Fault processing method and device of cloud computing platform, electronic equipment and storage medium

Abstract

The invention provides a fault processing method and device for a cloud computing platform, electronic equipment and a storage medium, and relates to the technical field of computers.The method comprises the steps that a target fault type, needing to be detected, of the cloud computing platform is determined; obtaining operation data of the cloud computing platform for data preprocessing to obtain a standard data set; performing clustering analysis on the standard data set to obtain a normal state data training machine learning algorithm model, and obtaining an anomaly detection model; inputting the operation data into an anomaly detection model to obtain historical fault data so as to train an initial prediction model and obtain an initial fault detection model; optimizing the initial fault detection model, and updating the initial fault detection model according to the real-time fault data to obtain a fault detection model; and performing fault detection on the cloud computing platform according to the fault detection model to obtain a fault detection result. According to the method, the efficiency and accuracy of fault diagnosis of the cloud computing platform are effectively improved, and the method can adapt to a complex and dynamic cloud computing environment.

Description

Technical Field

[0001] The present invention relates to the field of computer technology, and in particular to a fault handling method, device, electronic equipment and storage medium for a cloud computing platform. Background Art

[0002] Cloud computing is an Internet-based computing model that uses virtualization technology to provide computing resources (such as servers, storage, and networks) to users in the form of services. Users can use resources on demand without having to purchase and maintain physical hardware. With the rapid development of cloud computing technology, more and more companies and organizations are choosing to migrate their businesses and data to the cloud to enjoy the elasticity, efficiency, and cost advantages brought by cloud computing. However, as the complexity of the cloud environment continues to increase, the failure problems faced by system operations are becoming increasingly prominent. System failures not only affect the stability and availability of services, but may also lead to serious consequences such as data loss and business interruption.

[0003] Fault detection and repair in traditional cloud computing environments mainly rely on manual operations. Operation and maintenance personnel need to analyze problems based on system logs and performance monitoring data, locate the cause of the fault and take corresponding repair measures. However, this method is inefficient. Especially when facing complex distributed systems, problem location often takes a lot of time and effort, which can easily lead to long-term system downtime and service interruption. In recent years, the application of artificial intelligence technology in the field of cloud computing has gradually received attention. AI technology can be combined with automation tools to quickly execute repair measures after detecting a fault and reduce system downtime. However, due to the huge amount of data and strong heterogeneity in the cloud computing environment, the dynamic nature of the cloud computing environment makes the system state change all the time. The existing processing methods based on artificial intelligence technology cannot quickly and accurately detect and repair faults. Summary of the invention

[0004] In view of the above-mentioned deficiencies in the prior art, the present invention provides a fault handling method, device, electronic device and storage medium for a cloud computing platform, which effectively solves the problem that the existing processing methods based on artificial intelligence technology cannot quickly and accurately detect and repair faults.

[0005] In a first aspect, the present invention provides a method for troubleshooting a cloud computing platform, the method comprising:

[0006] Determine the target fault types that need to be detected on the cloud computing platform;

[0007] Acquire operation data of the cloud computing platform, perform data preprocessing on the operation data, and obtain a standard data set;

[0008] Performing cluster analysis on the standard data set according to the target fault type to obtain normal state data, and using the normal state data to train a machine learning algorithm model to obtain an abnormality detection model;

[0009] Inputting the operating data into the anomaly detection model to obtain historical fault data, establishing a historical data set based on the historical fault data, and using the historical data set to train an initial prediction model to obtain an initial fault detection model;

[0010] Optimizing the initial fault detection model, updating the initial fault detection model according to the real-time fault data output by the initial fault detection model, and obtaining a fault detection model;

[0011] Fault detection is performed on the cloud computing platform according to the fault detection model to obtain a fault detection result.

[0012] Furthermore, the method further comprises:

[0013] Establishing a fault repair system according to the target fault type and setting fault automatic repair rules;

[0014] Using a reinforcement learning model to perform reinforcement learning on the fault repair system to obtain an automatic fault repair strategy;

[0015] The corresponding automatic fault repair strategy is selected according to the fault detection result to automatically repair the fault.

[0016] Furthermore, performing cluster analysis on the standard data set according to the target fault type to obtain normal state data, and using the normal state data to train a machine learning algorithm model to obtain an abnormality detection model includes:

[0017] Using a cluster analysis algorithm to classify the standard data set according to the target fault type to obtain the normal state data;

[0018] The machine learning algorithm model is trained by the normal state data to learn the normal behavior pattern to detect data that deviates from the normal behavior pattern, thereby obtaining the abnormality detection model.

[0019] Furthermore, the historical data set is established according to the historical fault data output by the anomaly detection model, and the initial prediction model is trained using the historical data set to obtain the initial fault detection model, including:

[0020] Performing data preprocessing on the historical fault data output by the anomaly detection model, and dividing the preprocessed historical fault data into a training set and a test set;

[0021] The initial prediction model is trained by using the features and labels of the training set to learn the characteristic patterns of normal and fault conditions;

[0022] Performing cross-validation evaluation on the trained initial prediction model through the test set to obtain an evaluation result;

[0023] The parameters of the initial prediction model are adjusted according to the evaluation result, and the training set is used for retraining to obtain an initial fault prediction model.

[0024] Furthermore, the optimizing the initial fault detection model, updating the initial fault detection model according to the real-time fault data output by the initial fault detection model, and obtaining the fault detection model includes:

[0025] Performing hyperparameter adjustment and feature selection on the initial fault detection model to obtain an optimized initial fault detection model;

[0026] Performing incremental learning and periodic retraining on the optimized initial fault detection model according to the real-time fault data;

[0027] An online learning algorithm is used to dynamically adjust the model parameters of the initial fault detection model to obtain a fault detection model.

[0028] Furthermore, the reinforcement learning model is used to perform reinforcement learning on the fault repair system to obtain an automatic fault repair strategy, including:

[0029] Setting a reinforcement learning environment for the fault repair system, wherein the reinforcement learning environment includes a state space, an action space, and a reward function;

[0030] Using the reinforcement learning model to perform exploration sampling on the reinforcement learning environment to obtain sampling data;

[0031] The reinforcement learning model is trained using the sampled data so that the reinforcement learning model finds the optimal reward strategy of the reinforcement learning environment and obtains the automatic fault repair strategy.

[0032] In a second aspect, the present invention further provides a fault handling device for a cloud computing platform, the device comprising:

[0033] A demand analysis module is used to determine the target fault types that need to be detected on the cloud computing platform;

[0034] A data processing module is used to obtain the operation data of the cloud computing platform, perform data preprocessing on the operation data, and obtain a standard data set;

[0035] A first training module is used to perform cluster analysis on the standard data set according to the target fault type to obtain normal state data, and use the normal state data to train a machine learning algorithm model to obtain an abnormality detection model;

[0036] A second training module is used to input the operating data into the anomaly detection model to obtain historical fault data, establish a historical data set based on the historical fault data, and use the historical data set to train an initial prediction model to obtain an initial fault detection model;

[0037] A model optimization module, used to optimize the initial fault detection model, update the initial fault detection model according to the real-time fault data output by the initial fault detection model, and obtain a fault detection model;

[0038] The fault detection module is used to perform fault detection on the cloud computing platform according to the fault detection model to obtain a fault detection result.

[0039] Furthermore, the device also includes:

[0040] A system building module, used to establish a fault repair system according to the target fault type and set automatic fault repair rules;

[0041] A reinforcement learning module, used to perform reinforcement learning on the fault repair system using a reinforcement learning model to obtain an automatic fault repair strategy;

[0042] The fault repair module is used to select the corresponding automatic fault repair strategy according to the fault detection result to automatically repair the fault.

[0043] In a third aspect, the present invention further provides an electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the fault handling method for the cloud computing platform as described in the first aspect of the present invention.

[0044] In a fourth aspect, the present invention further provides a computer-readable storage medium having a computer program stored thereon, wherein when the computer program is executed by a processor, the fault handling method for the cloud computing platform as described in the first aspect of the present invention is implemented.

[0045] The fault handling method, device, electronic device and storage medium of the cloud computing platform provided by the present invention can analyze a large amount of historical operation data through machine learning and deep learning technology to establish an intelligent fault diagnosis model. These models can not only accurately identify the fault type, but also locate the root cause of the fault, thereby reducing the situation of false positives and false negatives. Through a data-driven approach, the efficiency and accuracy of fault diagnosis are significantly improved, and it can adapt to complex and dynamic cloud computing environments. At the same time, according to the fault diagnosis results, the repair strategy can be automatically selected and implemented, which not only shortens the recovery time of the cloud computing platform, but also improves the availability of the service. By continuously learning and accumulating historical fault data for self-optimization, improving fault detection and repair strategies, not only the intelligence level of the cloud computing platform is enhanced, but also it has stronger adaptability when facing new faults. Through continuous iteration and updating, cloud computing can maintain a high service quality and stability, and significantly improve the reliability, efficiency and user experience of the cloud computing platform. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for use in the embodiments are briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present invention and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without creative work.

[0047] Figure 1 This is a first schematic diagram of a process of a method for troubleshooting a cloud computing platform provided by an embodiment of the present invention;

[0048] Figure 2 This is a second schematic diagram of the process of the fault handling method of the cloud computing platform provided by an embodiment of the present invention;

[0049] Figure 3 This is a third schematic diagram of the process of the fault handling method of the cloud computing platform provided by an embodiment of the present invention;

[0050] Figure 4 This is a fourth schematic diagram of a fault handling method for a cloud computing platform provided by an embodiment of the present invention;

[0051] Figure 5 This is a fifth schematic diagram of a method for troubleshooting a cloud computing platform provided by an embodiment of the present invention;

[0052] Figure 6 This is a sixth schematic diagram of a fault handling method for a cloud computing platform provided by an embodiment of the present invention;

[0053] Figure 7 It is a first schematic diagram of the structure of a fault handling device for a cloud computing platform provided by an embodiment of the present invention;

[0054] Figure 8 is a second schematic diagram of the structure of a fault handling device for a cloud computing platform provided by an embodiment of the present invention;

[0055] Fig. 9 It is a structural schematic diagram of an electronic device provided by an embodiment of the present invention.

[0056] Description of main component symbols:

[0057] 700. Fault handling device of cloud computing platform; 710. Demand analysis module; 720. Data processing module; 730. First training module; 740. Second training module; 750. Model optimization module; 760. Fault detection module; 770. System construction module; 780. Reinforcement learning module; 790. Fault repair module; 900. Electronic device; 910. Processor; 920. Communication interface; 930. Memory; 940. Communication bus. DETAILED DESCRIPTION

[0058] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be further clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. It should be noted that the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0059] In addition, the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features defined as "first" and "second" may explicitly or implicitly include one or more of the features. In the description of the present invention, the meaning of "plurality" is two or more, unless otherwise clearly and specifically defined.

[0060] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by those skilled in the art to which this application belongs. The terms used in this specification are only for the purpose of describing specific embodiments and are not intended to limit the present invention. The term "and / or" used herein includes any and all combinations of one or more of the related listed items.

[0061] Fault detection and repair in traditional cloud computing environments mainly rely on manual operations. Operation and maintenance personnel need to analyze problems based on system logs and performance monitoring data, locate the cause of the fault and take corresponding repair measures. However, this method is inefficient. Especially when facing complex distributed systems, problem location often takes a lot of time and effort, which can easily lead to long-term system downtime and service interruption. In recent years, the application of artificial intelligence technology in the field of cloud computing has gradually received attention. AI technology can be combined with automation tools to quickly execute repair measures after detecting a fault and reduce system downtime. However, due to the huge amount of data and strong heterogeneity in the cloud computing environment, the dynamic nature of the cloud computing environment makes the system state change all the time. The existing processing methods based on artificial intelligence technology cannot quickly and accurately detect and repair faults.

[0062] Example 1

[0063] The embodiment of the present invention provides a fault handling method for a cloud computing platform, which effectively solves the problem that the existing processing methods based on artificial intelligence technology cannot quickly and accurately detect and repair faults. Figure 1 FIG. 1 is a first schematic diagram of a method for troubleshooting a cloud computing platform provided by an embodiment of the present invention. Figure 1 As shown, the method comprises the following steps:

[0064] S100: Determine a target fault type that needs to be detected on the cloud computing platform.

[0065] In order to determine the specific fault detection and repair requirements, it is first necessary to determine the target fault type of the cloud computing platform, which includes but is not limited to hardware failures, software anomalies, and network problems.

[0066] S200: Acquire operation data of the cloud computing platform, perform data preprocessing on the operation data, and obtain a standard data set.

[0067] In the embodiment of the present invention, log analysis and system monitoring tools can be used to collect and monitor the resource operation status of the cloud computing platform. For example, the open source system monitoring and alarm toolkit Prometheus and the open source data visualization and monitoring platform Grafana can be used to perform comprehensive log analysis and system monitoring to obtain operation data. The operation data includes but is not limited to system performance monitoring data such as CPU usage, memory usage, network throughput, and disk I / O, and can also include log data such as system logs and application logs.

[0068] Then, the operation data is preprocessed by data cleaning, feature extraction, and standardization. Data cleaning can remove duplicate data and invalid data (such as null values ​​and outliers). Feature extraction extracts fault-related features from the operation data, such as the maximum value, average value, and standard deviation of CPU usage. Standardization standardizes or normalizes the extracted features so that different features have the same scale and obtains a standard data set to facilitate subsequent model training.

[0069] S300. Perform cluster analysis on the standard data set according to the target fault type to obtain normal state data, and use the normal state data to train a machine learning algorithm model to obtain an anomaly detection model.

[0070] Figure 2 FIG. 2 is a second schematic diagram of a fault handling method for a cloud computing platform provided by an embodiment of the present invention. Figure 2 As shown in Figure 2, the anomaly detection model is obtained using the following steps:

[0071] S310 , using a cluster analysis algorithm to classify the standard data set according to the target fault type to obtain normal state data.

[0072] In the embodiment of the present invention, the cluster analysis algorithm can adopt a k-means clustering algorithm or a density-based clustering algorithm. The k-means clustering algorithm divides the standard data set into k clusters, each cluster represents a data state, analyzes the characteristics of each cluster to identify the clustering of normal state and abnormal state, thereby obtaining normal state data and abnormal state data. The density-based clustering algorithm performs clustering by identifying the core points, boundary points and noise points of the standard data set, and removes the noise points as potential abnormal points to obtain normal state data.

[0073] S320. Train the machine learning algorithm model through normal state data to learn the normal behavior pattern, so as to detect data that deviates from the normal behavior pattern and obtain an anomaly detection model.

[0074] In the embodiment of the present invention, according to business needs and data characteristics, an isolation forest model or an autoencoder model can be selected to use normal state data to train and learn normal behavior patterns, and unsupervised learning can be used to identify potential abnormal points and detect data that deviates from normal behavior patterns, thereby obtaining an anomaly detection model. The dual detection of the clustering algorithm and the machine learning algorithm model can improve the accuracy of anomaly detection and effectively reduce false positives and false negatives.

[0075] S400, inputting the operation data into the anomaly detection model to obtain historical fault data, establishing a historical data set according to the historical fault data, and using the historical data set to train an initial prediction model to obtain an initial fault detection model.

[0076] Figure 3FIG. 3 is a third schematic diagram of a fault handling method for a cloud computing platform provided by an embodiment of the present invention. Figure 3 As shown, the acquisition of the initial fault detection model includes the following steps:

[0077] S410: preprocess the historical fault data, and divide the preprocessed historical fault data into a training set and a test set.

[0078] The operation data of the cloud computing platform is input into the anomaly detection model for anomaly detection, and the obtained detection data is used as the historical fault data. The historical fault data is also preprocessed by data cleaning, feature extraction and standardization, and important features related to the fault, such as CPU usage and network latency, are extracted to obtain a standard data set. Optionally, the standard data set can be divided into a training set and a test set, where the training set is used to train the model, accounting for 70%-80% of the standard data set, and the test set is used to evaluate the model performance, accounting for 20%-30% of the standard data set.

[0079] S420: Train the initial prediction model using the features and labels of the training set to learn normal and fault feature patterns.

[0080] In an embodiment of the present invention, the initial prediction model can adopt a random forest algorithm model or a support vector machine algorithm model, and the initial prediction model can be trained using the features and labels of the training set to learn the characteristic patterns of normal and faulty conditions, and the parameters of the initial prediction model can be adjusted to minimize the loss function.

[0081] S430, performing cross-validation evaluation on the trained initial prediction model through the test set to obtain an evaluation result.

[0082] In the embodiment of the present invention, the generalization ability of the trained initial prediction model can be evaluated by cross-validation, and the cross-validation method includes but is not limited to K-fold cross-validation and leave-one-out cross-validation. By dividing the test set into multiple parts and training and testing the model on different parts, the performance of the model on different data can be more comprehensively understood, and indicators such as accuracy, recall rate and F1 score can be calculated as evaluation results.

[0083] S440, adjusting the parameters of the initial prediction model according to the evaluation results, and retraining using the training set to obtain an initial fault prediction model.

[0084] Optionally, determine in which aspects the initial prediction model performs poorly based on indicators such as accuracy, recall, and F1 score, such as overfitting or underfitting, or low prediction accuracy for certain specific categories. Determine the parameters that need to be adjusted based on the model type and evaluation results. For example, for the random forest algorithm model, the number of trees, maximum depth, and minimum number of samples may need to be adjusted. Retrain the initial prediction model with adjusted parameters using the training set to obtain the initial fault prediction model to ensure its accuracy in fault detection.

[0085] S500 , optimizing the initial fault detection model, updating the initial fault detection model according to the real-time fault data output by the initial fault detection model, and obtaining a fault detection model.

[0086] Figure 4 FIG. 4 is a fourth schematic diagram of a method for troubleshooting a cloud computing platform provided by an embodiment of the present invention. Figure 4 As shown, the acquisition of the fault detection model includes the following steps:

[0087] S510 , performing hyperparameter adjustment and feature selection on the initial fault detection model to obtain an optimized initial fault detection model.

[0088] Hyperparameters are parameters set before the initial fault detection model starts the learning process, and they have a significant impact on the performance of the model. In an embodiment of the present invention, methods such as cross-validation, grid search, and random search can be used to find the best hyperparameter combination of the initial fault detection model, thereby improving the performance of the model.

[0089] Feature selection is the process of identifying and retaining key features that have a positive impact on model performance while eliminating redundant or irrelevant features. In an embodiment of the present invention, key features can be identified through feature importance analysis, while redundant features can be eliminated, effectively simplifying the model structure, improving computational efficiency, and improving the accuracy of the model.

[0090] S520 , performing incremental learning and regular retraining on the optimized initial fault detection model according to the real-time fault data.

[0091] In the embodiment of the present invention, when new real-time fault data is obtained, an incremental learning algorithm is used to enable the model to predict the new data based on the existing parameters, and to update the parameters based on the difference between the predicted results and the true labels, so as to gradually learn and adapt to the new data. This incremental update process enables the initial fault detection model to gradually adapt to the new data without retraining the entire initial fault detection model, thereby greatly reducing the computational overhead and time consumption.

[0092] Over time, fault data may change, for example, new fault data categories may appear, or the distribution of existing fault data categories may change. These changes may cause the performance of the initial fault detection model to degrade. In order to maintain the predictive power of the model, the initial fault detection model needs to be retrained regularly with the latest data set to ensure that the initial fault detection model is always consistent with the latest data and adapts to data changes.

[0093] S530: dynamically adjust the model parameters of the initial fault detection model using an online learning algorithm to obtain a fault detection model.

[0094] In the embodiment of the present invention, an online learning algorithm such as stochastic gradient descent is used to randomly select some samples in each iteration to update the model parameters. The model can be optimized by adjusting parameters such as learning rate and batch size, thereby significantly reducing the amount of calculation and improving training efficiency. Optionally, the parameters can also be dynamically adjusted by monitoring the deviation between the model prediction results and the actual results, effectively improving the adaptability and accuracy of the model.

[0095] S600: Perform fault detection on the cloud computing platform according to the fault detection model to obtain a fault detection result.

[0096] In the embodiment of the present invention, the fault detection model is deployed to the cloud computing platform, the operation data of each node in the cloud computing platform is collected in real time, the real-time collected operation data is pre-processed and input into the fault detection model, and the fault detection model is used to analyze the operation data to determine whether there is a fault in the cloud computing platform and the type of fault, and obtain the fault detection result. Based on the real-time operation data and fault detection results, the cloud computing platform can give early warning and notify the operation and maintenance personnel through SMS, email or message platform, so as to ensure the stability and reliability of the cloud computing platform and provide users with high-quality services.

[0097] As a further implementation of the embodiment of the present invention, Figure 5 FIG. 5 is a fifth schematic diagram of a method for troubleshooting a cloud computing platform provided by an embodiment of the present invention. Figure 5 As shown, the method also includes the following steps:

[0098] S700: Establish a fault repair system according to the target fault type and set automatic fault repair rules.

[0099] In the embodiment of the present invention, an expert system or a rule-based engine is used to define automatic repair rules for common faults, and the automatic repair rules are set according to the target fault type.

[0100] S800, use the reinforcement learning model to perform reinforcement learning on the fault repair system to obtain the automatic fault repair strategy.

[0101] Figure 6 6 is a sixth schematic diagram of a method for troubleshooting a cloud computing platform provided by an embodiment of the present invention. Figure 6 As shown, the automatic fault repair strategy is obtained by the following steps:

[0102] S810. Setting a reinforcement learning environment for the fault repair system, where the reinforcement learning environment includes a state space, an action space, and a reward function.

[0103] In an embodiment of the present invention, the reinforcement learning environment of the fault repair system can be a cloud computing platform management system, the state space can be a real-time system performance indicator, the action space can be an adjustment of resource configuration, and the reward function can include reward mechanisms such as system stability and performance optimization. The defined reward function positively rewards behaviors that are beneficial to system stability and performance, and negative rewards are used to suppress behaviors that may cause failures.

[0104] S820. Use the reinforcement learning model to explore and sample the reinforcement learning environment to obtain sampling data.

[0105] In an embodiment of the present invention, the reinforcement learning model can adopt a Q-learning network model or a deep Q network model, and the reinforcement learning model is used for decision-making training in a complex environment. The reinforcement learning model continuously takes actions in the reinforcement learning environment, randomly explores the environment, and gradually accumulates experience, thereby collecting sampled data such as state, action, and reward.

[0106] S830. Use the sampled data to train the reinforcement learning model so that the reinforcement learning model can find the optimal reward strategy of the reinforcement learning environment and obtain the automatic fault repair strategy.

[0107] In the embodiment of the present invention, the reinforcement learning model is trained by using sampled data through experience replay and target network. The reinforcement learning model gradually finds the strategy that maximizes the long-term reward in the reinforcement learning environment through multiple iterations and updating the policy function. During the training process, the reinforcement learning environment strategy is continuously adjusted, and the action selection of the reinforcement learning environment is optimized by continuously interacting with the reinforcement learning environment until the optimal reward strategy of the reinforcement learning environment is found, and the automatic fault repair strategy is obtained.

[0108] S900: Select a corresponding automatic fault repair strategy according to the fault detection result to automatically repair the fault.

[0109] In the embodiments of the present invention, a corresponding automatic fault repair strategy can be selected according to the fault detection result, for example, based on the elastic scaling characteristics of the cloud computing platform, resources can be automatically adjusted (such as dynamically allocating virtual machines and restarting services); when a network-level fault occurs, traffic switching or service redundancy can be automatically performed to reduce downtime; for database and application-level faults, snapshots can be automatically restored and database recovery operations can be performed, etc. Optionally, after the fault is repaired, the repair process is recorded and the effect is analyzed to optimize the subsequent automatic repair strategy.

[0110] The fault handling method of the cloud computing platform provided in the embodiment of the present invention can analyze a large amount of historical operation data through machine learning and deep learning technology to establish an intelligent fault diagnosis model. These models can not only accurately identify the fault type, but also locate the root cause of the fault, thereby reducing false positives and false negatives. Through a data-driven approach, the efficiency and accuracy of fault diagnosis are significantly improved, and can adapt to complex and dynamic cloud computing environments. At the same time, according to the fault diagnosis results, the repair strategy can be automatically selected and implemented, which not only shortens the recovery time of the cloud computing platform, but also improves the availability of the service.

[0111] Example 2

[0112] Based on the same technical concept, an embodiment of the present invention provides a fault handling device for a cloud computing platform. Figure 7 is a first schematic diagram of the structure of a fault handling device for a cloud computing platform provided by an embodiment of the present invention, such as Figure 7 As shown, the fault handling device 700 of the cloud computing platform includes:

[0113] The demand analysis module 710 is used to determine the target fault type that needs to be detected on the cloud computing platform;

[0114] The data processing module 720 is used to obtain the operation data of the cloud computing platform, perform data preprocessing on the operation data, and obtain a standard data set;

[0115] The first training module 730 is used to perform cluster analysis on the standard data set according to the target fault type to obtain normal state data, and use the normal state data to train the machine learning algorithm model to obtain an abnormality detection model;

[0116] The second training module 740 is used to input the operation data into the anomaly detection model to obtain historical fault data, establish a historical data set based on the historical fault data, and use the historical data set to train the initial prediction model to obtain an initial fault detection model;

[0117] The model optimization module 750 is used to optimize the initial fault detection model, update the initial fault detection model according to the real-time fault data output by the initial fault detection model, and obtain the fault detection model;

[0118] The fault detection module 760 is used to perform fault detection on the cloud computing platform according to the fault detection model to obtain a fault detection result.

[0119] As a further implementation of the embodiment of the present invention, Figure 8 is a second schematic diagram of the structure of a fault handling device for a cloud computing platform provided by an embodiment of the present invention, such as Figure 8 As shown, the fault handling device 700 of the cloud computing platform also includes:

[0120] System construction module 770, used to establish a fault repair system according to the target fault type and set fault automatic repair rules;

[0121] A reinforcement learning module 780 is used to perform reinforcement learning on the fault repair system using a reinforcement learning model to obtain an automatic fault repair strategy;

[0122] The fault repair module 790 is used to select a corresponding automatic fault repair strategy according to the fault detection result to automatically repair the fault.

[0123] As a preferred implementation mode of an embodiment of the present invention, the fault handling device 700 of the cloud computing platform can be designed with a redundant system and a high-availability architecture to ensure the stability, reliability and scalability of the device, wherein the design of the redundant system includes multi-zone deployment and load balancing, and the high-availability architecture includes master-slave replication and distributed storage.

[0124] Multi-zone deployment deploys devices in different geographic areas or data centers to reduce the impact of single point failures. For example, if one area fails, other areas can take over traffic and services. Through distributed deployment, devices can automatically switch to other areas when a failure occurs in one area, ensuring continuous availability of services.

[0125] Load balancing uses a load balancer to distribute traffic to multiple backend servers, and can dynamically distribute the load based on the performance of the server, the type of request, and other strategies. Load balancing ensures that no single server is overloaded, improves the device's response speed and processing power, and provides failover capabilities. If a server fails, traffic can be automatically forwarded to other servers.

[0126] Master-slave replication sets up a master database and multiple slave databases at the database level. The master database handles write requests, while the slave database handles read requests. Data changes are synchronized to the slave databases in real time or periodically. Master-slave replication improves the concurrent performance of read operations, and when the master database fails, a slave database can be quickly promoted to the master database, thereby achieving rapid recovery.

[0127] Distributed storage uses a distributed storage system to store data on multiple nodes to ensure data redundancy and backup between nodes. Distributed storage provides high availability and data persistence. If a node fails, data can still be recovered from other nodes, reducing the risk of data loss.

[0128] The fault handling device of the cloud computing platform provided in the embodiment of the present invention performs self-optimization by continuously learning and accumulating historical fault data, and improves fault detection and repair strategies, which not only enhances the intelligence level of the cloud computing platform, but also has stronger adaptability when facing new types of faults. Through continuous iteration and updating, cloud computing can maintain a high service quality and stability, and significantly improve the reliability, efficiency and user experience of the cloud computing platform.

[0129] It can be understood that the implementation method of the fault handling method of the cloud computing platform described in the above embodiment 1 is also applicable to this embodiment and can achieve the same technical effect, so it will not be repeated here.

[0130] Example 3

[0131] Based on the same concept, an embodiment of the present invention further provides an electronic device, Fig. 9 is a schematic diagram of the structure of an electronic device provided by an embodiment of the present invention, such as Fig. 9 As shown, the electronic device 900 may include: a processor 910, a communication interface 920, a memory 930 and a communication bus 940, wherein the processor 910, the communication interface 920 and the memory 930 communicate with each other through the communication bus 940. The processor 910 may call the logic instructions in the memory 930 to execute the steps of the fault handling method of the cloud computing platform described in the above embodiments. For example, it includes:

[0132] S100, determining a target fault type to be detected on the cloud computing platform;

[0133] S200, obtaining operation data of the cloud computing platform, performing data preprocessing on the operation data, and obtaining a standard data set;

[0134] S300, performing cluster analysis on the standard data set according to the target fault type to obtain normal state data, and using the normal state data to train a machine learning algorithm model to obtain an anomaly detection model;

[0135] S400, inputting the operation data into the anomaly detection model to obtain historical fault data, establishing a historical data set based on the historical fault data, and using the historical data set to train an initial prediction model to obtain an initial fault detection model;

[0136] S500, optimizing the initial fault detection model, updating the initial fault detection model according to the real-time fault data output by the initial fault detection model, and obtaining a fault detection model;

[0137] S600: Perform fault detection on the cloud computing platform according to the fault detection model to obtain a fault detection result.

[0138] The processor 910 may be a central processing unit (CPU). The processor may also be other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, or a combination of the above chips.

[0139] In addition, the logic instructions in the above-mentioned memory 930 can be implemented in the form of a software functional unit and can be stored in a computer-readable storage medium when it is sold or used as an independent product. Based on such an understanding, the technical solution of the present invention can be essentially or partly embodied in the form of a software product that contributes to the prior art. The computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, etc. Various media that can store program codes.

[0140] The memory 930 may include a program storage area and a data storage area, wherein the program storage area may store an operating system, an application required for at least one function; the data storage area may store data created by the processor, etc. In addition, the memory may include a high-speed random access memory, and may also include a non-transient memory, such as at least one disk storage device, a flash memory device, or other non-transient solid-state storage device. In some embodiments, the memory may optionally include a memory remotely disposed relative to the processor, and these remote memories may be connected to the processor via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0141] Example 4

[0142] Based on the same concept, an embodiment of the present invention further provides a computer-readable storage medium, which stores a computer program, and the computer program includes at least one code segment, which can be executed by a main control device to control the main control device to implement the steps of the cloud computing platform fault handling method described in the above embodiments. For example, it includes:

[0143] S100, determining a target fault type to be detected on the cloud computing platform;

[0144] S200, obtaining operation data of the cloud computing platform, performing data preprocessing on the operation data, and obtaining a standard data set;

[0145] S300, performing cluster analysis on the standard data set according to the target fault type to obtain normal state data, and using the normal state data to train a machine learning algorithm model to obtain an anomaly detection model;

[0146] S400, inputting the operation data into the anomaly detection model to obtain historical fault data, establishing a historical data set based on the historical fault data, and using the historical data set to train an initial prediction model to obtain an initial fault detection model;

[0147] S500, optimizing the initial fault detection model, updating the initial fault detection model according to the real-time fault data output by the initial fault detection model, and obtaining a fault detection model;

[0148] S600: Perform fault detection on the cloud computing platform according to the fault detection model to obtain a fault detection result.

[0149] Based on the same technical concept, an embodiment of the present invention further provides a computer program, which is used to implement the above method embodiment when executed by a main control device.

[0150] The computer program may be stored in whole or in part on a computer-readable storage medium packaged with the processor, or may be stored in whole or in part on a memory not packaged with the processor.

[0151] Based on the same technical concept, an embodiment of the present invention further provides a processor, which is used to implement the above method embodiment. The above processor can be a chip.

[0152] In summary, the fault handling method, device, electronic device and storage medium of the cloud computing platform provided by the present invention can analyze a large amount of historical operation data through machine learning and deep learning technology to establish an intelligent fault diagnosis model. These models can not only accurately identify the fault type, but also locate the root cause of the fault, thereby reducing false positives and false negatives. Through a data-driven approach, the efficiency and accuracy of fault diagnosis have been significantly improved, and can adapt to complex and dynamic cloud computing environments. At the same time, according to the fault diagnosis results, the repair strategy can be automatically selected and implemented, which not only shortens the recovery time of the cloud computing platform, but also improves the availability of the service. By continuously learning and accumulating historical fault data for self-optimization, improving fault detection and repair strategies, not only the intelligence level of the cloud computing platform is enhanced, but also it has a stronger adaptability when facing new faults. Through continuous iteration and updating, cloud computing can maintain a high service quality and stability, and significantly improve the reliability, efficiency and user experience of the cloud computing platform.

[0153] Reference to "embodiments" herein means that a particular feature, structure, or characteristic described in conjunction with the embodiments may be included in at least one embodiment of the present application. The appearance of the phrase in various locations in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment that is mutually exclusive with other embodiments. It is explicitly and implicitly understood by those skilled in the art that the embodiments described herein may be combined with other embodiments.

[0154] The above-mentioned embodiments only express several implementation methods of the present invention, and the description thereof is relatively specific and detailed, but it cannot be understood as limiting the scope of the patent of the present invention. It should be pointed out that, for ordinary technicians in this field, several variations and improvements can be made without departing from the concept of the present invention, which all belong to the protection scope of the present invention. Therefore, the protection scope of the patent of the present invention shall be subject to the attached claims.

[0155] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for troubleshooting a cloud computing platform, characterized in that: The method comprises: Determine the target fault types that need to be detected on the cloud computing platform; Acquiring the operating data of the cloud computing platform, performing data preprocessing on the operating data, and obtaining a standard data set; Performing cluster analysis on the standard data set according to the target fault type to obtain normal state data, and using the normal state data to train a machine learning algorithm model to obtain an abnormality detection model; Inputting the operating data into the anomaly detection model to obtain historical fault data, establishing a historical data set based on the historical fault data, and using the historical data set to train an initial prediction model to obtain an initial fault detection model; Optimizing the initial fault detection model, updating the initial fault detection model according to the real-time fault data output by the initial fault detection model, and obtaining a fault detection model; Fault detection is performed on the cloud computing platform according to the fault detection model to obtain a fault detection result.

2. The method for troubleshooting a cloud computing platform according to claim 1, characterized in that: The method further comprises: Establishing a fault repair system according to the target fault type and setting fault automatic repair rules; Using a reinforcement learning model to perform reinforcement learning on the fault repair system to obtain an automatic fault repair strategy; The corresponding automatic fault repair strategy is selected according to the fault detection result to automatically repair the fault.

3. The method for troubleshooting a cloud computing platform according to claim 1, characterized in that: The method of performing cluster analysis on the standard data set according to the target fault type to obtain normal state data, and using the normal state data to train a machine learning algorithm model to obtain an abnormality detection model includes: Using a cluster analysis algorithm to classify the standard data set according to the target fault type to obtain the normal state data; The machine learning algorithm model is trained by the normal state data to learn the normal behavior pattern to detect data that deviates from the normal behavior pattern, thereby obtaining the abnormality detection model.

4. The method for troubleshooting a cloud computing platform according to claim 3, characterized in that: The step of establishing a historical data set according to the historical fault data, training an initial prediction model using the historical data set, and obtaining an initial fault detection model comprises: Performing data preprocessing on the historical fault data, and dividing the preprocessed historical fault data into a training set and a test set; The initial prediction model is trained by using the features and labels of the training set to learn the characteristic patterns of normal and fault conditions; Performing cross-validation evaluation on the trained initial prediction model through the test set to obtain an evaluation result; The parameters of the initial prediction model are adjusted according to the evaluation result, and the training set is used for retraining to obtain an initial fault prediction model.

5. The method for troubleshooting a cloud computing platform according to claim 1, characterized in that: The optimizing the initial fault detection model, updating the initial fault detection model according to the real-time fault data output by the initial fault detection model, and obtaining the fault detection model comprises: Performing hyperparameter adjustment and feature selection on the initial fault detection model to obtain an optimized initial fault detection model; Performing incremental learning and periodic retraining on the optimized initial fault detection model according to the real-time fault data; An online learning algorithm is used to dynamically adjust the model parameters of the initial fault detection model to obtain a fault detection model.

6. The method for troubleshooting a cloud computing platform according to claim 2, characterized in that: The method of using a reinforcement learning model to perform reinforcement learning on the fault repair system to obtain an automatic fault repair strategy includes: Setting a reinforcement learning environment for the fault repair system, wherein the reinforcement learning environment includes a state space, an action space, and a reward function; Using the reinforcement learning model to perform exploration sampling on the reinforcement learning environment to obtain sampling data; The reinforcement learning model is trained using the sampled data so that the reinforcement learning model finds the optimal reward strategy of the reinforcement learning environment and obtains the automatic fault repair strategy.

7. A fault handling device for a cloud computing platform, characterized in that: The device comprises: A demand analysis module is used to determine the target fault types that need to be detected on the cloud computing platform; A data processing module, used to obtain the operation data of the cloud computing platform, perform data preprocessing on the operation data, and obtain a standard data set; A first training module is used to perform cluster analysis on the standard data set according to the target fault type to obtain normal state data, and use the normal state data to train a machine learning algorithm model to obtain an abnormality detection model; A second training module is used to input the operating data into the anomaly detection model to obtain historical fault data, establish a historical data set based on the historical fault data, and use the historical data set to train an initial prediction model to obtain an initial fault detection model; A model optimization module, used to optimize the initial fault detection model, update the initial fault detection model according to the real-time fault data output by the initial fault detection model, and obtain a fault detection model; The fault detection module is used to perform fault detection on the cloud computing platform according to the fault detection model to obtain a fault detection result.

8. The fault handling device for a cloud computing platform according to claim 7, characterized in that: The device also includes: A system building module, used to establish a fault repair system according to the target fault type and set automatic fault repair rules; A reinforcement learning module, used to perform reinforcement learning on the fault repair system using a reinforcement learning model to obtain an automatic fault repair strategy; The fault repair module is used to select the corresponding automatic fault repair strategy according to the fault detection result to automatically repair the fault.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: The processor executes the computer program to implement the fault handling method for the cloud computing platform according to any one of claims 1 to 6.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the fault handling method of the cloud computing platform according to any one of claims 1 to 6 is implemented.

Images