Methods, apparatus, equipment, and storage media for processing network cutover tasks

Abstract

This application discloses a method, apparatus, device, and storage medium for processing network cutover tasks. By receiving network cutover requests initiated by operators, the method obtains the operation event logs and configuration files of the target network devices based on the requests. Then, it automatically generates and executes the corresponding cutover script using a pre-built script generation model, avoiding the errors and lack of standardization issues associated with manually written scripts. Furthermore, it identifies high-risk instructions in the cutover script using a pre-built knowledge graph and determines the cutover risk assessment report level based on these instructions, providing feedback to the operators. This achieves automatic identification and quantitative assessment of high-risk operations during the cutover process. This application can improve the automation level and risk control capabilities of the cutover process, reduce the probability of cutover failure due to human error, and thus improve the overall efficiency and security of network operation and maintenance. The technical solution of this application can be widely applied in the field of network technology.

Description

Technical Field

[0001] This application relates to the field of network technology, and in particular to a method, apparatus, device, and storage medium for processing network cutover tasks. Background Technology

[0002] With the rapid evolution of communication networks and the continuous expansion of business scale, network operation and maintenance faces increasingly complex challenges. Network cutover, as a routine task in operator network operation and maintenance, involves multiple professional fields such as wireless, transmission, data, access, and power environment. The cutover processes and operation methods of different professional fields vary significantly, but all require the smooth migration of the system from the old network architecture to the new network architecture while ensuring that user services are not affected.

[0003] In related technologies, network cutover processes mainly rely on process-oriented support platforms for management. However, key aspects such as cutover plan formulation, script writing, command execution, and result verification still require significant manual intervention, resulting in overall efficiency that needs improvement. Furthermore, different operators may have varying understandings and execution standards for the same cutover scenario, easily introducing human error and hindering the reuse and standardization of experience. This not only affects the success rate and efficiency of cutovers but also increases network operation and maintenance costs and security risks. Summary of the Invention

[0004] This application provides a method, apparatus, device, and storage medium for processing network cutover tasks, which can improve the automation level and risk control capabilities of the cutover process, reduce the probability of cutover failure due to human error, and thus improve the overall efficiency and security of network operation and maintenance.

[0005] One aspect of this application provides a method for processing network cutover tasks, the method comprising: Receive network cutover request initiated by operator; wherein, the network cutover request is used to initiate a cutover task on the target network device; Based on the network cutover request, obtain the operation event logs and configuration files of the target network device; Based on the operation event log and the configuration file, a cutover script corresponding to the network cutover request is generated using a pre-built script generation model; High-risk instructions in the cutover script are identified by using a pre-built knowledge graph; Based on the high-risk instruction, determine the cutover risk assessment report level corresponding to the network cutover request, provide feedback on the cutover risk assessment report level to the operator, and execute the cutover script.

[0006] For example, in some embodiments, the script generation model is trained through the following steps: Collect sample operation event logs, sample configuration files, and actual network cutover operation commands from multiple sample network devices to obtain raw data; According to different cutover scenarios, the original data is labeled, and the labeled original data is cleaned and standardized to obtain a labeled dataset. The initialized script generation model is trained using the labeled dataset to obtain a trained script generation model.

[0007] For example, in some embodiments, generating the cutover script corresponding to the network cutover request based on the operation event log and the configuration file using a pre-built script generation model includes: Based on the network cutover request, determine the cutover scenario type corresponding to the current cutover task; Based on the cutover scenario type corresponding to the cutover task, the script generation model is fine-tuned using few-shot learning techniques to obtain the fine-tuned script generation model. Based on the operation event log and the configuration file, the cutover script corresponding to the network cutover request is generated through the fine-tuned script generation model. The knowledge graph is used to perform semantic verification and optimization of the cutover script.

[0008] Exemplarily, in some embodiments, the method further includes: Before executing the cutover script, historical indicator data of the target network device is collected; Based on the historical indicator data, the predicted indicator data of the target network device is determined by a pre-built trend prediction model; After executing the cutover script, collect the actual performance data of the target network device; Based on the consistency between the predicted indicator data and the actual indicator data, or based on the ratio between the actual indicator data and the historical indicator data, it is determined whether the target network device has any indicator anomalies.

[0009] Exemplarily, in some embodiments, the method further includes: During the execution of the cutover script, the actual operation commands of the target network device are monitored; The actual operation instruction is matched with a preset risk operation instruction identification rule to determine whether the actual operation instruction is a risk operation instruction. When the actual operation instruction is determined to be a risky operation instruction, query the operator's operation permissions for the risky operation instruction; If the operator has the authority to execute the risk operation instruction, the risk operation instruction shall be executed and the corresponding operation log shall be recorded. If the operator does not have the authority to execute the risk operation instruction, the risk operation instruction shall be intercepted.

[0010] Exemplarily, in some embodiments, the method further includes: Based on the device IP of the target network device, obtain all device node information of the target network device and its downstream devices, and analyze and filter out the device nodes affected by the cutover task. Based on the device IP of the target network device, obtain the service information accessed by the target network device and its downstream devices, and filter out a list of users affected by the cutover task based on the service information. Obtain the service link information and physical link information involved in the target network device, and filter out the list of circuits affected by the cutover task based on the service link information and the physical link information.

[0011] Exemplarily, in some embodiments, the method further includes: After executing the cutover script, network performance indicators, alarm information, and user online status data of the target network device and its downstream devices are collected as verification indicator data. The data to be verified is input into a pre-established network indicator cutover verification model. The network indicator cutover verification model automatically compares the data to be verified with the baseline data before the cutover to determine whether the data to be verified is abnormal, and determines whether the network and services of the target network device have returned to normal based on the judgment result. After confirming that the network and services of the target network device have returned to normal, network operation data of the target network device is continuously collected within a preset monitoring period. The network operation data is then subjected to multi-dimensional comprehensive analysis using a pre-built post-cutover evaluation model. The multi-dimensional comprehensive analysis includes at least one of the following: poor network quality analysis, poor service quality analysis, user online status analysis, or alarm correlation analysis.

[0012] On the other hand, embodiments of this application provide a processing apparatus for network cutover tasks, the apparatus comprising: A receiving unit is used to receive a network cutover request initiated by an operator; wherein the network cutover request is used to initiate a cutover task on a target network device; The acquisition unit is used to acquire the operation event logs and configuration files of the target network device based on the network cutover request. The generation unit is used to generate the cutover script corresponding to the network cutover request based on the operation event log and the configuration file through a pre-built script generation model. The identification unit is used to identify high-risk instructions in the cutover script by using a pre-built knowledge graph; The feedback unit is used to determine the cutover risk assessment report level corresponding to the network cutover request based on the high-risk instruction, to provide feedback of the cutover risk assessment report level to the operator, and to execute the cutover script.

[0013] On the other hand, embodiments of this application provide an electronic device, including a processor and a memory; The memory is used to store computer programs; The processor executes the computer program to implement the aforementioned method for processing network cutover tasks.

[0014] On the other hand, embodiments of this application provide a computer-readable storage medium storing a computer program, which is executed by a processor to implement the aforementioned network cutover task processing method.

[0015] On the other hand, embodiments of this application also provide a computer program product, which includes a computer program stored in a computer-readable storage medium. The processor of a computer device reads the computer program from the computer-readable storage medium and executes the computer program, causing the computer device to perform the processing method for implementing the aforementioned network cutover task.

[0016] The embodiments of this application include at least the following beneficial effects: This application provides a method, apparatus, device, and storage medium for processing network cutover tasks. This application receives network cutover requests initiated by operators, obtains the operation event logs and configuration files of the target network device based on the request, and then automatically generates and executes the corresponding cutover script using a pre-built script generation model, avoiding the problems of error-prone and difficult-to-standardize manual script writing. Furthermore, it identifies high-risk instructions in the cutover script through a pre-built knowledge graph and determines the cutover risk assessment report level based on the high-risk instructions, providing feedback to the operator. This achieves automatic identification and quantitative assessment of high-risk operations during the cutover process. Compared to existing methods that rely on manual judgment and post-event investigation, this application can proactively identify and grade potential risks before the cutover task is executed, effectively improving the automation level and risk control capabilities of the cutover process, reducing the probability of cutover failure due to human error, and thus improving the overall efficiency and security of network operation and maintenance. Attached Figure Description

[0017] The accompanying drawings are used to provide a further understanding of the technical solutions of this application and constitute a part of the specification. They are used together with the embodiments of this application to explain the technical solutions of this application and do not constitute a limitation on the technical solutions of this application.

[0018] Figure 1 This is a schematic representation of the verification details of a manual cutover provided in the embodiments of this application; Figure 2 This is a flowchart illustrating a method for processing a network cutover task provided in an embodiment of this application. Figure 3 This is a schematic diagram of an AI capability center of a system provided in an embodiment of this application; Figure 4 This is a schematic diagram illustrating an automatic cutover script generation method provided in an embodiment of this application; Figure 5 This is a schematic diagram of a high-risk instruction set identified in an embodiment of this application; Figure 6 This is a schematic diagram of a training script generation model provided in an embodiment of this application; Figure 7 This is a schematic diagram illustrating the principle of risk monitoring provided in the embodiments of this application; Figure 8 This is a schematic diagram illustrating the principle of cutover impact analysis provided in an embodiment of this application; Figure 9 This is a schematic diagram illustrating the principle of an indicator verification provided in an embodiment of this application; Figure 10 This is a schematic diagram illustrating the principle of quality monitoring provided in the embodiments of this application; Figure 11 This is a structural block diagram of a network cutover task processing device provided in an embodiment of this application; Figure 12 This is a structural block diagram of an electronic device provided in an embodiment of this application. Detailed Implementation

[0019] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0020] It is understood that the terms “first,” “second,” etc., used in this application may be used to describe various concepts herein, but unless otherwise stated, these concepts are not limited by these terms. These terms are used only to distinguish one concept from another.

[0021] As used in this application, the terms "at least one", "multiple", "each", "any", etc., "at least one" includes one, two or more, "multiple" includes two or more, "each" refers to each of the corresponding multiples, and "any" refers to any one of the multiples.

[0022] With the rapid evolution of communication networks and the continuous expansion of business scale, network operation and maintenance faces increasingly complex challenges. Network cutover, as a routine task in operator network operation and maintenance, involves multiple professional fields such as wireless, transmission, data, access, and power environment. The cutover processes and operation methods of different professional fields vary significantly, but all require the smooth migration of the system from the old network architecture to the new network architecture while ensuring that user services are not affected.

[0023] In related technologies, network cutover processes mainly rely on process-oriented support platforms for management. However, key aspects such as cutover plan formulation, script writing, command execution, and result verification still require significant manual intervention, resulting in overall efficiency that needs improvement. Furthermore, different operators may have varying understandings and execution standards for the same cutover scenario, easily introducing human error and hindering the reuse and standardization of experience. This not only affects the success rate and efficiency of cutovers but also increases network operation and maintenance costs and security risks.

[0024] For example, please refer to Figure 1 , Figure 1 This illustration shows a schematic representation of the verification details of a manual cutover provided in an embodiment of this application, such as... Figure 1 As shown, the current cutover task requires manual scripting for each step, followed by individual checks and verifications, resulting in low execution efficiency and high costs.

[0025] In view of this, this application provides a method, apparatus, device, and storage medium for processing network cutover tasks. This application receives network cutover requests initiated by operators, obtains the operation event logs and configuration files of the target network device based on the request, and then automatically generates and executes the corresponding cutover script using a pre-built script generation model. This avoids the problems of error-prone and difficult-to-standardize manually written scripts. Furthermore, a pre-built knowledge graph is used to identify high-risk instructions in the cutover script, and the cutover risk assessment report level is determined based on the high-risk instructions and fed back to the operator, achieving automatic identification and quantitative assessment of high-risk operations during the cutover process. Compared to existing methods that rely on manual judgment and post-event investigation, this application can proactively identify and grade potential risks before the cutover task is executed, effectively improving the automation level and risk control capabilities of the cutover process, reducing the probability of cutover failure due to human error, and thus improving the overall efficiency and security of network operation and maintenance.

[0026] General Description of Embodiments in this Application Please refer to Figure 2 , Figure 2 This illustration shows a flowchart of a network cutover task processing method provided in an embodiment of this application. Figure 2 As shown, a method for processing a network cutover task according to an embodiment of this application includes, but is not limited to, the following steps: Step 210: Receive a network cutover request initiated by the operator; wherein, the network cutover request is used to initiate a cutover task on the target network device; Step 220: Based on the network cutover request, obtain the operation event logs and configuration files of the target network device; Step 230: Based on the operation event log and configuration file, generate the cutover script corresponding to the network cutover request through the pre-built script generation model; Step 240: Identify high-risk instructions in the cutover script using a pre-built knowledge graph; Step 250: Determine the cutover risk assessment report level corresponding to the high-risk instruction, provide feedback on the cutover risk assessment report level to the operator, and execute the cutover script.

[0027] In this application embodiment, a method for processing network cutover tasks is provided. This method aims to improve the problems in related technologies where the network cutover process relies heavily on manually written scripts, has weak risk control of high-risk instructions, and lacks an automated evaluation mechanism, resulting in low cutover efficiency, frequent human errors, and prominent security risks. By constructing a risk pre-assessment framework driven by automatic script generation and knowledge graph, the method achieves intelligent processing and risk control of cutover tasks.

[0028] Specifically, this method first receives a network cutover request initiated by an operator, which is used to initiate a cutover task on the target network device. Then, based on this request, it obtains the target network device's operation event logs and configuration files. On this basis, a pre-built script generation model automatically generates and executes the corresponding cutover script based on the obtained operation event logs and configuration files, replacing the traditional manual scripting method. Furthermore, a pre-built knowledge graph is used to identify high-risk instructions in the cutover script, and the cutover risk assessment report level is determined based on the identified high-risk instructions, which is then fed back to the operator. This method combines automatic script generation with knowledge graph risk identification, achieving automated production of cutover scripts and pre-emptive quantitative assessment of high-risk instructions. This effectively reduces the error rate of manually written scripts, improves the security of cutover operations, and significantly enhances the overall efficiency and reliability of network cutover.

[0029] Below, in conjunction with Figure 2 The following describes and explains each step of the network cutover task processing method in the embodiments of this application.

[0030] In step 210, a network cutover request initiated by the operator is received; wherein, the network cutover request is used to initiate a cutover task on the target network device.

[0031] In this embodiment, receiving a network cutover request initiated by an operator is the starting point for the entire process. This step aims to transform the operator's cutover intention into a standardized request that the system can recognize and process, laying the foundation for subsequent automated processing. For example, in this embodiment, a unified request entry point can be used to incorporate the management of cutover tasks into a standardized process.

[0032] Specifically, operators can initiate network cutover requests through the interactive interface of the integrated dispatch system or other operation and maintenance management platforms. This request can include at least the information needed to uniquely identify the target network device to be operated, such as the device's IP address, device name, or device serial number. In addition, the network cutover request can also include basic attribute information of the cutover task, such as the cutover scenario type (version upgrade, link expansion, routing adjustment, etc.), the planned cutover time window, and the operator's identification.

[0033] After receiving the network cutover request, the system can verify the validity of the request. For example, it can verify the operator's permissions and check whether the target network device is operable. Once the verification is successful, the cutover request will be formally accepted.

[0034] In step 220, based on the network cutover request, the operation event logs and configuration files of the target network device are obtained.

[0035] In this embodiment of the application, after receiving the cutover request, the system needs to obtain historical operating data and current configuration information related to the target network device. This data is the basic material for subsequent intelligent generation of cutover scripts.

[0036] Specifically, in this embodiment, based on the device identification information carried in the cutover request, the target network device can be automatically accessed through the network management interface or network management system to retrieve the device's operation event logs over a past period, as well as the device's current configuration file. It is easy to understand that by automatically collecting raw data at the device level, the tedious process of operators manually reviewing and organizing information is avoided, and accurate and complete input is provided for the script generation model.

[0037] In this embodiment, the operation event log records the historical operation records of the target network device, including previously executed configuration change commands, system alarm events, device restart records, etc., which can reflect the device's operating status and historical change trajectory. The configuration file contains the complete parameter settings of the target network device currently in operation, such as interface configuration, routing protocol configuration, VLAN division, ACL rules, etc. This data will be organized into a standardized data format for subsequent script generation model processing and analysis.

[0038] In step 230, based on the operation event log and configuration file, the cutover script corresponding to the network cutover request is generated through a pre-built script generation model.

[0039] In this embodiment of the application, the use of a pre-trained script generation model to automatically generate a cutover script for the current cutover scenario based on the acquired device data is one of the important improvements of this application. This step replaces the traditional method of manually writing scripts. By using a deep learning model to understand the device context and cutover requirements, it outputs directly executable script instructions, which greatly improves the efficiency and accuracy of script generation.

[0040] Specifically, in this embodiment, the script generation model can be pre-trained based on deep learning and natural language processing techniques. During the training phase, this model learns the mapping relationship between operation commands and device configurations from a large number of historical cutover cases. In the inference phase, the script generation model receives the operation event logs and configuration files obtained in step 220 as input, and, combined with the cutover scenario type specified in the cutover request, automatically infers and generates a series of ordered operation instructions to form a complete cutover script. After format verification, the generated script can be automatically pushed to the target network device for execution.

[0041] For example, please refer to Figure 3 and Figure 4 , Figure 3 This illustration shows a schematic diagram of an AI capability center of a system provided in an embodiment of this application. Figure 4 This illustration shows a schematic diagram of an automatically generated cutover script provided in an embodiment of this application. The AI ​​capability center in this embodiment can integrate the fragmented data collection and strategy definition stages to construct a closed-loop system from data input to intelligent output. The center first gathers multi-source data such as device information, configuration files, and operation commands from the cutover data collection stage, and then uses this data to perform scenario matching and drive the intelligent script generation model.

[0042] At the training level of the script generation model, the system leverages historical network cutover data and combines advanced technologies such as BERT, knowledge graphs, and NLP for deep feature extraction and intelligent or manual annotation, thereby creating a script generation model with scene understanding capabilities. Ultimately, the script generation model can automatically generate suitable cutover scripts based on specific input cutover requirements and pass the output to the backend strategy definition stage, realizing the transformation of cutover scripts from experience-driven to data-intelligent driven.

[0043] In step 240, high-risk instructions in the cutover script are identified using a pre-built knowledge graph.

[0044] In this embodiment, during the generation and execution of the cutover script, a pre-built knowledge graph can be used to identify the risks of each instruction in the script. In this embodiment, the knowledge graph stores the relationships between high-risk instructions and risk levels in the script, as well as the chain reactions that various instructions may cause, providing structured knowledge support for the automatic identification of high-risk instructions.

[0045] Specifically, a knowledge graph can organize a large amount of professional knowledge in the field of network device operation in the form of a graph structure, including entities such as device type, operation instructions, instruction parameters, risk level, and historical failure cases, as well as their interrelationships. The system can match each instruction in the cutover script with the instruction nodes in the knowledge graph, and determine whether the instruction is a high-risk instruction based on the matching results. For example, instructions involving device restart, core routing protocol reset, and batch interface shutdown are usually marked as high-risk instructions. The knowledge graph can also dynamically assess the actual risk level of an instruction based on its specific parameters and the context information of the current device, avoiding a simplistic, one-size-fits-all classification. The identified high-risk instructions and their risk levels will be recorded as a basis for subsequent risk assessments.

[0046] For example, please refer to Figure 5 , Figure 5 The illustration shows a schematic diagram of a set of high-risk instructions identified in an embodiment of this application. These high-risk instructions can be systematically layered according to different operational impact granularities, such as global level, board level, port level, and protocol level. This layering approach aims to accurately identify and quantify the business risks that various instructions may cause, starting from different levels of the network architecture.

[0047] Specifically, global-level commands typically involve restarting the core control plane of a device or clearing global configurations, affecting the entire device and potentially causing an instantaneous interruption of all services. Board-level commands primarily operate on individual hardware boards; for example, shutting down or resetting a board directly impacts user sessions or service traffic carried on that board. Port-level commands act on physical or logical interfaces, involving configuration clearing or state switching, easily causing local link failures or service impairments. Protocol-level commands are related to routing and forwarding policies; for example, modifying routing policies or network protocol parameters can lead to routing oscillations, traffic black holes, or abnormal network topology. Through this multi-dimensional hierarchical identification, the system can provide granular evidence for subsequent automated risk assessments, ensuring more accurate and comprehensive control of high-risk commands.

[0048] In step 250, the cutover risk assessment report level corresponding to the network cutover request is determined according to the high-risk instruction, the cutover risk assessment report level is fed back to the operator, and the cutover script is executed.

[0049] In this embodiment, after identifying high-risk instructions, the risk level of the entire cutover task can be comprehensively assessed based on the identification results, and the assessment conclusions can be provided to the operators in a visual manner. This step allows operators to gain an intuitive understanding of the risk level of the operation, assisting them in making decisions such as whether to continue execution or whether additional protective measures are needed.

[0050] Specifically, based on the number of high-risk instructions identified in step 240, the distribution of risk levels, and the correlation between instructions, a preset risk assessment algorithm can be used to calculate the comprehensive risk score of this cutover task and map it to a corresponding risk assessment report level, such as low risk, medium risk, and high risk. The determination of the cutover risk assessment report level can consider not only the high-risk instructions themselves but also factors such as historical failure rates of similar scenarios recorded in the knowledge graph and equipment health status, used for weighted calculation. The generated cutover risk assessment report level will be displayed visually on the operator's interface, for example, through color coding (green for low risk, yellow for medium risk, and red for high risk) or level labels. Based on the feedback risk level, operators can choose to continue, postpone, or cancel the cutover task, or use auditing and monitoring tools to track and record activities in the system to promptly detect and respond to potential safety incidents, thereby achieving proactive management and pre-emptive control of cutover risks.

[0051] Furthermore, during the execution of the cutover script, the system can monitor the script's execution progress and return results in real time, ensuring that each instruction takes effect as expected. If an exception occurs during execution, error information can be recorded and the corresponding exception handling process can be triggered.

[0052] It is understood that the network cutover task processing method provided in this application embodiment receives a network cutover request initiated by an operator, obtains the operation event logs and configuration files of the target network device based on the request, and then automatically generates and executes the corresponding cutover script using a pre-built script generation model. This avoids the problems of error-prone and difficult-to-standardize manual script writing. Furthermore, a pre-built knowledge graph is used to identify high-risk instructions in the cutover script, and the cutover risk assessment report level is determined based on the high-risk instructions and fed back to the operator, achieving automatic identification and quantitative assessment of high-risk operations during the cutover process. Compared to existing methods that rely on manual judgment and post-event investigation, this method can identify and grade potential risks before the cutover task is executed, effectively improving the automation level and risk management capabilities of the cutover process, reducing the probability of cutover failure due to human error, and thus improving the overall efficiency and security of network operation and maintenance.

[0053] Specifically, in some embodiments, the script generation model is trained through the following steps: Collect sample operation event logs, sample configuration files, and actual network cutover operation commands from multiple sample network devices to obtain raw data; Based on different cutover scenarios, the original data is labeled, and the labeled original data is cleaned and standardized to obtain a labeled dataset. The initial script generation model is trained using a labeled dataset to obtain a trained script generation model.

[0054] This application provides a training method for a script generation model, which can construct a model for intelligently generating cutover scripts for network devices by using deep learning, natural language processing (NLP) technology and AI algorithms, combined with knowledge graphs.

[0055] Specifically, please refer to Figure 6 , Figure 6 This illustration shows a flowchart of a training script generation model provided in an embodiment of this application. In this embodiment, when training the script generation model, training data is first collected. Specifically, sample operation event logs, sample configuration files, and actual network cutover operation commands from multiple sample network devices are collected to obtain raw data. This raw data requires rigorous annotation. For example, it is tagged according to different cutover scenarios (e.g., cutover of equipment from the same manufacturer, capacity expansion cutover, routing changes, etc.), and the raw data undergoes preprocessing such as cleaning and standardization. After completing the preprocessing of the raw data, a labeled dataset is obtained.

[0056] In this embodiment, the script generation model can be built using the pre-trained BERT model from the field of natural language processing. Combined with domain-specific word embeddings related to network devices, the feature representation capability can be further enhanced. Furthermore, knowledge graph technology can be introduced to improve the model's understanding of the relationship between network devices and configurations, thereby improving the overall performance of the model.

[0057] When training the script generation model, the labeled dataset can be divided into training, validation, and test sets in a 7:2:1 ratio. To improve the model's generalization ability, data augmentation techniques can be used to expand the training dataset. During training, the mainstream PyTorch deep learning framework is used, and hyperparameters are adjusted using methods such as grid search and random search. For algorithm optimization, commonly used algorithms such as Adam and SGD are selected to improve the model training effect. The model's performance is evaluated using the validation set, with metrics such as accuracy, recall, and F1 score used for evaluation. Based on the evaluation results, the model structure and parameters are continuously adjusted and iteratively optimized to obtain a well-trained script generation model.

[0058] Specifically, in some embodiments, based on operation event logs and configuration files, a pre-built script generation model is used to generate a cutover script corresponding to the network cutover request, including: Based on the network cutover request, determine the cutover scenario type corresponding to the current cutover task; Based on the type of cutover scenario corresponding to the cutover task, the script generation model is fine-tuned using few-shot learning techniques to obtain the fine-tuned script generation model. Based on the operation event logs and configuration files, the cutover scripts corresponding to the network cutover requests are generated using a fine-tuned script generation model. The cutover script is semantically validated and optimized using a knowledge graph.

[0059] This application provides a method for generating cutover scripts based on few-shot learning and knowledge graph optimization. This method can further improve the accuracy and security of generated scripts by fine-tuning the script generation model through scene adaptation and semantic verification, based on the script generation model that has been initially trained.

[0060] Specifically, in this embodiment, after receiving a network cutover request initiated by an operator, the type of cutover scenario corresponding to the current cutover task can first be determined based on the request, such as capacity expansion cutover, routing change, or version upgrade. It is easy to understand that different cutover scenarios have different requirements for the logical structure and instruction content of the script; therefore, it is necessary to adapt the general script generation model to specific scenarios.

[0061] After determining the type of cutover scenario, a few-shot learning technique can be used to fine-tune the pre-trained script generation model. Specifically, in this embodiment, a small amount of sample data related to the cutover scenario type can be collected, such as historically successful cutover operation commands and configuration change records in that scenario. This sample data can be used as training material for fine-tuning, allowing for lightweight adjustments and optimizations to the parameters of the script generation model. Through few-shot learning, even when the available training data is extremely limited in a new scenario, the model can quickly adapt to and master the unique instruction patterns and operational rules of that scenario, thereby generating scripts that meet the needs of the current cutover scenario. This effectively overcomes the problem of traditional models performing poorly in new scenarios due to data sparsity.

[0062] After fine-tuning the model, the system generates a preliminary cutover script based on the acquired operation event logs and configuration files of the target network device, using the fine-tuned script generation model. Furthermore, this embodiment introduces a knowledge graph to perform semantic verification and optimization of the generated cutover script. The knowledge graph stores structured knowledge such as topological relationships between network devices, configuration dependency rules, and historical fault cases. Each line of the generated cutover script is compared with the rules in the knowledge graph to verify the logical rationality of the instructions and the correctness of the operation sequence. For example, it checks whether the configuration was correctly backed up before executing the board reset and whether redundant paths were confirmed before modifying the routing policy. For instructions that do not conform to the knowledge graph rules, the system automatically corrects or replaces them, ensuring that the final output cutover script is logically rigorous and operationally safe, thereby minimizing the cutover risk caused by script defects.

[0063] Specifically, in some embodiments, the method further includes: Before executing the cutover script, collect historical performance data of the target network device; Based on historical indicator data, predictive indicator data for the target network devices are determined through a pre-built trend prediction model. After executing the cutover script, collect the actual performance data of the target network device; Based on the consistency between predicted and actual indicator data, or based on the ratio of actual to historical indicator data, determine whether the target network device has any indicator anomalies.

[0064] In this embodiment of the application, a method for detecting anomalies in indicators before and after cutover based on a trend prediction model is provided. This method can use historical data to predict the expected trend of equipment indicators before cutover is performed, and compare the actual indicators with the predicted indicators after cutover is performed, thereby automatically and accurately determining whether the cutover operation has caused abnormal impact on network equipment.

[0065] Specifically, in this embodiment, before executing the cutover script, the system first collects historical performance data of the target network device. This historical performance data may include key performance parameters such as device traffic, optical power, CPU utilization, and memory usage. The time range for collection can be data from the week prior to the cutover or a longer period, to ensure that the normal fluctuation patterns and periodic characteristics of device operation can be captured.

[0066] After acquiring sufficient historical indicator data, the system inputs it into a pre-built trend prediction model. This trend prediction model can employ time series forecasting algorithms or machine learning regression models to predict the target network equipment's indicator data over a future period (e.g., during and shortly after a cutover). This predicted indicator data represents the trajectory of the equipment's indicators should exhibit under normal conditions without cutover interference.

[0067] After executing the cutover script, the system collects actual performance data from the target network device again, covering the same time window as the predicted performance data. The system then compares the actual and predicted performance data for consistency, such as calculating the percentage deviation or the sum of squared residuals. If there is a significant deviation between the actual and predicted performance data, such as a sudden drop in traffic exceeding a preset threshold or abnormal fluctuations in optical power, it can be determined that the target network device has abnormal performance.

[0068] Furthermore, the system can perform year-on-year and month-on-month comparisons between actual indicator data and historical indicator data before the cutover. For example, it can compare the flow rate at the same time after the cutover with the flow rate at the same time one week before the cutover, or compare the instantaneous change before and after the cutover. When the year-on-year or month-on-month change exceeds a dynamic threshold, the indicator can also be identified as abnormal. Through this multi-angle comparison mechanism, the system can effectively distinguish between normal fluctuations caused by the cutover operation and genuine abnormal states, avoiding false alarms or missed alarms, and providing reliable data support for the automated verification of cutover results.

[0069] Specifically, in some embodiments, the method further includes: During the execution of the cutover script, monitor the actual operation commands of the target network device; The actual operation instructions are matched with the preset risk operation instruction identification rules to determine whether the actual operation instructions are risk operation instructions. When an actual operation instruction is determined to be a risky operation instruction, query the operator's operation permissions for the risky operation instruction; If the operator has the authority to execute risk operation instructions, execute the risk operation instructions and record the corresponding operation log. If the operator does not have the authority to execute risky operation instructions, the risky operation instructions should be intercepted.

[0070] This application provides a real-time protection method based on risk operation instruction identification and permission control. It can dynamically monitor and verify the permissions of the actual instructions issued during the execution of the cutover script, thereby effectively preventing security incidents caused by misoperation or unauthorized operation.

[0071] Specifically, please refer to Figure 7 , Figure 7 This illustration shows a schematic diagram of a risk monitoring principle provided in an embodiment of this application. In this embodiment, during the execution of the cutover script, the system monitors in real time the actual operation instructions received by the target network device. These actual operation instructions include both instructions automatically issued by the script generation model and instructions manually entered by the operator during execution. The system matches and analyzes each actual operation instruction using preset risk operation instruction identification rules. Risk operation instruction identification rules can be defined based on dimensions such as the keyword of the instruction, the command category to which the instruction belongs, and the parameter range of the instruction. For example, instructions involving sensitive operations such as device restart, configuration erasure, and routing protocol reset can be marked as risk operation instructions. Furthermore, the system can also combine screen risk operation identification rules to verify the legitimacy of the operator's login to the network management system, such as determining whether the login account is correct and whether the login source IP is within the whitelist, thereby reducing risk at the entry point.

[0072] When an actual operation instruction is determined to be a risky operation instruction, the system immediately triggers a permission query process. This process checks whether the current operator has the necessary permissions to execute the risky instruction. Permission information can be pre-stored in the system's permission management module and configured in a refined manner based on the operator's role, rank, department, and other attributes.

[0073] If the query results show that the operator has the authority to execute the risky operation instruction, the system allows the instruction to continue execution and automatically records the corresponding operation log. The operation log can contain detailed information such as operator identification, operation instruction content, operation time, and risk level, for subsequent audit tracing and incident analysis. Conversely, if the query results show that the operator does not have the authority to execute the risky operation instruction, the system immediately intercepts the instruction, preventing it from being sent to the target network device, and provides the operator with feedback on the reason for the interception and warning information.

[0074] It is understood that, in this embodiment of the application, through this mechanism that combines real-time monitoring and permission verification, the system can promptly block risky operations before they occur, avoiding irreversible damage to the network caused by permission abuse or misoperation, and significantly improving the security and controllability of the cutover process.

[0075] Specifically, in some embodiments, the method further includes: Based on the target network device's IP address, obtain all device node information of the target network device and its downstream devices, and analyze and filter out the device nodes affected by the cutover task. Based on the target network device's IP address, obtain the service information accessed by the target network device and its downstream devices, and filter out a list of users affected by the cutover task based on the service information. Obtain the service link information and physical link information involved in the target network device, and filter out the list of circuits affected by the cutover task based on the service link information and physical link information.

[0076] This application provides an automatic analysis method for the scope of cutover impact based on device IP. It can comprehensively screen potentially affected device nodes, users and circuits before executing the cutover script, thereby providing operators with a global view of cutover risks and assisting them in formulating more comprehensive emergency plans.

[0077] Specifically, please refer to Figure 8 , Figure 8 This illustration shows a schematic diagram of a cutover impact analysis principle provided in an embodiment of this application. In this embodiment, firstly, based on the target network device's IP address, all device node information of the device and its downstream devices is obtained from the network management system or resource management database. Here, downstream devices refer to devices that have a physical or logical connection with the target network device, such as switches, routers, base stations, etc., directly connected via fiber optic or network cables. The system analyzes each device node by establishing rules to determine whether it is a downstream sub-device node of the target network device, or whether it belongs to the resource range that needs to be blocked during the cutover process, thereby filtering out a list of device nodes affected by the cutover task. The identification results of these affected device nodes can help operators understand the network scope that the cutover operation may affect.

[0078] Based on the completed device node analysis, the system further obtains all service information connected to the target network device and its downstream devices according to the device IP address. Service information may include customer name, service type, and service level agreement (SLA). The system establishes rules to analyze whether each service is carried by the target network device and whether the customers associated with the service have a direct or indirect dependency relationship with the target network device, thereby filtering out a list of users affected by the cutover task. This user list can provide a basis for subsequent notification work, such as sending service interruption warnings to affected users before the cutover.

[0079] In addition, the system also acquires all service link information and physical link information related to the target network device. Service link information describes the logical transmission path of data flow in the network, while physical link information reflects the actual physical connection relationship between devices. By establishing rules to analyze whether each link is a physical link or service link associated with the target network device, such as determining whether a fiber optic link is directly connected to a port of the target network device, or whether a service link is forwarded through the target network device, the system can filter out the list of circuits affected by the cutover task.

[0080] It is understood that, in this embodiment of the application, through comprehensive analysis of three dimensions—device nodes, users, and circuits—the system can provide operators with a complete profile of the cutover impact, enabling them to prepare in advance for tasks such as service backup, traffic switching, or user notification, effectively reducing the impact of the cutover on existing network services.

[0081] Specifically, in some embodiments, the method further includes: After executing the cutover script, network performance indicators, alarm information, and user online status data of the target network device and its downstream devices are collected as verification indicator data. Input the data of the indicator to be verified into the pre-established network indicator cutover verification model. The network indicator cutover verification model will automatically compare the data of the indicator to be verified with the corresponding baseline data before the cutover, determine whether there is any abnormality in the data of the indicator to be verified, and determine whether the network and services of the target network device have returned to normal based on the judgment result. After confirming that the network and services of the target network device have returned to normal, network operation data of the target network device is continuously collected within a preset monitoring period. The network operation data is then analyzed in multiple dimensions using a pre-built post-cutover evaluation model. The multi-dimensional comprehensive analysis includes at least one of the following: poor network quality analysis, poor service quality analysis, user online status analysis, or alarm correlation analysis.

[0082] In this application embodiment, a post-cutover evaluation method based on automated verification and continuous monitoring is also provided. After the cutover script is executed, it can automatically determine whether the network and services have returned to normal through intelligent indicator comparison and multi-dimensional analysis. After confirming the recovery, it can carry out continuous quality monitoring for several days to ensure the long-term stability of the cutover effect.

[0083] Specifically, please refer to Figure 9 and Figure 10 , Figure 9 The diagram illustrates the principle of an indicator verification method provided in an embodiment of this application. Figure 10This illustration shows a schematic diagram of a quality monitoring principle provided in an embodiment of this application. In this embodiment, after the cutover script is executed, network performance indicators, alarm information, and user online status data of the target network device and its connected devices are automatically collected, and these data are used as indicator data to be verified. Network performance indicators may include key parameters such as device traffic, optical power, CPU utilization, and memory usage; alarm information may include alarm events of various levels reported by the device; and user online status data reflects the access and activity of users within the cutover range.

[0084] After data collection, the data of the metrics to be verified is input into a pre-established network metric cutover verification model. This model contains baseline data collected before the cutover, such as the values ​​of various metrics for the same period one hour or day before the cutover. The model automatically compares the data of the metrics to be verified with the baseline data item by item, such as calculating the percentage deviation of traffic, the increase or decrease in the number of alarms, and the change in user online rate. If the comparison results show that all metrics are within the preset normal fluctuation range, the network and services are considered to have returned to normal; if a certain metric deviates significantly, an anomaly is identified and the problem point is marked for further investigation by operators.

[0085] After confirming that the network and services of the target network device have returned to normal, the system does not immediately end monitoring but enters a continuous quality monitoring phase. Within a preset monitoring period, such as seven consecutive days after the cutover, the system continuously collects network operation data from the target network device. This data is input into a pre-built post-cutover evaluation model, which performs multi-dimensional comprehensive analysis.

[0086] In this embodiment of the application, the multi-dimensional comprehensive analysis may include: poor network quality analysis, used to identify whether there are areas or devices with degraded performance in the network; poor service quality analysis, used to detect whether the quality of a specific service has declined; user online status analysis, used to observe abnormal changes in user access and online behavior; and alarm correlation analysis, used to explore the inherent relationship between different alarm events and determine whether there are potential hidden dangers.

[0087] Understandably, through this continuous, multi-dimensional monitoring and analysis, the system can comprehensively assess the impact of the cutover on the network within a relatively long time window after the cutover, promptly identify delayed issues, and ensure the long-term stability of network quality and service quality after the cutover.

[0088] It is understood that the technical solution of this application has at least the following advantages: 1. Improved the pain points of manual cutover script writing being prone to errors and difficult to reuse: By building a script generation model based on deep learning and natural language processing technology, the cutover script adapted to the current cutover scenario is automatically generated according to the device operation log and configuration file, replacing the traditional manual writing method. This effectively avoids logical and grammatical errors caused by human negligence, fatigue or misunderstanding of requirements, while realizing the standardized reuse of cutover experience.

[0089] 2. Significantly improved cutover risk management capabilities: By using a pre-built knowledge graph to automatically identify high-risk instructions in the cutover script, and combining this with operator permissions for real-time verification and interception, high-risk instructions can be detected in advance and blocked during the process. This makes up for the shortcomings of inaccurate risk assessment and lack of emergency response in traditional methods, and greatly reduces the risk of business interruption caused by the misexecution of high-risk instructions.

[0090] 3. Enhanced efficiency and accuracy of cutover verification: By introducing a network indicator cutover verification model, the indicator data after the cutover is automatically collected and compared with the baseline data before the cutover, replacing the traditional verification method that relies on manual verification one by one. This eliminates the problem of inconsistent judgment standards among different verification personnel. At the same time, through continuous quality monitoring and multi-dimensional comprehensive analysis over several days, hidden faults and delay issues can be detected in a timely manner.

[0091] 4. Significantly improved overall cutover efficiency: Through a fully automated closed loop of script generation, risk assessment, and verification, the need for manual intervention is significantly reduced, and the success rate and execution speed of a single cutover are significantly improved.

[0092] 5. Wide range of applications: The solution proposed in this application is applicable to various cutover scenarios such as version upgrades, link expansion, route adjustment, board expansion, firewall connectivity, port mirroring, batch configuration distribution of devices, PON port cutover, OLT cutover between old and new metropolitan areas, and new device network access. It has good versatility and large-scale promotion value.

[0093] Reference Figure 11 In this embodiment of the application, a network cutover task processing apparatus is also provided, which includes: The receiving unit 1110 is used to receive a network cutover request initiated by an operator; wherein, the network cutover request is used to initiate a cutover task on the target network device. The acquisition unit 1120 is used to acquire the operation event logs and configuration files of the target network device based on the network cutover request; The generation unit 1130 is used to generate the cutover script corresponding to the network cutover request based on the operation event log and configuration file through a pre-built script generation model. The identification unit 1140 is used to identify high-risk instructions in the cutover script through a pre-built knowledge graph; Feedback unit 1150 is used to determine the cutover risk assessment report level corresponding to the network cutover request based on the high-risk instruction, provide feedback on the cutover risk assessment report level to the operator, and execute the cutover script.

[0094] It is understandable that, such as Figure 2 The content of the network cutover task processing method embodiments shown herein is applicable to the network cutover task processing device embodiments. The specific functions implemented by the network cutover task processing device embodiments are as follows: Figure 2 The method for processing network cutover tasks shown in the embodiment is the same, and the beneficial effects achieved are the same as those described above. Figure 2 The beneficial effects achieved by the network cutover task processing method embodiment shown are also the same.

[0095] Reference Figure 12 This application also discloses an electronic device, including: At least one processor 1210; At least one memory 1220 is used to store at least one program; When at least one program is executed by at least one processor 1210, such that at least one processor 1210 performs as follows: Figure 2 The example shown illustrates a method for handling network cutover tasks.

[0096] The electronic device in this application embodiment may be a mobile phone, a computer device, or a server device.

[0097] Understandable Figure 2 The methods for processing network cutover tasks shown in the embodiments are all applicable to the embodiments of this electronic device. The specific functions implemented in the embodiments of this electronic device are the same as those in the embodiments of this electronic device. Figure 2 The method for processing the network cutover task shown in the embodiment is the same, and the beneficial effects achieved are the same. Figure 2 The beneficial effects achieved by the network cutover task processing method embodiment shown are also the same.

[0098] This application also discloses a computer-readable storage medium storing a processor-executable program, which, when executed by a processor, is used to implement, for example... Figure 2 The example shown illustrates a method for handling network cutover tasks.

[0099] Understandable Figure 2 The methods for processing network cutover tasks shown in the embodiments are applicable to the embodiments of this computer-readable storage medium. The specific functions implemented in the embodiments of this computer-readable storage medium are the same as those in the embodiments of this computer-readable storage medium. Figure 2The method for processing the network cutover task shown in the embodiment is the same, and the beneficial effects achieved are the same. Figure 2 The beneficial effects achieved by the network cutover task processing method embodiment shown are also the same.

[0100] This application also discloses a computer program product or computer program, which includes computer instructions stored in the aforementioned computer-readable storage medium. Figure 12 The processor of the illustrated electronic device can read the computer instructions from the aforementioned computer-readable storage medium, and the processor executes the computer instructions, causing the computer device to perform... Figure 2 The example shown illustrates a method for handling network cutover tasks.

[0101] Understandable Figure 2 The methods for handling network cutover tasks shown in the embodiments are all applicable to this computer program product or computer program embodiment. The specific functions implemented by this computer program product or computer program embodiment are the same as those described above. Figure 2 The method for processing the network cutover task shown in the embodiment is the same, and the beneficial effects achieved are the same. Figure 2 The beneficial effects achieved by the network cutover task processing method embodiment shown are also the same.

[0102] In some alternative embodiments, the functions / operations mentioned in the block diagrams may not occur in the order shown in the operation diagrams. For example, depending on the functions / operations involved, two consecutively shown blocks may actually be executed substantially simultaneously, or the blocks may sometimes be executed in reverse order. Furthermore, the embodiments presented and described in the flowcharts of this application are provided by way of example to provide a more comprehensive understanding of the technology. The disclosed methods are not limited to the operations and logic flows presented herein. Alternative embodiments are contemplated in which the order of various operations is changed and sub-operations described as part of a larger operation are executed independently.

[0103] Furthermore, although this application is described in the context of functional modules, it should be understood that, unless otherwise stated to the contrary, one or more of the functions and / or features may be integrated into a single physical device and / or software module, or one or more functions and / or features may be implemented in a separate physical device or software module. It is also understood that a detailed discussion of the actual implementation of each module is unnecessary for understanding this application. Rather, given the properties, functions, and internal relationships of the various functional modules in the apparatus disclosed herein, the actual implementation of the module will be understood within the scope of conventional technology for an engineer. Therefore, those skilled in the art can implement the application set forth in the claims using ordinary techniques without excessive experimentation. It is also understood that the specific concepts disclosed are merely illustrative and not intended to limit the scope of this application, which is determined by the full scope of the appended claims and their equivalents.

[0104] If a function is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0105] In the embodiments of this application, the terms "module" or "unit" refer to a computer program or part of a computer program that has a predetermined function and works with other related parts to achieve a predetermined goal, and can be implemented wholly or partially using software, hardware (such as processing circuitry or memory), or a combination thereof. Similarly, a processor (or multiple processors or memory) can be used to implement one or more modules or units. Furthermore, each module or unit can be part of an overall module or unit that includes the functionality of that module or unit.

[0106] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as a sequenced list of executable instructions for implementing logical functions, and can be embodied in any computer-readable storage medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (such as a computer-based system, a processor-included system, or other system that can fetch and execute instructions from, an instruction execution system, apparatus, or device). For the purposes of this specification, "computer-readable storage medium" can be any means that can contain, store, communicate, propagate, or transmit programs for use by, or in conjunction with, an instruction execution system, apparatus, or device.

[0107] It should be understood that various parts of this application can be implemented using hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented using software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, it can be implemented using any one or a combination of the following techniques known in the art: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.

[0108] In the foregoing description of this specification, the references to terms such as "one embodiment," "another embodiment," or "some embodiments," etc., indicate that a specific feature, structure, material, or characteristic described in connection with an embodiment or example is included in at least one embodiment or example of this application. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0109] Although embodiments of this application have been shown and described, those skilled in the art will understand that various changes, modifications, substitutions and variations can be made to these embodiments without departing from the principles and spirit of this application, the scope of which is defined by the claims and their equivalents.

[0110] The above is a detailed description of the preferred embodiments of this application, but this application is not limited to the embodiments. Those skilled in the art can make various equivalent modifications or substitutions without departing from the spirit of this application, and these equivalent modifications or substitutions are all included within the scope defined by the claims of this application.

Claims

1. A method for processing network cutover tasks, characterized in that, The method includes: Receive network cutover request initiated by operator; wherein, the network cutover request is used to initiate a cutover task on the target network device; Based on the network cutover request, obtain the operation event logs and configuration files of the target network device; Based on the operation event log and the configuration file, a cutover script corresponding to the network cutover request is generated using a pre-built script generation model; High-risk instructions in the cutover script are identified by using a pre-built knowledge graph; Based on the high-risk instruction, determine the cutover risk assessment report level corresponding to the network cutover request, provide feedback on the cutover risk assessment report level to the operator, and execute the cutover script.

2. The method for processing network cutover tasks according to claim 1, characterized in that, The script generation model is trained through the following steps: Collect sample operation event logs, sample configuration files, and actual network cutover operation commands from multiple sample network devices to obtain raw data; According to different cutover scenarios, the original data is labeled, and the labeled original data is cleaned and standardized to obtain a labeled dataset. The initialized script generation model is trained using the labeled dataset to obtain a trained script generation model.

3. The method for processing network cutover tasks according to claim 2, characterized in that, The step of generating the cutover script corresponding to the network cutover request based on the operation event log and the configuration file using a pre-built script generation model includes: Based on the network cutover request, determine the cutover scenario type corresponding to the current cutover task; Based on the cutover scenario type corresponding to the cutover task, the script generation model is fine-tuned using few-shot learning techniques to obtain the fine-tuned script generation model. Based on the operation event log and the configuration file, the cutover script corresponding to the network cutover request is generated through the fine-tuned script generation model. The knowledge graph is used to perform semantic verification and optimization on the cutover script.

4. The method for processing network cutover tasks according to claim 1, characterized in that, The method further includes: Before executing the cutover script, historical indicator data of the target network device is collected; Based on the historical indicator data, the predicted indicator data of the target network device is determined through a pre-built trend prediction model; After executing the cutover script, collect the actual performance data of the target network device; Based on the consistency between the predicted indicator data and the actual indicator data, or based on the ratio between the actual indicator data and the historical indicator data, it is determined whether the target network device has any indicator anomalies.

5. The method for processing network cutover tasks according to claim 1, characterized in that, The method further includes: During the execution of the cutover script, the actual operation commands of the target network device are monitored; The actual operation instruction is matched with a preset risk operation instruction identification rule to determine whether the actual operation instruction is a risk operation instruction. When the actual operation instruction is determined to be a risky operation instruction, query the operator's operation permissions for the risky operation instruction; If the operator has the authority to execute the risk operation instruction, the risk operation instruction shall be executed and the corresponding operation log shall be recorded. If the operator does not have the authority to execute the risk operation instruction, the risk operation instruction shall be intercepted.

6. The method for processing network cutover tasks according to claim 1, characterized in that, The method further includes: Based on the device IP of the target network device, obtain all device node information of the target network device and its downstream devices, and analyze and filter out the device nodes affected by the cutover task. Based on the device IP of the target network device, obtain the service information accessed by the target network device and its downstream devices, and filter out a list of users affected by the cutover task based on the service information. Obtain the service link information and physical link information involved in the target network device, and filter out the list of circuits affected by the cutover task based on the service link information and the physical link information.

7. The method for processing network cutover tasks according to any one of claims 1-6, characterized in that, The method further includes: After executing the cutover script, network performance indicators, alarm information, and user online status data of the target network device and its downstream devices are collected as verification indicator data. The data to be verified is input into a pre-established network indicator cutover verification model. The network indicator cutover verification model automatically compares the data to be verified with the baseline data before the cutover to determine whether the data to be verified is abnormal, and determines whether the network and services of the target network device have returned to normal based on the judgment result. After confirming that the network and services of the target network device have returned to normal, network operation data of the target network device is continuously collected within a preset monitoring period. The network operation data is then subjected to multi-dimensional comprehensive analysis using a pre-built post-cutover evaluation model. The multi-dimensional comprehensive analysis includes at least one of the following: poor network quality analysis, poor service quality analysis, user online status analysis, or alarm correlation analysis.

8. A processing apparatus for network cutover tasks, characterized in that, The device includes: A receiving unit is used to receive a network cutover request initiated by an operator; wherein the network cutover request is used to initiate a cutover task on a target network device; The acquisition unit is used to acquire the operation event logs and configuration files of the target network device based on the network cutover request. The generation unit is used to generate the cutover script corresponding to the network cutover request based on the operation event log and the configuration file through a pre-built script generation model. The identification unit is used to identify high-risk instructions in the cutover script by using a pre-built knowledge graph; The feedback unit is used to determine the cutover risk assessment report level corresponding to the network cutover request based on the high-risk instruction, to provide feedback of the cutover risk assessment report level to the operator, and to execute the cutover script.

9. An electronic device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the network cutover task processing method according to any one of claims 1 to 7.

10. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements the network cutover task processing method according to any one of claims 1 to 7.

Images