] In the present invention, by applying graph network analysis, the accuracy and speed of fault source locating are improved, which not only makes fault isolation faster, but also reduces system downtime and avoids large-scale data loss. Automation and intelligence are introduced in the judgment of fault type and severity through decision trees, and the adaptability and accuracy of the repair process are optimized. Real-time dynamic adjustment of resource allocation ensures the optimal performance of the system under various loads, improves resource utilization efficiency, and reduces operation and maintenance costs, thereby strengthening the disaster recovery process of the intelligent computing center, effectively protecting key data and ensuring business continuity and service stability. BRIEF DESCRIPTION OF THE DRAWINGS

[0060] Figure 1 A flowchart of a method for automatically managing the disaster recovery process of an intelligent computing center is proposed for the present invention;

[0061] Figure 2 A flowchart for identifying key dependencies and data transmission paths in the present invention;

[0062] Figure 3 A flow chart for determining and isolating a fault source in the present invention;

[0063] Figure 4 This is a flow chart for determining the fault type and severity in the present invention;

[0064] Figure 5 A flow chart of automatically scheduling resources to repair faults in the present invention;

[0065] Figure 6 A flowchart of dynamic adjustment of computing resources and network bandwidth allocation in the present invention;

[0066] Figure 7 A flowchart for updating resource configuration of the intelligent computing center in the present invention;

[0067] Figure 8 This is a flow chart for obtaining the service stability evaluation results in the present invention. DETAILED DESCRIPTION

[0068] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0069] In the description of the present invention, it should be understood that the terms "length", "width", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside", etc., indicating positions or positional relationships, are based on the positions or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as limiting the present invention. In addition, in the description of the present invention, the meaning of "multiple" is two or more, unless otherwise clearly and specifically defined.

[0070] Example: See Figure 1 The present invention provides a technical solution: a method for automatically managing the disaster recovery process of an intelligent computing center, comprising the following steps:

[0071] S1: Based on the software architecture of the intelligent computing center, through graph network analysis, identify the key dependencies and data transmission paths between architectures, compare the differences between the current software running status and the fault standard, determine the fault source and isolate it, and generate fault isolation records;

[0072] S2: Based on the fault isolation records, the decision tree is used to determine the fault type and severity, and the matching repair process is determined, including restarting the affected services or deploying data backups, and automatically scheduling resources for fault repair to obtain repair confirmation configuration;

[0073] S3: Based on the repair confirmation configuration, real-time monitoring of software resource usage, dynamic adjustment of computing resources and network bandwidth allocation, minimization of operation interruption, and update of intelligent computing center resource configuration according to current needs, generating resource allocation optimization information;

[0074] S4: Based on the resource allocation optimization information, check the status of the intelligent computing center after repair, verify the effectiveness of the resource configuration, determine whether there are new fault points, and update the operation and maintenance plan according to the verification results to generate service stability evaluation results.

[0075] Fault isolation records include the time of fault occurrence, fault impact scope, and isolation measures. Repair confirmation configuration includes the selected repair process, resource scheduling plan, and estimated recovery time. Resource allocation optimization information includes adjusted CPU utilization, memory usage, and network bandwidth configuration. Service stability assessment results include operation stability information, service availability after recovery, and maintenance cycle.

[0076] See also Figure 2 ,The steps for identifying key dependencies and data transmission paths are as follows:

[0077] S111: Based on the software architecture of the intelligent computing center, collect software operation status and interaction data, including the call frequency and data transmission volume between software, and obtain an overview of software interaction;

[0078] In the process of collecting all software and their interaction data, based on the on-site server and cloud storage data, the operation logs and data transmission records of each software are obtained through API calls. The records list the frequency of calls between software and the size of data transmission. Then, key information is extracted from the logs, including timestamp, source software identifier, target software identifier and the amount of data transmitted. The data can be used to preliminarily construct an interaction map between software. The map reflects the communication frequency and data flow between software, generates an overview of software interactions, and provides a data basis for subsequent dependency analysis.

[0079] S112: Based on the software interaction overview, the formula is used:

[0080] ;

[0081] Calculate the connection strength between software , and obtain the connection strength graph, where It's software To Software The calling frequency, is the amount of data transferred, and It is the maximum value of call frequency and data volume;

[0082] The frequency of calls from software A to software B The maximum call frequency is 100 times. The data transfer volume from software A to software B is 500 times The maximum data transfer volume is 200MB The value is 1000MB. Substituting it into the formula, we get:

[0083] ;

[0084] ;

[0085] ;

[0086] The result shows that the connection strength between software A and software B is 0.69, indicating that their connection relationship in the system is relatively important and helps to further determine the key data transmission path of the system.

[0087] S113: based on the connection strength graph, extract the software with the best connection strength as the key dependency, and determine the transmission path with the highest data flow as the key path, so as to obtain the key dependency and transmission path information;

[0088] In the process of identifying key dependencies and data transmission paths from the connection strength graph, according to the software connection strength graph, the software pairs whose strength values ​​exceed the preset threshold are screened out. The software pairs represent key dependencies. Then, the data flow between the software pairs is further analyzed, especially those paths with large and frequent data traffic, so as to locate the transmission path with the largest data traffic and mark the path as the key data transmission path. The key path is usually related to performance bottlenecks and potential risk points, and has a significant impact on system stability and efficiency.

[0089] See also Figure 3 , the steps to determine the source of the fault and isolate it are as follows:

[0090] S121: collecting operation performance data, including CPU usage, memory usage, network traffic, and disk I / O operations, comparing real-time data with fault standards, determining whether there are abnormalities, and generating preliminary fault detection records;

[0091] In the process of monitoring performance data, key performance indicators such as CPU usage, memory usage, network traffic and disk I / O operations are captured regularly. By comparing them with pre-defined fault standards in real time, each indicator is analyzed to see if it exceeds the predetermined threshold. The analysis helps determine whether a fault alarm is triggered, thereby generating a preliminary fault detection report. The report identifies all detected abnormal software and its corresponding performance indicators, provides necessary data support for further analysis of the source of the fault, and prepares isolation measures.

[0092] S122: Based on the preliminary fault detection records, use the formula:

[0093] ;

[0094] Calculate the fault source index ,in, Representative The deviation between the monitoring index and the fault standard, It is The weight coefficient of each monitoring indicator is is the number of monitoring indicators.

[0095] There are three monitoring indicators of deviation , , , the corresponding weight , , , the calculation process is:

[0096] ;

[0097] ;

[0098] ;

[0099] The result shows that the overall failure index is 0.7, indicating that there is a high risk of failure in the system and isolation measures are needed to avoid further system impacts.

[0100] S123: According to the fault source index, the isolation process is automatically executed to cut off or limit the network access rights of the fault source, optimize the diffusion of the fault impact, and obtain the fault isolation record;

[0101] When the monitored performance indicators are abnormal and exceed the failure standard, an alarm is automatically triggered and all relevant data is recorded. The data includes the CPU overload time, the duration of memory exceeding the normal range, and the specific circumstances of the network traffic abnormality. Through the monitoring of indicators, potential faulty software can be located. According to the comparative analysis of the monitoring data and the failure standard, which software has faulty behavior can be identified. Focus on those software that have the greatest impact on system stability, such as key database servers or core application servers, to ensure that the operating status of key components is monitored in real time, so that after confirming the source of the fault, a quick response can be made to cut off or limit the network access rights of the faulty software and optimize the spread of the fault impact. All decisions are recorded in the fault isolation record to provide a basis for subsequent fault analysis and prevention, which not only improves the reliability of the system, but also reduces the potential risk of business interruption.

[0102] See also Figure 4 , the specific steps for determining the fault type and severity are as follows:

[0103] S211: collecting key information of the fault isolation record, including the timestamp of the fault occurrence, the affected service type, the duration of the fault, and the error code, to obtain a preliminary fault data summary;

[0104] Collect key information from fault isolation records, including the timestamp of the fault, the type of affected service, the duration of the fault, and the error code in the log. After extracting the key information, analyze the specific performance of each data, such as whether the frequency of fault occurrence shows a specific periodicity, whether the error code matches the fault mode in the historical data, and whether the duration of the monitored fault exceeds the preset allowable range. The data points help build a comprehensive fault data summary, which will be analyzed to determine the scope and severity of the fault. The generated fault data summary will be used for further fault assessment and diagnosis processes.

[0105] S212: Based on the preliminary fault data summary, a decision tree is used to perform fault analysis using the formula:

[0106] ;

[0107] Calculating the Fault Severity Index , get the fault assessment data, where is the sensitivity weight for each fault code, is the deviation value currently monitored, is the adjustment factor for the rate of change of deviation, Indicates the number of fault indicators, is the baseline sensitivity, adjusted according to the overall system stability, is the current operating load factor, which is used to adjust the baseline sensitivity of fault assessment;

[0108] There are three fault indicators: , and , corresponding to the weight , and , adjustment factor , and , system operation load , baseline sensitivity , then the calculation is as follows:

[0109] ;

[0110] ;

[0111] ;

[0112] ;

[0113] The result shows that the severity of the current fault is 3.22, indicating that the system needs to take immediate repair measures to prevent the fault from further expanding.

[0114] S213: Based on the fault assessment data, by comparing and analyzing with known fault modes, the accuracy of the fault type and severity is determined to obtain a fault verification result;

[0115] Verify the evaluation results by matching them with the fault patterns in the historical database, comparing the similarity of fault type, frequency and severity, analyzing whether the evaluation results conform to the established fault patterns, and confirming their accuracy. During the matching process, similar faults in the historical data are introduced for comparison to ensure the consistency and reliability of the evaluation results. In addition, the system load at the time of the fault is analyzed for similarity with the load in the historical records to confirm whether the load change has a direct impact on the fault. After comparison and verification, the type and severity of the fault are confirmed.

[0116] See also Figure 5 ,The specific steps for automatically scheduling resources for fault repair are:

[0117] S221: Retrieve a maintenance process that matches the fault type and severity. For a service interruption fault, select a service restart process. For a data corruption fault, select a data backup process to obtain a repair process plan.

[0118] Retrieve the maintenance process that matches the type and severity of the fault, determine the repair strategy suitable for service interruption and data corruption, and when selecting the restart service process, first confirm the scope and duration of the service interruption to ensure that the restart action will not affect other dependent services. The data backup process needs to confirm the size of the affected data and the availability of backup storage. The plan will list the required resources such as computing power and storage space. The steps ensure that the repair measures are both quick and will not introduce new problems. The entire process requires the system to strictly monitor and schedule the use and impact of various resources in order to quickly and effectively restore services and protect data integrity, and obtain a repair process plan.

[0119] S222: Based on the repair process plan, optimize resource scheduling using the formula:

[0120] ;

[0121] Optimized resource scheduling plan ,in, is the resource demand weight defined according to the fault type, is the basic amount of computing resources required for the repair operation, is an acceleration factor based on the severity of the fault, Indicates the amount of computing resources currently available. is the amount of spare resources available;

[0122] There is a failure that requires restarting the service, and its resource demand weight , the required computing resources , acceleration factor , currently available resources , spare resources , the calculation process is:

[0123] ;

[0124] ;

[0125] ;

[0126] ;

[0127] The result shows that the amount of resources allocated to handle the current fault is 16, which ensures that the fault can be handled promptly and effectively.

[0128] S223: Execute the optimized resource scheduling plan, monitor the implementation of service restart or data backup, verify whether the operation is successfully executed, and record the repair activities and status changes to obtain the repair confirmation configuration;

[0129] Monitor the implementation process of restarting services or data backup to ensure that the operation not only complies with the predetermined execution plan but can also be completed successfully. Each step of the execution process will be recorded, including the start and end time, resource usage, and any exceptions that occur. The records help confirm the successful execution of the operation and provide important diagnostic information for potential failures. Get repair confirmation configuration, ensure that all changes and repair activities are accurately recorded, and provide a complete status change and activity log.

[0130] See also Figure 6 ,The specific steps for dynamic adjustment of computing resources and network bandwidth allocation are as follows:

[0131] S311: Implement real-time monitoring according to the repair confirmation configuration, collect software resource usage data, including CPU usage, memory usage, and network bandwidth consumption, and generate real-time resource usage information;

[0132] Implement real-time monitoring to collect software resource usage data, regularly capture key indicators of CPU usage, memory usage, and network bandwidth consumption, aggregate and record data through automated monitoring, and convert information into real-time resource usage status, so that you can intuitively view the current resource load. Through real-time data updates, detailed real-time resource usage information is automatically generated. This information is used for subsequent resource adjustments and optimization decisions to ensure the efficiency and accuracy of resource allocation, which not only improves the transparency of resource management, but also optimizes resource utilization.

[0133] S312: Based on the real-time information of resource usage, analyze the data and calculate resource allocation using the formula:

[0134] ;

[0135] .Generate resource adjustment configuration ,in, Representative resources The priority weight of Represents the current usage of the resource. is the resource usage sensitivity adjustment factor, Represents the total available resources;

[0136] There are three resource types, with usage rates as follows: , , The priority weights are , , , adjustment factor , total resources , the calculation process is:

[0137] ;

[0138] ;

[0139] ;

[0140] ;

[0141] The result shows that the current adjustment value is 0.045, which means that based on the current resource utilization, dynamic allocation of resources will increase efficiency by about 4.5%, helping to reduce the pressure on high-load resources and optimize overall resource utilization.

[0142] S313: Based on the resource adjustment configuration, dynamically adjust the computing resources and network bandwidth allocation, optimize the performance bottleneck area, continuously monitor the resource adjustment effect, and generate the resource adjustment execution result;

[0143] Based on the collected real-time information on resource usage, resource adjustment configuration is performed, and the data is used to analyze the current performance bottleneck area. Subsequently, based on the performance data and the predetermined optimization strategy, the allocation of computing resources and network bandwidth is dynamically adjusted. The optimal resource allocation is calculated based on the real-time data, and the effect of the adjustment is continuously monitored to ensure that the implemented measures can effectively solve the performance bottleneck. Detailed results of the resource adjustment execution are generated. The results record the performance improvement before and after the adjustment and the improvement of resource utilization efficiency, providing a basis for subsequent resource management and adjustment decisions.

[0144] See also Figure 7 , the specific steps to update the resource configuration of the intelligent computing center are:

[0145] S321: monitor the current resource usage of the intelligent computing center, collect data, including resource consumption and identified performance bottlenecks and resource waste areas, and generate a resource usage table;

[0146] Monitor the resource usage of the intelligent computing center and collect real-time usage data of various resources, including but not limited to CPU usage, memory usage, and network bandwidth consumption. By setting monitoring points and data collection frequency, it is able to continuously track the dynamic changes in resource consumption. The collected data is processed by analysis software to identify performance bottleneck areas and waste points in resource usage, and is updated in a timely manner to the resource usage table. The table records the consumption details and performance status of each resource, which becomes the basis for optimizing resource allocation and adjusting plans.

[0147] S322: Based on the resource usage table, balance resource usage and waste using the formula:

[0148] ;

[0149] Get resource allocation optimization information, where: and They represent the utilization rate and waste rate of resource i respectively, and is the weight of the utilization rate and waste rate of resource i, is a positive number, It is the result of resource optimization. is the total number of resource types;

[0150] The following data was collected:

[0151] , , and ;

[0152] Weight , , and ;

[0153] ;

[0154] The calculation is as follows:

[0155] ;

[0156] ;

[0157] ;

[0158] ;

[0159] The results show that there are major deficiencies in the current resource allocation, and further adjustments are needed to reduce resource waste and improve utilization efficiency.

[0160] See also Figure 8 , the specific steps for obtaining the service stability evaluation results are:

[0161] S411: Based on the resource allocation optimization information, monitor multiple key performance indicators of the intelligent computing center, including CPU usage, memory usage, disk IO and network bandwidth usage, and obtain current operation status information;

[0162] Based on the latest resource allocation optimization information, the intelligent computing center starts real-time monitoring of key performance indicators, including CPU utilization, memory usage, disk IO, and network bandwidth usage. During the monitoring process, performance data is automatically collected and processed in real time, and the current data is compared with the historical performance baseline to identify indicators that deviate from the normal range. The data is integrated into a comprehensive operation status report. The report lists the real-time usage of each resource in detail, points out performance bottlenecks and areas of resource waste. This information is crucial for the operation and maintenance team to evaluate the current status and optimize resource allocation. The generated operation status information is then used for performance evaluation and the formulation of fault prevention strategies.

[0163] S412: Analyze the current operation status information using the formula:

[0164] ;

[0165] Evaluate resource configuration effectiveness and performance bottlenecks, and obtain performance and configuration evaluation results, where: is the performance efficiency index, is the total number of resource types monitored by the intelligent computing center. Indicates The utilization rate of the resource, It is The weight factor of the resource type, Representative The maximum ideal utilization rate of a resource;

[0166] There are three types of resources, CPU, memory and disk, and the corresponding actual usage , , , the corresponding resource importance weight , , , ideal maximum utilization , , ;

[0167] The calculation is as follows:

[0168]

[0169] ;

[0170] ;

[0171] The results show that the resource allocation efficiency of the intelligent computing center is 0.806, or 80.6%, indicating that the overall efficiency of resource utilization is high but there is slight room for adjustment to achieve the ideal allocation efficiency.

[0172] S413: Verify the performance and configuration evaluation results, compare historical maintenance records and fault data, determine whether the newly discovered performance bottlenecks or fault points are known, update the operation and maintenance plan based on the comparison results, and generate service stability evaluation results;

[0173] After completing the preliminary assessment of the performance and configuration of the intelligent computing center, compare the latest maintenance records with known fault data. The comparison operation involves matching and analyzing the current performance bottlenecks and potential fault points with historical data, checking whether the newly discovered problems are known faults or caused by recent configuration changes, and updating the operation and maintenance plan based on the comparison results. Readjust resource allocation to address identified problems and generate service stability assessment results that record all current and potential problems.

[0174] An automatic management system for disaster recovery process of an intelligent computing center, the system comprising:

[0175] The fault isolation module is based on the intelligent computing center software architecture. Through graph network analysis, it determines the key dependencies and data transmission paths, compares the difference between the current software running status and the fault standard, determines the fault source and isolates it, and generates fault isolation records.

[0176] The repair scheduling module uses a decision tree to determine the fault type and severity based on the fault isolation record, determines the matching repair process, including restarting or performing data backup, and automatically schedules resources for fault repair and generates a repair confirmation configuration;

[0177] The resource adjustment module confirms the configuration based on the repair, monitors the usage of software resources, dynamically adjusts computing resources and network bandwidth, minimizes operation interruptions, and updates the resource configuration of the intelligent computing center according to current needs, generating resource allocation optimization information;

[0178] The status verification module checks the status of the intelligent computing center after repair based on the resource allocation optimization information, verifies the validity of resource configuration, identifies potential new fault points, and generates status verification results;

[0179] The operation and maintenance process optimization module updates the operation and maintenance plan based on the status verification results, and continuously verifies the continuity and stability of the service to obtain the service stability evaluation results.

[0180] The above are only preferred embodiments of the present invention and are not intended to limit the present invention in other forms. Any technician familiar with the profession may use the technical contents disclosed above to change or modify them into equivalent embodiments with equivalent changes and apply them to other fields. However, any simple modification, equivalent change and modification made to the above embodiments based on the technical essence of the present invention without departing from the technical solution of the present invention still falls within the protection scope of the technical solution of the present invention.

Claims

1. A method for automatically managing the disaster recovery process of an intelligent computing center, characterized in that: The following steps are involved: Based on the software architecture of the intelligent computing center, through graph network analysis, key dependencies and data transmission paths between architectures are identified, the difference between the current software running status and the fault standard is compared, the fault source is determined and isolated, and a fault isolation record is generated; Based on the fault isolation record, a decision tree is used to determine the fault type and severity, and a matching repair process is determined, including restarting the affected service or deploying data backup, and resources are automatically scheduled for fault repair to obtain a repair confirmation configuration; Based on the repair confirmation configuration, real-time monitoring of software resource usage, dynamic adjustment of computing resources and network bandwidth allocation, minimization of operational interruptions, and updating of resource configuration of the intelligent computing center according to current needs, generating resource allocation optimization information; Based on the resource allocation optimization information, check the status of the intelligent computing center after repair, verify the effectiveness of resource configuration, determine whether there are new fault points, and update the operation and maintenance plan according to the verification results to generate service stability evaluation results; The steps of identifying the key dependencies and data transmission paths are specifically as follows: Based on the software architecture of the intelligent computing center, collect software operation status and interaction data, including the call frequency and data transmission volume between software, and obtain an overview of software interaction; Based on the software interaction overview, the formula is adopted: ； Calculate the connection strength between software , and obtain the connection strength graph, where It's software To Software The calling frequency, is the amount of data transferred, and It is the maximum value of call frequency and data volume; Based on the connection strength graph, the software with the best connection strength is extracted as the key dependency, and the transmission path with the maximum data flow is determined as the key path, so as to obtain the key dependency and transmission path information.

2. The automatic management method of the disaster recovery process of the intelligent computing center according to claim 1 is characterized in that: The steps of determining the fault source and isolating it are specifically as follows: Collects operational performance data, including CPU usage, memory usage, network traffic, and disk I / O operations, compares real-time data with fault criteria to determine if anomalies exist, and generates preliminary fault detection records; Based on the preliminary fault detection record, the formula is used: ； Calculate the fault source index ,in, Representative The deviation between the monitoring index and the fault standard, It is The weight coefficient of each monitoring indicator is is the number of monitoring indicators; According to the fault source index, the isolation process is automatically executed, the network access rights of the fault source are cut off or restricted, the diffusion of the fault impact is optimized, and the fault isolation record is obtained.

3. The automatic management method of the intelligent computing center disaster recovery process according to claim 1 is characterized in that: The steps for determining the fault type and severity are specifically as follows: Collect key information of the fault isolation record, including the timestamp of the fault occurrence, the affected service type, the duration of the fault and the error code, to obtain a preliminary fault data summary; Based on the preliminary fault data summary, a decision tree is used to perform fault analysis using the formula: ； Calculating the Fault Severity Index , get the fault assessment data, where is the sensitivity weight for each fault code, is the deviation value currently monitored, is the adjustment factor for the rate of change of deviation, Indicates the number of fault indicators, is the baseline sensitivity, is the current operating load factor; Based on the fault assessment data, by comparing and analyzing with known fault modes, the accuracy of the fault type and severity is judged to obtain a fault verification result.

4. The automatic management method of the disaster recovery process of the intelligent computing center according to claim 1 is characterized in that: The steps of automatically scheduling resources to repair faults are specifically as follows: Retrieve the maintenance process that matches the fault type and severity, select the service restart process for the service interruption fault, and select the data backup process for the data corruption fault, and obtain the repair process plan; Based on the repair process plan, resource scheduling optimization is performed using the formula: ； Optimized resource scheduling plan ,in, is the resource demand weight defined according to the fault type, is the basic amount of computing resources required for the repair operation, is an acceleration factor based on the severity of the fault, Indicates the amount of computing resources currently available. is the amount of spare resources available; Execute the optimized resource scheduling plan, monitor the implementation of service restart or data backup, verify whether the operation is successfully executed, and record the repair activities and status changes to obtain the repair confirmation configuration.

5. The automatic management method of disaster recovery process of intelligent computing center according to claim 1 is characterized in that: The steps of dynamically adjusting the computing resources and network bandwidth allocation are specifically as follows: Implement real-time monitoring according to the repair confirmation configuration, collect software resource usage data, including CPU usage, memory usage, and network bandwidth consumption, and generate real-time resource usage information; Based on the real-time information of resource usage, the data is analyzed and resource allocation is calculated using the formula: ； Generate resource adjustment configuration ,in, Representative resources The priority weight of Represents the current utilization of resources. is the resource usage sensitivity adjustment factor, Represents the total available resources; Based on the resource adjustment configuration, the computing resources and network bandwidth allocation are dynamically adjusted, the performance bottleneck area is optimized, and the resource adjustment effect is continuously monitored to generate the resource adjustment execution result.

6. The automatic management method of disaster recovery process of intelligent computing center according to claim 1 is characterized in that: The steps of updating the resource configuration of the intelligent computing center are specifically as follows: Monitor the current resource usage of the intelligent computing center, collect data, including resource consumption and identified performance bottlenecks and resource waste areas, and generate a resource usage table; Based on the resource usage table, balance resource usage and waste, using the formula: ； Get resource allocation optimization information, where: and They represent the utilization rate and waste rate of resource i respectively, and is the weight of the utilization rate and waste rate of resource i, It is the result of resource optimization. Is the total number of resource types.

7. The automatic management method of disaster recovery process of intelligent computing center according to claim 1 is characterized in that: The steps for obtaining the service stability evaluation result are specifically as follows: Based on the resource allocation optimization information, monitor multiple key performance indicators of the intelligent computing center, including CPU usage, memory usage, disk IO and network bandwidth usage, and obtain current operating status information; The current running status information is analyzed using the formula: ； Evaluate resource configuration effectiveness and performance bottlenecks, and obtain performance and configuration evaluation results, where: is the performance efficiency index, is the total number of resource types monitored by the intelligent computing center. Indicates The utilization rate of the resource, It is The weight factor of the resource type, Representative The maximum ideal utilization rate of a resource; The performance and configuration evaluation results are verified, historical maintenance records and fault data are compared, it is determined whether new performance bottlenecks or fault points are known, the operation and maintenance plan is updated based on the comparison results, and the service stability evaluation results are generated.

8. An automatic management system for disaster recovery process of intelligent computing center, characterized in that: According to any one of claims 1 to 7, the method for automatically managing the disaster recovery process of an intelligent computing center is implemented, and the system includes: The fault isolation module is based on the intelligent computing center software architecture. Through graph network analysis, it determines the key dependencies and data transmission paths, compares the difference between the current software running status and the fault standard, determines the fault source and isolates it, and generates fault isolation records. The repair scheduling module uses a decision tree to determine the fault type and severity based on the fault isolation record, determines a matching repair process, including restarting or performing data backup, and automatically schedules resources for fault repair and generates a repair confirmation configuration; The resource adjustment module confirms the configuration based on the repair, monitors the usage of software resources, dynamically adjusts computing resources and network bandwidth, minimizes operation interruptions, and updates the resource configuration of the intelligent computing center according to current needs, generating resource allocation optimization information; The status verification module checks the status of the intelligent computing center after repair based on the resource allocation optimization information, verifies the validity of resource configuration, identifies potential new fault points, and generates status verification results; The operation and maintenance process optimization module updates the operation and maintenance plan based on the status verification result, and continuously verifies the continuity and stability of the service to obtain a service stability evaluation result.

Images