Fault self-recovery method and device for cloud native platform

Abstract

The invention provides a fault self-recovery method and device for a cloud native platform, and the method comprises the steps: inputting real-time data into a fault prediction model after the real-time data of a monitoring index of the cloud native platform is obtained, so as to determine a fault prediction result of the cloud native platform through the fault prediction model, and under the condition of determining that a target fault occurs in the cloud native platform through the fault prediction result, obtaining diagnosis data related to the target fault in the cloud native platform, inputting the diagnosis data into the large model to determine a repair suggestion for the target fault through the large model, and performing self-healing processing on the cloud native platform according to the repair suggestion. Therefore, under the condition that it is determined that the cloud native platform breaks down, the diagnosis data related to the fault is analyzed through the large model, so that the trouble of manual analysis is reduced, repair suggestions for the fault are rapidly given through the large model, the repair efficiency of the cloud native platform is improved, and the time required for the cloud native platform to return to normal is shortened.

Description

Technical Field

[0001] This application relates to the field of computer technology, and in particular to fault self-healing methods and apparatus for cloud-native platforms. Background Technology

[0002] Cloud-native platforms are susceptible to various types of failures during operation, such as network failures, hardware failures, service unresponsiveness, and container failures. These failures can impact user experience and even cause significant financial losses. In related technologies, after a cloud-native platform failure, the diagnostic data related to the failure is typically analyzed manually to provide remediation suggestions. Based on these suggestions, the cloud-native platform then attempts to self-heal. However, manual analysis of diagnostic data is time-consuming and relies on experience-based judgment, resulting in a slow troubleshooting and remediation process and a prolonged time required for the cloud-native platform to return to normal operation. Summary of the Invention

[0003] This application proposes a fault self-healing method and apparatus for cloud-native platforms.

[0004] One embodiment of this application proposes a fault self-healing method for a cloud-native platform. The method includes: acquiring real-time data of monitoring indicators of the cloud-native platform; inputting the real-time data into a fault prediction model to determine a fault prediction result for the cloud-native platform; if the fault prediction result determines that a target fault has occurred in the cloud-native platform, acquiring diagnostic data related to the target fault in the cloud-native platform; inputting the diagnostic data into a large model to determine repair suggestions for the target fault; and performing self-healing processing on the cloud-native platform according to the repair suggestions.

[0005] Another embodiment of this application proposes a fault self-healing device for a cloud-native platform. The device includes: a first acquisition module for acquiring real-time data of monitoring indicators of the cloud-native platform; a first determination module for inputting the real-time data into a fault prediction model to determine a fault prediction result for the cloud-native platform through the fault prediction model; a second acquisition module for acquiring diagnostic data related to the target fault in the cloud-native platform when the fault prediction result determines that a target fault has occurred in the cloud-native platform; a second determination module for inputting the diagnostic data into a large model to determine a repair suggestion for the target fault through the large model; and a self-healing processing module for performing self-healing processing on the cloud-native platform according to the repair suggestion.

[0006] Another embodiment of this application proposes an electronic device, including: 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 self-healing method for cloud-native platforms disclosed in the embodiments of this application.

[0007] Another embodiment of this application proposes a computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the fault self-healing method for cloud-native platforms disclosed in the embodiments of this application.

[0008] Another embodiment of this application proposes a computer program product, including a computer program that, when executed by a processor, implements the fault self-healing method for cloud-native platforms disclosed in the embodiments of this application.

[0009] The technical solutions provided by the embodiments of this application bring at least the following beneficial effects:

[0010] After acquiring real-time monitoring data of the cloud-native platform, the data is input into a fault prediction model to determine the fault prediction result. If the fault prediction result confirms a target fault in the cloud-native platform, diagnostic data related to the target fault is acquired and input into a large-scale model to determine remediation suggestions. Based on these suggestions, the cloud-native platform undergoes self-healing. Therefore, by analyzing fault-related diagnostic data through a large-scale model when a fault is determined based on the fault prediction model, the tediousness of manual analysis is reduced. The large-scale model quickly provides remediation suggestions for the fault, improving the remediation efficiency of the cloud-native platform and reducing the time required for it to return to normal operation.

[0011] Other effects of the above-mentioned alternative methods will be described below in conjunction with specific embodiments. Attached Figure Description

[0012] The accompanying drawings are provided for a better understanding of this solution and do not constitute a limitation of this application. Wherein:

[0013] Figure 1 This is a flowchart illustrating a fault self-healing method for a cloud-native platform according to an embodiment of this application.

[0014] Figure 2 This is a flowchart illustrating a fault self-healing method for a cloud-native platform according to another embodiment of this application.

[0015] Figure 3 This is a flowchart illustrating a fault self-healing method for a cloud-native platform according to another embodiment of this application.

[0016] Figure 4 This is a flowchart illustrating a fault self-healing method for a cloud-native platform according to another embodiment of this application.

[0017] Figure 5 This is a schematic diagram of the structure of a fault self-healing device for a cloud-native platform according to an embodiment of this application.

[0018] Figure 6 This is a schematic diagram of the structure of an electronic device shown in an exemplary embodiment of this application. Detailed Implementation

[0019] Embodiments of the present invention are described in detail below. Examples of these embodiments are illustrated in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and intended to explain this application, and should not be construed as limiting this application.

[0020] The following describes a fault self-healing method and apparatus for a cloud-native platform according to embodiments of this application, with reference to the accompanying drawings.

[0021] Figure 1 This is a flowchart illustrating a fault self-healing method for a cloud-native platform according to an embodiment of this application. It should be noted that the execution entity of the fault self-healing method for a cloud-native platform provided in this embodiment is a fault self-healing device for the cloud-native platform. This fault self-healing device can be implemented in software and / or hardware. In this embodiment, the fault self-healing device for the cloud-native platform can be an electronic device, or it can be configured within an electronic device.

[0022] In this example embodiment, the electronic device may include a terminal device, a server, etc. The terminal device may be a PC (Personal Computer), a mobile device, a tablet computer, etc., and this embodiment does not specifically limit it.

[0023] like Figure 1 As shown, the fault self-healing method of this cloud-native platform may include:

[0024] Step 101: Obtain real-time data of monitoring metrics for the cloud-native platform.

[0025] In some embodiments, real-time data of monitoring metrics of the cloud-native platform can be obtained from the metrics monitoring system in the cloud-native platform.

[0026] In this embodiment, there may be one or more monitoring indicators, and the number of monitoring indicators is not specifically limited in this embodiment.

[0027] Step 102: Input real-time data into the fault prediction model to determine the fault prediction result of the cloud-native platform through the fault prediction model.

[0028] It should be noted that the above-mentioned fault prediction model was pre-trained.

[0029] In some embodiments, one possible way to train the fault prediction model is to combine sample data of monitoring metrics of the cloud-native platform and sample faults of the cloud-native platform determined based on the sample data to train the initial fault prediction model to obtain the fault prediction model.

[0030] Step 103: If the fault prediction results indicate that a target fault has occurred on the cloud-native platform, obtain the diagnostic data related to the target fault in the cloud-native platform.

[0031] It should be noted that the data source types for diagnostic data in this embodiment may include at least two.

[0032] In other words, the diagnostic data mentioned above was obtained from different types of data acquisition devices.

[0033] As an example, target logs related to the target fault in the cloud-native platform can be obtained from a first data acquisition terminal used to store logs of the cloud-native platform, and target content usage information related to the target fault in the cloud-native platform can be obtained from a second data acquisition terminal that stores memory usage information in the cloud-native platform.

[0034] Step 104: Input the diagnostic data into the large model to determine repair recommendations for the target fault.

[0035] In this embodiment, corresponding prompts can be generated based on diagnostic data. These prompts instruct the analysis of the diagnostic data to identify the cause of the fault and, based on the analysis results, to provide repair suggestions for the target fault. The prompts are then input into a large model, enabling the model to analyze the diagnostic data based on the prompts and provide repair suggestions for the target fault. Thus, by guiding the large model with prompts, the repair suggestions generated by the large model for the target fault are improved.

[0036] In some embodiments, to further improve the efficiency of the large model in obtaining repair suggestions for the target fault, key data can be extracted from the diagnostic data to obtain target key data. This target key data is then input into the large model to determine repair suggestions for the target fault. This reduces the amount of data input to the large model while ensuring the accuracy of the repair suggestions obtained by the large model for the target fault.

[0037] It should be noted that the repair suggestions in this embodiment can be in a preset format, such as a JOSN format. That is, the repair suggestions can be in a JOSN format. The preset format can also be other types of formats. This embodiment does not specifically limit the preset format.

[0038] Step 105: Perform self-healing on the cloud-native platform according to the repair recommendations.

[0039] In some embodiments, step 105 may be implemented as follows: obtaining the self-healing workflow template associated with the target fault; determining the self-healing workflow for the target fault based on the repair suggestions and the self-healing workflow template; and performing self-healing processing on the cloud-native platform according to the self-healing workflow. Thus, based on the repair suggestions and the self-healing workflow template, a self-healing workflow required to repair the target fault in the cloud-native platform is generated, and the cloud-native platform is self-healed according to the self-healing workflow, thereby achieving automatic self-healing processing of the cloud-native platform, reducing manual operations during the restoration of the cloud-native platform to normal operation, improving the repair efficiency of the cloud-native platform, and reducing the time required for the cloud-native platform to return to normal operation.

[0040] The fault self-healing method for cloud-native platforms provided in this application, after acquiring real-time data of monitoring indicators of the cloud-native platform, inputs the real-time data into a fault prediction model to determine the fault prediction result of the cloud-native platform. If the fault prediction result determines that a target fault has occurred in the cloud-native platform, it acquires diagnostic data related to the target fault in the cloud-native platform and inputs the diagnostic data into a large model to determine repair suggestions for the target fault. Based on the repair suggestions, the cloud-native platform undergoes self-healing processing. Therefore, when a fault is determined to have occurred in the cloud-native platform based on the fault prediction model, the large model analyzes the fault-related diagnostic data, reducing the burden of manual analysis. The large model quickly provides repair suggestions for the fault, helping to improve the repair efficiency of the cloud-native platform and reduce the time required for the cloud-native platform to return to normal.

[0041] In some embodiments, to avoid the potential for new problems to arise from directly performing self-healing processing on the cloud-native platform based on a self-healing workflow, a target fault can be simulated in a digital twin model of the cloud-native platform. The digital twin model is then subjected to self-healing processing according to the self-healing workflow. If the service availability of the self-healed digital twin model exceeds a preset threshold, the cloud-native platform is then subjected to self-healing processing again according to the self-healing workflow. This reduces the likelihood of new problems arising from directly performing self-healing processing on the cloud-native platform based on a self-healing workflow. To clarify this process, the following section combines... Figure 2The fault self-healing method of the cloud-native platform in this embodiment is further described in an exemplary manner.

[0042] Figure 2 This is a flowchart illustrating a fault self-healing method for a cloud-native platform according to another embodiment of this application.

[0043] like Figure 2 As shown, the method may include:

[0044] Step 201: Obtain real-time data of monitoring metrics for the cloud-native platform.

[0045] Step 202: Input real-time data into the fault prediction model to determine the fault prediction result of the cloud-native platform through the fault prediction model.

[0046] Step 203: If the fault prediction results indicate that a target fault has occurred on the cloud-native platform, obtain the diagnostic data related to the target fault in the cloud-native platform.

[0047] Step 204: Input the diagnostic data into the large model to determine repair recommendations for the target fault.

[0048] It should be noted that for a detailed description of steps 201 to 204, please refer to the relevant descriptions in other embodiments, which will not be repeated here.

[0049] Step 205: Obtain the self-healing workflow template associated with the target fault.

[0050] Step 206: Based on the repair suggestions and self-healing workflow template, determine the self-healing workflow for the target fault.

[0051] In some embodiments, the self-healing workflow template may include parameter value filling positions for target parameters. Correspondingly, a possible implementation of the self-healing workflow that determines the target fault based on the repair suggestion and the self-healing workflow template is as follows: determine the target parameter value of the target parameter based on the repair suggestion; fill the parameter value into the parameter value filling positions to obtain the self-healing workflow. This allows the parameter value of the target parameter in the self-healing workflow template to be dynamically configured based on the target parameter value determined by the repair suggestion, avoiding the problem of static parameters in the self-healing workflow.

[0052] Step 207: Determine the digital twin model of the cloud-native platform and inject the target fault into the digital twin model to simulate the target fault in the digital twin model.

[0053] In some embodiments, digital twin modeling can be performed on the cloud-native platform to obtain a digital twin model of the cloud-native platform. Thus, through digital twin technology, an accurate digital twin model of the cloud-native platform is constructed.

[0054] Step 208: Perform self-healing processing on the digital twin model according to the self-healing workflow.

[0055] In some embodiments, to ensure the accuracy of subsequent processing, the target parameter value of the target parameter in the self-healing workflow can be verified for compliance. If the target parameter value passes the compliance verification, the digital twin model is self-healed according to the self-healing workflow.

[0056] It should be noted that if the target parameter value fails the compliance check, a check failure message will be output.

[0057] Step 209: If the service availability of the self-healing digital twin model exceeds a preset threshold, perform self-healing processing on the cloud-native platform according to the self-healing workflow.

[0058] The preset threshold is set according to actual needs. For example, the preset threshold can be 0.95, 0.92, or 0.97, etc. This embodiment does not specifically limit the value of the preset threshold.

[0059] In this embodiment, after determining the self-healing workflow required for fault recovery of the cloud-native platform, the target fault is simulated in the digital twin model of the cloud-native platform, and self-healing processing is performed on the digital twin model according to the self-healing workflow. If the service availability of the digital twin model after self-healing exceeds a preset threshold, self-healing processing is then performed on the cloud-native platform according to the self-healing workflow. This reduces the possibility that self-healing processing of the cloud-native platform based directly on the self-healing workflow may cause new problems to arise in the cloud-native platform.

[0060] Figure 3 This is a flowchart illustrating a fault self-healing method for a cloud-native platform according to another embodiment of this application.

[0061] like Figure 3 As shown, the method may include:

[0062] Step 301: Obtain real-time data of monitoring metrics for the cloud-native platform.

[0063] Step 302: Input real-time data into the fault prediction model to determine the fault prediction result of the cloud-native platform through the fault prediction model.

[0064] Step 303: If the fault prediction results indicate that a target fault has occurred on the cloud-native platform, obtain the diagnostic data related to the target fault in the cloud-native platform.

[0065] Step 304: Input the diagnostic data into the large model to determine repair recommendations for the target fault through the large model.

[0066] Step 305: Obtain the self-healing workflow template associated with the target fault.

[0067] Step 306: Based on the repair suggestions and self-healing workflow template, determine the self-healing workflow for the target fault.

[0068] Step 307: Perform self-healing processing on the cloud-native platform according to the self-healing workflow.

[0069] It should be noted that for a detailed description of steps 301 to 307, please refer to the relevant descriptions in other embodiments, which will not be repeated here.

[0070] Step 308: If the self-healing process based on the cloud-native platform determines that the cloud-native platform has successfully self-healed, obtain the historical self-healing success rate of the cloud-native platform.

[0071] Step 309: Obtain the resource cost consumed by performing self-healing processing on the cloud-native platform based on the self-healing workflow.

[0072] Step 310: Determine the corresponding target reward value based on resource costs and historical self-healing success rate.

[0073] In some embodiments, reward values ​​can be calculated based on resource costs and historical self-healing success rates to obtain the corresponding target reward value.

[0074] In some embodiments, the corresponding target reward value can be calculated based on resource costs and historical self-healing success rates using a reward value calculation formula.

[0075] As an example, the formula for calculating the target reward value can be:

[0076] reward = A × 0.7 - B × 0.3

[0077] In the formula, A represents the historical self-healing success rate, and B represents the resource cost.

[0078] Step 311: Train the fault prediction model using reinforcement learning based on the target reward value.

[0079] In some embodiments, if the self-healing process based on the cloud-native platform determines that the self-healing has failed, the target fault and the corresponding self-healing workflow are saved to a preset database. This facilitates subsequent research and processing based on the faults and corresponding self-healing workflows in the preset database.

[0080] In this embodiment, the model parameters of the fault prediction model can be adjusted according to the target reward value, and reinforcement learning can be applied to the adjusted fault prediction model. This enables autonomous optimization of the fault prediction model, which helps improve the accuracy of fault prediction using the reinforcement learning-trained model.

[0081] It should be noted that the above steps in this embodiment can be performed by a single electronic device, or by different electronic devices, or by different modules within an electronic device. This embodiment does not specifically limit the subject performing the above steps.

[0082] To ensure a clear understanding of this application, the following will be combined with... Figure 4 The fault self-healing method of the cloud-native platform in this embodiment is described by way of example.

[0083] Figure 4 This is a flowchart illustrating a fault self-healing method for a cloud-native platform according to another embodiment of this application. It should be noted that the fault self-healing device for the cloud-native platform in this embodiment may include: a fault prediction terminal, an intelligent diagnostic agent terminal, a large model, a self-healing workflow terminal, a sandbox verification terminal, and an optimization strategy terminal. It should also be noted that the fault prediction terminal, intelligent diagnostic agent, sandbox verification terminal, and optimization strategy terminal in this embodiment can communicate with the cloud-native platform.

[0084] It should be noted that the cloud-native platform in this embodiment is a Kubernetes (K8s) cluster as its core, which is used to build an overall technical system that supports containerized application deployment, elastic scaling and automated operation and maintenance.

[0085] like Figure 4 As shown, the method may include:

[0086] Step 401: The fault prediction terminal obtains real-time data of the monitoring metrics of the cloud-native platform.

[0087] In some embodiments, real-time data of monitoring metrics of the cloud-native platform can be obtained from the metrics monitoring system in the cloud-native platform.

[0088] In this embodiment, there may be one or more monitoring indicators, and the number of monitoring indicators is not specifically limited in this embodiment.

[0089] Step 402: The fault prediction end determines the fault prediction result of the cloud-native platform based on the real-time data of the monitoring indicators through the fault prediction model.

[0090] Step 403: If the fault prediction results determine that a target fault has occurred on the cloud-native platform, the intelligent diagnostic agent obtains diagnostic data related to the target fault from multiple different data collection terminals used to collect data from the cloud-native platform.

[0091] It should be noted that different data collection terminals may collect different data dimensions when collecting data from cloud-native platforms. For example, one data collection terminal may collect log data from the native platform, while another may collect distributed tracing data from the cloud-native platform.

[0092] Step 404: The intelligent diagnostic agent preprocesses the acquired diagnostic data.

[0093] The preprocessing may include at least one of the following: data cleaning, key data extraction, etc. This embodiment does not specifically limit the preprocessing.

[0094] Step 405: The intelligent diagnostic agent processes the pre-processed diagnostic data through a large model to determine repair recommendations for the target fault.

[0095] In some embodiments, the intelligent diagnostic agent provides preprocessed diagnostic data to the large model. Correspondingly, the large model processes the preprocessed diagnostic data to obtain repair suggestions and provides the repair suggestions to the intelligent diagnostic agent.

[0096] In other embodiments, the intelligent diagnostic agent can generate corresponding prompts based on preprocessed diagnostic data and preset prompt templates. These prompts instruct the analysis of the preprocessed diagnostic data to obtain repair suggestions for the target fault. Correspondingly, the prompts are provided to the large model, which then analyzes the preprocessed diagnostic data based on the prompts to obtain repair suggestions for the target fault. Therefore, by providing prompts to the large model, it is possible to effectively guide the model to analyze the fault causes based on the preprocessed diagnostic data and generate repair suggestions for the target fault based on the analysis results, thus improving the accuracy of the repair suggestions obtained by the large model.

[0097] It should be noted that the repair suggestions in this embodiment can be in JSON format.

[0098] Step 406: The intelligent diagnostic agent determines the target parameter value of the target parameter based on the repair suggestion and provides the target parameter value to the self-healing workflow end, so that the self-healing workflow end adds the target parameter value to the parameter value filling position corresponding to the target parameter in the self-healing workflow template associated with the target fault, thereby obtaining the self-healing workflow.

[0099] In this embodiment, a predefined self-healing workflow template is used to support quick invocation and reduce the hassle of repeated configuration.

[0100] The parameter value filling position refers to the position where the parameter value is to be filled.

[0101] It should be noted that the target parameter in this embodiment can be one or more, and the number of target parameters is not specifically limited in this embodiment.

[0102] Step 407: The self-healing workflow sends a verification request to the sandbox verification end, wherein the verification request is used to instruct the self-healing workflow used to repair the target fault to be verified through the digital twin model of the cloud-native platform.

[0103] Step 408: The sandbox verification end performs compliance verification on the target parameter values ​​of the target parameters in the self-healing workflow.

[0104] Step 409: If the target parameter value passes the compliance verification, the sandbox verification terminal performs digital twin processing on the cloud-native platform to obtain a digital twin model of the cloud-native platform, and injects the target fault into the digital twin model to simulate the target fault in the digital twin model.

[0105] Step 410: The sandbox verification end performs self-healing processing on the digital twin model according to the self-healing workflow, and obtains the self-healing processing result of the digital twin model.

[0106] Step 411: The sandbox verification end provides the self-healing processing results of the digital twin model to the self-healing workflow end.

[0107] Step 412: If the self-healing workflow determines that the service availability of the digital twin model exceeds a preset threshold based on the self-healing processing result of the digital twin model, the sandbox verification end provides the self-healing workflow to the cloud-native platform, and the cloud-native platform executes the self-healing workflow accordingly.

[0108] The preset threshold is set in advance according to actual needs; for example, the preset threshold can be 85%.

[0109] Step 413: The cloud-native platform provides the self-healing processing results to the self-healing workflow.

[0110] Step 414: The self-healing workflow provides the self-healing processing results to the optimization strategy.

[0111] Step 415: Based on the self-healing results, the optimization strategy determines the corresponding target reward value when the cloud-native platform is confirmed to have successfully self-healed, according to the historical self-healing success rate of the cloud-native platform and the resource cost consumed in performing self-healing on the cloud-native platform based on the self-healing workflow.

[0112] In other embodiments, when the self-healing process fails based on the optimization strategy, the target fault and the corresponding self-healing workflow are added to a preset database to facilitate subsequent processing based on the faults and corresponding self-healing workflows in the preset database.

[0113] Step 416: The optimization strategy determines the adjustment parameters of the fault prediction model based on the target reward value.

[0114] Step 417: The optimization strategy adjusts the parameters of the fault prediction model according to the adjustment parameters, so as to perform the next round of fault prediction through the adjusted fault prediction model.

[0115] In this embodiment, the separation between prediction and execution in traditional solutions is resolved by associating the faults predicted by the fault prediction model with the corresponding self-healing workflow template. Furthermore, this embodiment determines the target parameter values ​​based on the corresponding repair suggestions and adds these values ​​to the self-healing workflow template, thus addressing the issue of static parameters in self-healing workflow templates in related technologies. Additionally, this embodiment supports multiple monitoring metrics to determine fault prediction for cloud-native platforms, replacing traditional single-metric alarm mechanisms. Moreover, this embodiment uses a large model to analyze the causes of faults from diagnostic data collected from multiple different data acquisition terminals, reducing the need for manual analysis of diagnostic data to obtain repair suggestions and improving the efficiency of obtaining repair suggestions. Furthermore, in this embodiment, after obtaining the self-healing workflow, sandbox verification is performed on the workflow, thereby verifying its robustness. During sandbox verification, the target reference values ​​of the target parameters in the workflow are first verified for compliance, thus intercepting obvious violations. Additionally, during sandbox verification, the target faults from the cloud-native platform are injected into the digital twin model of the cloud-native platform, ensuring that the state of the digital twin model is consistent with the state of the cloud-native platform, further improving the robustness of the verified self-healing workflow. Moreover, in this embodiment, after the self-healing workflow is executed on the cloud-native platform, if the self-healing is successful based on the cloud-native platform's self-healing results, the optimization strategy determines the corresponding target reward value. Based on the target reward value, reinforcement learning is applied to the fault prediction model. If the cloud-native platform fails to self-heal, the target fault and the corresponding self-healing workflow are saved to the database to accumulate experience.

[0116] Figure 5 This is a schematic diagram of the structure of a fault self-healing device for a cloud-native platform according to an embodiment of this application.

[0117] like Figure 5 As shown, the fault self-healing device 500 of the cloud-native platform may include a first acquisition module 501, a first determination module 502, a second acquisition module 503, a second determination module 504, and a self-healing processing module 505, wherein:

[0118] The first acquisition module 501 is used to acquire real-time data of monitoring metrics of the cloud-native platform.

[0119] The first determining module 502 is used to input real-time data into the fault prediction model in order to determine the fault prediction result of the cloud-native platform through the fault prediction model.

[0120] The second acquisition module 503 is used to acquire diagnostic data related to the target fault in the cloud-native platform when the target fault is determined to have occurred in the cloud-native platform through fault prediction results.

[0121] The second determination module 504 is used to input diagnostic data into the large model in order to determine repair recommendations for the target fault through the large model.

[0122] The self-healing module 505 is used to perform self-healing on the cloud-native platform based on repair suggestions.

[0123] In one embodiment of this application, the self-healing processing module 505 includes:

[0124] The acquisition unit is used to acquire the self-healing workflow template associated with the target fault;

[0125] The determination unit is used to determine the self-healing workflow for the target fault based on the repair recommendations and the self-healing workflow template.

[0126] The self-healing unit is used to perform self-healing processing on the cloud-native platform according to the self-healing workflow.

[0127] In one embodiment of this application, the self-healing workflow template includes: a parameter value filling position for the target parameter and a determination unit, specifically used for: determining the target parameter value of the target parameter according to the repair suggestion; and filling the target parameter value into the parameter value filling position to obtain the self-healing workflow.

[0128] In one embodiment of this application, the apparatus further includes: a processing module, configured to determine a digital twin model of a cloud-native platform, inject a target fault into the digital twin model to simulate the target fault in the digital twin model; and perform self-healing processing on the digital twin model according to a self-healing workflow;

[0129] The self-healing module 505 is specifically used to: perform self-healing processing on the cloud-native platform according to the self-healing workflow when the service availability of the digital twin model after self-healing exceeds a preset threshold.

[0130] In one embodiment of this application, the processing module is specifically used to: perform digital twin modeling on the cloud-native platform to obtain a digital twin model of the cloud-native platform.

[0131] In one embodiment of this application, the device may further include:

[0132] The compliance verification module is used to verify the compliance of the target parameter values ​​in the self-healing workflow.

[0133] The specific method by which the processing module performs self-healing processing on the digital twin model according to the self-healing workflow is as follows: when the target parameter value of the target parameter passes the compliance verification, the digital twin model performs self-healing processing according to the self-healing workflow.

[0134] In one embodiment of this application, the device may further include: a data extraction module, used to extract key data from diagnostic data to obtain target key data;

[0135] The second determining module 504 is specifically used for:

[0136] The target's key data is input into a large model to determine repair recommendations for the target fault.

[0137] In one embodiment of this application, the device may further include:

[0138] The reinforcement learning processing module is used to obtain the historical self-healing success rate of the cloud-native platform when the self-healing processing result based on the cloud-native platform determines that the self-healing of the cloud-native platform is successful; obtain the resource cost consumed by performing self-healing processing on the cloud-native platform based on the self-healing workflow; determine the corresponding target reward value based on the resource cost and the historical self-healing success rate; and perform reinforcement learning training on the fault prediction model based on the target reward value.

[0139] In one embodiment of this application, the device may further include: a storage module, used to save the target fault and the corresponding self-healing workflow to a preset database when the self-healing process based on the cloud-native platform determines that the cloud-native platform self-healing has failed.

[0140] It should be noted that the foregoing explanation of the fault self-healing method embodiment for cloud-native platforms also applies to the fault self-healing device for cloud-native platforms in this embodiment, and will not be repeated here.

[0141] The cloud-native platform fault self-healing device provided in this application, after acquiring real-time data of monitoring indicators of the cloud-native platform, inputs the real-time data into a fault prediction model to determine the fault prediction result of the cloud-native platform. If the fault prediction result determines that a target fault has occurred in the cloud-native platform, it acquires diagnostic data related to the target fault in the cloud-native platform and inputs the diagnostic data into a large model to determine repair suggestions for the target fault. Based on the repair suggestions, the cloud-native platform performs self-healing processing. Therefore, when a fault is determined to have occurred in the cloud-native platform based on the fault prediction model, the large model analyzes the fault-related diagnostic data, reducing the hassle of manual analysis. The large model quickly provides repair suggestions for the fault, helping to improve the repair efficiency of the cloud-native platform and reduce the time required for the cloud-native platform to return to normal.

[0142] According to embodiments of this application, this application also provides an electronic device, wherein the electronic device includes: a processor; a memory for storing processor-executable instructions; wherein the processor is configured to execute instructions to implement the fault self-healing method for a cloud-native platform as proposed in any of the foregoing embodiments.

[0143] As an example, Figure 6 This is a schematic diagram of the structure of an electronic device shown in an exemplary embodiment of this application.

[0144] like Figure 6 As shown, the aforementioned electronic device 600 may further include:

[0145] The memory 610 and processor 620 are connected by a bus 630, which connects different components (including the memory 610 and the processor 620). The memory 610 stores a computer program, and when the processor 620 executes the program, it implements the fault self-healing method of the cloud-native platform according to the embodiments of this application.

[0146] Bus 630 represents one or more of several bus architectures, including a memory bus or memory controller, a peripheral bus, a graphics acceleration port, a processor, or a local bus using any of the various bus architectures. Examples of these architectures include, but are not limited to, the Industry Standard Architecture (ISA) bus, the Micro Channel Architecture (MAC) bus, the Enhanced ISA bus, the Video Electronics Standards Association (VESA) local bus, and the Peripheral Component Interconnect (PCI) bus.

[0147] Electronic device 600 typically includes a variety of electronic device readable media. These media can be any available media that can be accessed by electronic device 600, including volatile and non-volatile media, removable and non-removable media.

[0148] Memory 610 may also include computer system readable media in the form of volatile memory, such as random access memory (RAM) 640 and / or cache memory 650. Electronic device 600 may further include other removable / non-removable, volatile / non-volatile computer system storage media. By way of example only, storage system 660 can be used to read and write non-removable, non-volatile magnetic media (… Figure 6 Not shown; usually referred to as a "hard drive"). Although Figure 6 Not shown, a disk drive for reading and writing to a removable non-volatile disk (e.g., a "floppy disk") and an optical disk drive for reading and writing to a removable non-volatile optical disk (e.g., a CD-ROM, DVD-ROM, or other optical media) may be provided. In these cases, each drive may be connected to bus 630 via one or more data media interfaces. Memory 610 may include at least one program product having a set (e.g., at least one) of program modules configured to perform the functions of the embodiments of this application.

[0149] A program / utility 680 having a set (at least one) of program modules 670 may be stored in, for example, memory 610. Such program modules 670 include—but are not limited to—an operating system, one or more application programs, other program modules, and program data. Each or some combination of these examples may include an implementation of a network environment. Program modules 670 typically perform the functions and / or methods described in the embodiments of this application.

[0150] Electronic device 600 can also communicate with one or more external devices 690 (e.g., keyboard, pointing device, display 691, etc.), and with one or more devices that enable a user to interact with electronic device 600, and / or with any device that enables electronic device 600 to communicate with one or more other computing devices (e.g., network card, modem, etc.). This communication can be performed via input / output (I / O) interface 692. Furthermore, electronic device 600 can also communicate with one or more networks (e.g., local area network (LAN), wide area network (WAN), and / or public networks, such as the Internet) via network adapter 693. As shown, network adapter 693 communicates with other modules of electronic device 600 via bus 630. It should be understood that, although not shown in the figures, other hardware and / or software modules can be used in conjunction with electronic device 600, including but not limited to: microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, and data backup storage systems.

[0151] The processor 620 executes various functional applications and data processing by running programs stored in the memory 610.

[0152] It should be noted that the implementation process and technical principles of the electronic device in this embodiment are explained in the foregoing description of the fault self-healing method of the cloud-native platform in this application embodiment, and will not be repeated here.

[0153] In an exemplary embodiment, a computer-readable storage medium including instructions is also provided, such as a memory including instructions, which can be executed by a processor of an electronic device to perform the fault self-healing method for a cloud-native platform proposed in any of the above embodiments. Optionally, the computer-readable storage medium may be a ROM, random access memory (RAM), CD-ROM, magnetic tape, floppy disk, and optical data storage device, etc.

[0154] In an exemplary embodiment, a computer program product is also provided, including a computer program / instructions, which, when executed by a processor, implement the fault self-healing method for a cloud-native platform proposed in any of the above embodiments.

[0155] Other embodiments of this application will readily occur to those skilled in the art upon consideration of the specification and practice of the invention disclosed herein. This application is intended to cover any variations, uses, or adaptations of this application that follow the general principles of this application and include common knowledge or customary techniques in the art not disclosed herein. The specification and examples are to be considered exemplary only, and the true scope and spirit of this application are indicated by the following claims.

[0156] It should be understood that this application is not limited to the precise structure described above and shown in the accompanying drawings, and various modifications and changes can be made without departing from its scope. The scope of this application is limited only by the appended claims.

Claims

1. A fault self-healing method for a cloud-native platform, characterized in that, The method includes: Obtain real-time data of monitoring metrics from cloud-native platforms; The real-time data is input into the fault prediction model to determine the fault prediction result of the cloud-native platform through the fault prediction model; If the cloud-native platform is determined to have a target fault based on the fault prediction results, diagnostic data related to the target fault in the cloud-native platform is obtained. The diagnostic data is input into a large model to determine repair recommendations for the target fault. Based on the aforementioned repair recommendations, the cloud-native platform is subjected to self-healing processing.

2. The method as described in claim 1, characterized in that, The self-healing process for the cloud-native platform based on the repair recommendations includes: Obtain the self-healing workflow template associated with the target fault; Based on the repair recommendations and the self-healing workflow template, determine the self-healing workflow for the target fault; The cloud-native platform is self-healed according to the self-healing workflow.

3. The method as described in claim 2, characterized in that, The self-healing workflow template includes: parameter value padding for the target parameter; the step of determining the self-healing workflow for the abnormal monitoring indicator based on the repair suggestion and the self-healing workflow template includes: Based on the repair recommendations, determine the target parameter values ​​for the target parameters; The target parameter value is filled into the parameter value filling position to obtain the self-healing workflow.

4. The method as described in claim 3, characterized in that, Before performing self-healing processing on the cloud-native platform according to the self-healing workflow, the method further includes: A digital twin model of the cloud-native platform is determined, and the target fault is injected into the digital twin model to simulate the target fault in the digital twin model; The digital twin model is subjected to self-healing processing according to the self-healing workflow; The self-healing process for the cloud-native platform based on the self-healing workflow includes: If the service availability of the digital twin model after self-healing exceeds a preset threshold, the cloud-native platform will undergo self-healing processing according to the self-healing workflow.

5. The method as described in claim 4, characterized in that, Determining the digital twin model of the cloud-native platform includes: A digital twin model of the cloud-native platform is performed to obtain the digital twin model of the cloud-native platform.

6. The method as described in claim 4, characterized in that, Prior to performing self-healing processing on the digital twin model according to the self-healing workflow, the method further includes: Perform compliance verification on the target parameter values ​​of the target parameters in the self-healing workflow; The step of performing self-healing processing on the digital twin model according to the self-healing workflow includes: If the target parameter value passes the compliance verification, the digital twin model is subjected to self-healing processing according to the self-healing workflow.

7. The method as described in claim 2, characterized in that, The method further includes: Key data are extracted from the diagnostic data to obtain the target key data; The step of inputting the diagnostic data into a large model to determine repair recommendations for the target fault through the large model includes: The target key data is input into a large model to determine repair recommendations for the target fault.

8. The method according to any one of claims 2-7, characterized in that, After performing self-healing processing on the cloud-native platform according to the self-healing workflow, the method further includes: If the self-healing process of the cloud-native platform determines that the self-healing of the cloud-native platform is successful, the historical self-healing success rate of the cloud-native platform is obtained. Obtain the resource cost consumed by performing self-healing processing on the cloud-native platform based on the self-healing workflow; Based on the resource costs and the historical self-healing success rate, determine the corresponding target reward value; The fault prediction model is trained using reinforcement learning based on the target reward value.

9. The method as described in claim 8, characterized in that, The method further includes: If the self-healing process of the cloud-native platform determines that the self-healing of the cloud-native platform has failed, the target fault and the corresponding self-healing workflow are saved to a preset database.

10. A fault self-healing device for a cloud-native platform, characterized in that, The device includes: The first acquisition module is used to acquire real-time data of monitoring metrics of the cloud-native platform; The first determining module is used to input the real-time data into the fault prediction model in order to determine the fault prediction result of the cloud-native platform through the fault prediction model. The second acquisition module is used to acquire diagnostic data related to the target fault in the cloud-native platform when the fault prediction result determines that a target fault has occurred in the cloud-native platform. The second determining module is used to input the diagnostic data into a large model to determine repair recommendations for the target fault through the large model; The self-healing module is used to perform self-healing on the cloud-native platform according to the repair suggestions.

Images