A fault template generation method and apparatus

Abstract

The application discloses a fault template generation method and device, and relates to the technical field of computers. A specific implementation manner of the method comprises the following steps: obtaining fault information, determining a product corresponding to the fault information, determining a fault atom and parameters adapted to the fault atom according to the fault information and business information of the product, generating a fault template according to the fault information, the fault atom and the adapted parameters, simulating and practicing the fault template by using original fault data in the fault information, comparing a simulation and practice result with a fault result in the original fault data to obtain a difference value, readjusting the fault atom and the adapted parameters, generating the fault template and performing the simulation and practice operation until the difference value does not meet a preset adjustment condition, and obtaining a target fault template. The implementation manner can automatically generate the fault template, and can ensure that the fault template can highly restore an actual fault condition, so that the quality of the fault template is effectively guaranteed.

Description

Technical Field

[0001] This invention relates to the field of computer technology, and in particular to a method and apparatus for generating fault templates. Background Technology

[0002] Chaos engineering platforms are commonly used tools by testing and operations personnel to improve system stability. However, the programming of fault phenomena is quite difficult, and currently, it mainly relies on experts to build or verify them based on experience. This leads to a lack of reliable basis for fault template creation and inconsistent quality. To solve these problems, a standardized and systematic fault template generation system needs to be established to ensure that the quantity and quality of fault templates meet actual needs and improve the efficiency of selection and use. Summary of the Invention

[0003] In view of this, embodiments of the present invention provide a method and apparatus for generating fault templates, which can at least solve the problems of unreliable basis for creation and inconsistent quality in the prior art of manually creating fault templates.

[0004] To achieve the above objectives, according to one aspect of the present invention, a fault template generation method is provided, comprising:

[0005] Obtain fault information, determine the product corresponding to the fault information, and determine the fault atom and the parameters that are compatible with the fault atom based on the fault information and the product's business information;

[0006] Based on the fault information, fault atoms, and adaptation parameters, a fault template is generated, and the original fault data in the fault information is used to simulate and exercise the fault template.

[0007] By comparing the simulation results with the fault results in the original fault data, the difference values ​​are obtained;

[0008] In response to the difference value meeting the preset adjustment conditions, the fault atoms and adaptation parameters are readjusted, the fault template is regenerated, and a simulation exercise is performed until the difference value no longer meets the preset adjustment conditions, thus obtaining the target fault template.

[0009] Optionally, the method further includes determining a fault information group; wherein the process of determining the fault information group includes:

[0010] By using the historical fault management platform, a list of resolved faults is collected, and the category of each fault information in the fault list is determined so that fault information of the same category is grouped together.

[0011] Fault records are obtained from each fault information, and fault information in the same category and with the same or similar fault records are grouped together; wherein, fault records include one or more of the following: fault root cause records and fault solution records.

[0012] Optionally, determining the fault atom and parameters compatible with the fault atom based on fault information and product business information includes:

[0013] Using feature analysis, business characteristics are extracted from the business information; and fault characteristics are extracted from the fault information.

[0014] Based on the fault characteristics and business characteristics, and combined with the fault atom characteristics of the chaos platform, the fault atom and the parameters adapted to the fault atom are determined.

[0015] Optionally, after obtaining the target fault template, the method further includes:

[0016] Determine the fault impact score based on the fault information;

[0017] Determine the category to which the fault information belongs, count the number of faults of the same category occurring within a preset time period, and determine the number of business users of the product;

[0018] The additional attribute score for the fault is determined based on the number of occurrences, business information, number of business users, and fault information.

[0019] The score of the target fault template is determined based on the fault impact score and the fault additional attribute score; the score is used to measure the importance of the target fault template.

[0020] Optionally, the method further includes:

[0021] The pre-configured weights for the fault impact score and the fault additional attribute score are added to the usage weight information of the target fault template.

[0022] Receive information on weight modifications, and based on the modified weights, redetermine the score of the target fault template and the weight information to be used.

[0023] Optionally, after obtaining the target fault template, the method further includes:

[0024] Using text analysis methods, business text features are extracted from the business information;

[0025] Determine the category to which the fault information belongs, and use text analysis to determine the fault text features based on the category information, fault information, and fault atoms;

[0026] Based on the characteristics of business text and fault text, the classification of target fault templates is determined.

[0027] Optionally, the method further includes:

[0028] Based on the classification of the target fault template, query the fault module library to see if there is an existing fault template of the same classification.

[0029] If the query result indicates that the existing fault template and the target fault template exist, the existing fault template and the target fault template will be integrated and the template information will be updated based on the integrated content; or the existing fault template will be replaced with the target fault template.

[0030] To achieve the above objectives, according to another aspect of the present invention, a fault template generation apparatus is provided, comprising:

[0031] The analysis module is used to obtain fault information, determine the product corresponding to the fault information, and determine the fault atom and the parameters that are compatible with the fault atom based on the fault information and the product's business information.

[0032] The exercise module is used to generate a fault template based on fault information, fault atoms, and adaptation parameters, and to conduct simulation exercises on the fault template using the original fault data in the fault information.

[0033] The adjustment module is used to compare the simulation results with the fault results in the original fault data to obtain the difference value. In response to the difference value meeting the preset adjustment conditions, the fault atoms and adaptation parameters are readjusted, the fault template is regenerated, and the simulation operation is performed until the difference value no longer meets the preset adjustment conditions, thus obtaining the target fault template.

[0034] Optionally, the apparatus further includes determining a fault information group; wherein the process of determining the fault information group includes:

[0035] By using the historical fault management platform, a list of resolved faults is collected, and the category of each fault information in the fault list is determined so that fault information of the same category is grouped together.

[0036] Fault records are obtained from each fault information, and fault information in the same category and with the same or similar fault records are grouped together; wherein, fault records include one or more of the following: fault root cause records and fault solution records.

[0037] Optionally, the analysis module is used for:

[0038] Using feature analysis, business characteristics are extracted from the business information; and fault characteristics are extracted from the fault information.

[0039] Based on the fault characteristics and business characteristics, and combined with the fault atom characteristics of the chaos platform, the fault atom and the parameters adapted to the fault atom are determined.

[0040] Optionally, the device further includes a scoring module for:

[0041] Determine the fault impact score based on the fault information;

[0042] Determine the category to which the fault information belongs, count the number of faults of the same category occurring within a preset time period, and determine the number of business users of the product;

[0043] The additional attribute score for the fault is determined based on the number of occurrences, business information, number of business users, and fault information.

[0044] The score of the target fault template is determined based on the fault impact score and the fault additional attribute score; the score is used to measure the importance of the target fault template.

[0045] Optionally, the scoring module is further configured to:

[0046] The pre-configured weights for the fault impact score and the fault additional attribute score are added to the usage weight information of the target fault template.

[0047] Receive information on weight modifications, and based on the modified weights, redetermine the score of the target fault template and the weight information to be used.

[0048] Optionally, the device further includes a classification module for:

[0049] Using text analysis methods, business text features are extracted from the business information;

[0050] Determine the category to which the fault information belongs, and use text analysis to determine the fault text features based on the category information, fault information, and fault atoms;

[0051] Based on the characteristics of business text and fault text, the classification of target fault templates is determined.

[0052] Optionally, the classification module is further configured to:

[0053] Based on the classification of the target fault template, query the fault module library to see if there is an existing fault template of the same classification.

[0054] If the query result indicates that the existing fault template and the target fault template exist, the existing fault template and the target fault template will be integrated and the template information will be updated based on the integrated content; or the existing fault template will be replaced with the target fault template.

[0055] To achieve the above objectives, according to another aspect of the present invention, a fault template generation electronic device is provided.

[0056] The electronic device of this invention includes: one or more processors; and a storage device for storing one or more programs, which, when executed by the one or more processors, cause the one or more processors to implement any of the fault template generation methods described above.

[0057] To achieve the above objectives, according to another aspect of the present invention, a computer-readable medium is provided having a computer program stored thereon, which, when executed by a processor, implements any of the fault template generation methods described above.

[0058] To achieve the above objectives, according to another aspect of the present invention, a computing program product is provided. One such computing program product includes a computer program that, when executed by a processor, implements the fault template generation method provided in this embodiment of the invention.

[0059] According to the solution provided by the present invention, one embodiment of the invention has the following advantages or beneficial effects: by automatically generating and optimizing fault templates, it ensures that the generated fault templates can highly reproduce the actual fault situation, thereby improving the accuracy and efficiency of fault diagnosis and handling. Specifically, it can automatically adjust fault atoms and adaptation parameters, making the simulation results as consistent as possible with the original fault data. The final target fault template can be used more reliably for fault reproduction and analysis, improving the effectiveness and efficiency of fault drills. In this way, it solves the problems of unreliable basis, lack of unified standards, and inconsistent quality caused by relying on manual experience to generate fault templates, effectively ensuring the quality of fault templates. Moreover, this method can generate fault templates based on fault classification, thereby increasing the number of fault templates and covering as many fault phenomena as possible.

[0060] The further effects of the aforementioned unconventional alternative methods will be explained below in conjunction with specific implementation methods. Attached Figure Description

[0061] The accompanying drawings are provided to better understand the invention and are not intended to unduly limit the scope of the invention. Wherein:

[0062] Figure 1 This is a schematic diagram of the main process of a fault template generation method according to an embodiment of the present invention;

[0063] Figure 2 This is a flowchart illustrating the process of generating fault templates;

[0064] Figure 3 A flowchart illustrating an optional fault template generation method according to an embodiment of the present invention;

[0065] Figure 4 This is a flowchart illustrating the process of determining the score for the fault template;

[0066] Figure 5 This is a flowchart illustrating another optional fault template generation method according to an embodiment of the present invention;

[0067] Figure 6This is a flowchart illustrating the process of determining the classification of fault templates;

[0068] Figure 7 This is a flowchart illustrating a specific fault template generation method according to an embodiment of the present invention;

[0069] Figure 8 This is a schematic diagram of the main modules of a fault template generation device according to an embodiment of the present invention;

[0070] Figure 9 This is an exemplary system architecture diagram in which embodiments of the present invention can be applied;

[0071] Figure 10 This is a schematic diagram of the structure of a computer system suitable for implementing the embodiments of the present invention, such as a mobile device or server. Detailed Implementation

[0072] The following description, in conjunction with the accompanying drawings, illustrates exemplary embodiments of the present invention, including various details to aid understanding. These details should be considered merely exemplary. Therefore, those skilled in the art will recognize that various changes and modifications can be made to the embodiments described herein without departing from the scope and spirit of the invention. Similarly, for clarity and brevity, descriptions of well-known functions and structures are omitted in the following description.

[0073] It should be noted that the collection, gathering, updating, analysis, processing, use, transmission, and storage of user personal information involved in this disclosed technical solution all comply with relevant laws and regulations, are used for legitimate purposes, and do not violate public order and good morals. Necessary measures are taken to prevent unauthorized access to user personal information data and to safeguard user personal information security, network security, and national security.

[0074] Chaos engineering platforms are tools and frameworks used to test and improve system resilience. By intentionally introducing faults and uncertainties, they simulate problems in real-world environments to observe system performance and responses. They help development teams identify potential weaknesses, optimize system design, and ensure stability and availability in the face of unexpected events. Therefore, chaos engineering platforms are commonly used verification tools by testing and operations personnel to improve system stability, and also serve as an auxiliary tool for managers to prepare contingency plans and assess operational efficiency.

[0075] However, the compilation of fault templates (or fault simulation templates, fault phenomenon (or fault scenario) templates, or fault phenomenon simulation templates) in chaos engineering is too difficult for relevant personnel. It requires experts with in-depth understanding of the system to provide relevant advice based on their experience to ensure effective verification and protection of system stability and rapid identification of system weaknesses. This leads to the following problems:

[0076] 1) The creation of fault templates lacks reliable basis, leading to uncertainties in the evaluation of system stability and reliability. 2) The limited number of fault templates results in incomplete fault drills, failing to fully cover all possible fault phenomena. This reduces the number of discoverable system stability optimization points, further resulting in fewer emergency plans and more potential faults when system failures occur. 3) Fault templates are mainly contributed by various parties, lacking unified standards and measurement methods, leading to inconsistent quality and difficulty in ensuring their reliability and effectiveness. 4) The classification of fault templates is chaotic, relying mainly on individual subjective judgment without unified standards. This makes it difficult for relevant personnel to choose templates, further affecting the efficiency of template management and use.

[0077] To address these issues, a standardized and systematic fault template generation system needs to be established to ensure that the quantity and quality of templates meet actual needs, and to improve the efficiency of template selection and use through scientific and reasonable classification methods.

[0078] See Figure 1 The diagram shows the main flowchart of a fault template generation method provided by an embodiment of the present invention, which includes the following steps:

[0079] S101: Obtain fault information, determine the product corresponding to the fault information, and determine the fault atom and the parameters that are compatible with the fault atom based on the fault information and the product's business information;

[0080] S102: Generate a fault template based on the fault information, fault atoms, and adaptation parameters, and conduct a simulation exercise on the fault template using the original fault data in the fault information;

[0081] S103: Compare the simulation results with the fault results in the original fault data to obtain the difference value;

[0082] S104: In response to the difference value meeting the preset adjustment conditions, readjust the fault atoms and adaptation parameters, regenerate the fault template and perform simulation exercises until the difference value no longer meets the preset adjustment conditions, and obtain the target fault template.

[0083] In the above implementation, the chaos engineering fault simulation covers various types of networks, devices, operating systems, components, services, and applications. Selecting the appropriate fault atom for different fault phenomena is a key challenge. To effectively address this challenge, this solution proposes an automated fault phenomenon template generation mechanism. Through root cause analysis of fault phenomena, combined with business architecture and business attributes, the mechanism accurately reproduces fault phenomena in the form of fault templates, thereby improving the reliability and accuracy of fault templates.

[0084] In this context, a fault phenomenon refers to the abnormal behavior or state exhibited by a system or device when a fault occurs. It is an external manifestation that can be directly observed by the user or monitoring system, and usually reflects the existence of internal problems. By analyzing fault phenomena, the type and possible causes of the fault can be preliminarily determined, and corresponding diagnostic and repair measures can then be taken. The core steps of the fault template generation mechanism are as follows:

[0085] For step S101, the first step is to collect a list of resolved faults through a historical fault management platform. This might involve interfacing with multiple products to collect fault information when these products malfunction, and then providing solutions for handling them. The fault list includes multiple fault entries, for example, 100 fault entries. These are the faults that have been resolved; unresolved or unsuccessfully resolved faults are not considered in this solution. Taking the aforementioned 100 fault entries as an example, the category to which each fault entry belongs is determined. Then, fault entries within the same category are grouped together. Assuming there are 5 fault categories, this results in 5 groups.

[0086] Furthermore, the root causes and solutions for different faults also differ. Therefore, after grouping based on classification, further grouping can be performed based on root causes and solutions, that is, fault information with the same or similar classification, root causes, and solutions can be grouped together. In practice, it is also possible to consider only the root causes or only the similarity of solutions. This approach preferably considers both root causes and similarities of solutions. Determining whether fault records are similar can be done using a vector approach. For example, fault records can be converted into vector form, and the similarity between different fault record vectors can be calculated. A high similarity indicates similarity, otherwise, it indicates dissimilarity.

[0087] Multiple fault information groups are obtained through the aforementioned steps, such as 7 fault information groups. Then, a fault template is generated based on each fault information group, thus generating 7 fault templates. A fault information group can contain one or more fault information items. However, when there are fewer fault information items, such as a group containing only one fault information item, the fault template is essentially generated based on the fault information. The specific operations are as follows: Step 1: Obtain the business attributes and business architecture of the product, and extract the business characteristics of the business attributes and business architecture; Step 2: Extract fault characteristics from the fault information; Step 3: Based on the fault characteristics and business characteristics, combined with the fault atom characteristics of the chaos platform, determine the fault atom and the parameters adapted to the fault atom.

[0088] For step 1: First, identify the product corresponding to the fault information, i.e., the product at the time the fault occurred. In the context of chaos engineering and fault simulation, the term "product" typically refers to a specific software system, service, or application. Each fault corresponds to a specific business system or component where the fault occurred. Next, it's necessary to obtain the business attributes and business architecture of this product to calculate business features using feature analysis algorithms, thereby more accurately describing and classifying the fault phenomenon. If the fault information group includes multiple fault entries, assuming there are 10 fault entries in the group, then the business features of these 10 fault entries need to be integrated to obtain the business features corresponding to the group.

[0089] Suppose a failure occurs in the order processing module of an e-commerce system. First, determine that this module belongs to the e-commerce product category and obtain its business attributes (such as high concurrency and real-time requirements) and business architecture (such as microservice architecture and distributed database). Then, using feature analysis algorithms, extract business characteristics based on these attributes and architecture, such as "high concurrency processing," "order management," and "distributed transactions." These business characteristics can help to more accurately pinpoint the cause of the failure and optimize the failure response strategy.

[0090] Feature analysis algorithms include, but are not limited to, TF-IDF (Term Frequency-Inverse Document Frequency), TextRank (a graph-based ranking model for text processing), LDA (Latent Dirichlet Allocation), Word2Vec (Word to Vector model), and BERT (Bidirectional Encoder Representations from Transformers). These models and algorithms are widely used in the field of natural language processing for tasks such as feature extraction, text classification, and keyword extraction.

[0091] For step 2: This solution also needs to extract fault features from the fault information, such as configuration, system changes, hosts, Kubernetes, network devices, etc. The extraction method can use the aforementioned feature analysis algorithm, or other methods; this solution does not impose any restrictions. If the fault information group includes multiple fault information entries, assuming there are 10 fault information entries in the group, then the fault feature characteristics of these 10 fault information entries need to be integrated. The fault information mainly includes:

[0092] 1. Raw Fault Data: Records detailed data at the time of the fault, such as logs and monitoring metrics. Unlike other fault information, raw fault data is user feedback, such as records of platform inaccessibility, while other fault information is generated based on further analysis and processing after the fault occurs.

[0093] 2. Fault Root Cause Record: Describe the root cause of the fault, such as code defects, hardware failures, configuration errors, etc.

[0094] 3. Faulty host change information: Record the status changes of the host before and after the fault occurred, such as CPU usage, memory usage, etc.

[0095] 4. Fault Solution Record: Record the specific steps and measures to resolve the fault, including temporary measures and long-term solutions.

[0096] 5. Scope of Fault Impact: Describe the scope of the fault's impact on the business system, such as the number of affected users and the duration of service interruption.

[0097] 6. MTTR (Mean Time To Repair): Records the time required from the occurrence of a fault to full recovery. It includes the time for fault detection, isolation, repair, and restart. The shorter the MTTR, the better the system's maintainability and the faster it can restore service.

[0098] 7. Fault Level: Classified according to the severity of the fault, such as severe, high, medium and low.

[0099] 8. Fault Explosion Radius: Describes the possible range of the fault's spread and its potential impact.

[0100] By synthesizing this fault information, key fault characteristics can be extracted. For example, changes in system configuration can be extracted from the raw fault data, the history of system changes can be extracted from the root cause records, host performance metrics can be extracted from the faulty host change information, and the state changes of relevant components can be extracted from the information of Kubernetes and network devices. The extraction of these characteristics helps to understand the fault phenomena more comprehensively and provides accurate data support for subsequent fault simulation and reproduction.

[0101] For step 3: After obtaining the fault characteristics and business characteristics, the fault atom characteristics of the chaos platform are superimposed, and an algorithm is used to simulate and calculate the generation of suitable fault atoms and adaptation parameters. Here, a fault atom refers to the smallest fault unit in a system or software that can be independently identified and processed. It is a fundamental concept in fault management, used to describe and classify specific fault phenomena, helping to quickly locate problems and carry out effective repairs. Based on the determined fault atoms, a fault phenomenon can be generated.

[0102] Suppose an e-commerce system's order processing module experiences a high-concurrency request timeout failure. Extracted business characteristics include "high-concurrency processing," "order management," and "distributed transactions." Fault information indicates the failure is related to a recent Kubernetes (k8s) configuration change, affecting multiple host nodes and causing brief network instability. Features extracted from the fault information include configuration changes (Kubernetes resource limit adjustments), system changes (recent code deployments), host changes (abnormal load on multiple nodes), and network device instability. Combining the fault atomic characteristics of a chaos platform, such as network latency, increased CPU load, memory leaks, and disk corruption, algorithms are used for simulation calculations (e.g., matching, hits) to select "network latency" and "increased CPU load" as suitable fault atoms. Adaptation parameters are set to a random network latency of 100ms to 500ms and to increase CPU load to 80%-90%.

[0103] For steps S102-S104, the above steps automatically analyze the detailed fault records and fault phenomena, combine them with AI large-scale model recommendation algorithms, and summarize more information about the causes and architecture of such faults from big data. The fault is then automatically matched with fault atoms that can trigger it. This step describes how to automatically orchestrate and generate fault templates. Specifically, after obtaining the fault atoms and adaptation parameters, a fault template can be generated based on the fault information (which may not include resolution records; these can be appended to the end of the template, or multiple fault information resolution records can be integrated and appended), fault atoms, and adaptation parameters. The adaptation parameters are used to better simulate the fault; faults have parameters, and different fault adaptation parameters will produce different fault phenomena.

[0104] Continuing with the previous example, the fault template assumes the following information: 1) Fault scenario name: High-concurrency order processing module request timeout; 2) Fault description: Due to Kubernetes configuration changes and temporary network device instability, the order processing module experiences request timeouts under high concurrency; 3) Fault atoms: Network latency and increased CPU load; 4) Adaptation parameters: Random network latency of 100ms to 500ms, CPU load increased to 80%-90%; 5) Affected components: Order processing module, Kubernetes cluster, network devices; 6) Fault impact scope: Multiple host nodes; 7) Fault level: High; 8) Fault MTTR: ​​30 minutes; 9) Fault solution: Optimize Kubernetes resource configuration, adjust network device parameters, and enhance system fault tolerance.

[0105] After generating the fault template, simulation exercises are required. For example, using the template mentioned above, the original fault data from the simulation information is used to simulate the template, obtaining simulation results. The simulation results are then compared with the original fault results in the original fault data to obtain the difference values. However, it should be noted that when the template information group includes multiple template information, assuming 10 template information, a fault template needs to be generated based on these 10 template information, initial atoms, and adaptation parameters. Similarly, the original fault data from these 10 template information also needs to be used to test the fault template to obtain 10 simulation results, or any number of original fault data points can be randomly selected from these 10 template information for testing.

[0106] For the test results, it can be determined whether they all meet the preset adjustment conditions, such as at least one difference value being greater than a preset threshold, or the number of difference values ​​greater than the preset threshold being greater than 2 (this is just an example, and can be adjusted in practice). Alternatively, the average difference value can be calculated, and it can be determined whether the average difference value is greater than the preset threshold. If the preset adjustment conditions are met, the initial atoms and adaptation parameters need to be adjusted.

[0107] Assuming the simulation results show that the order processing module did indeed experience request timeouts after introducing network latency and increased CPU load, affecting multiple host nodes, the MTTR was 25 minutes, slightly lower than the 30 minutes in the original fault data. Furthermore, the peak load on some host nodes in the simulation reached 92%, while the highest load in the original fault data was 88%. Comparing the simulation results with the original fault data reveals the following differences: the MTTR decreased by 5 minutes, and the peak CPU load increased by 4 percentage points. These differences indicate that while the simulation largely reproduced the original fault phenomena, some deviations remain in certain details, requiring further optimization of the fault template to achieve the most accurate fault simulation possible.

[0108] See the overall process Figure 2 As shown:

[0109] 1. Obtain fault information and extract fault characteristics from the fault information; the fault information mainly includes original fault data, fault root cause records and fault solution records.

[0110] 2. Determine the product corresponding to the fault information, obtain the business attributes and business architecture of the product, and extract the business characteristics of the business attributes and business architecture;

[0111] 3. Based on the fault characteristics and business characteristics, and combined with the fault atom characteristics of the chaos platform, determine the fault atom and the parameters that are compatible with the fault atom;

[0112] 4. Generate a fault template based on the fault information, fault atoms, and adaptation parameters. Use the original fault data in the fault information to simulate and practice the fault template.

[0113] 5. Compare the simulation results with the fault results in the original fault data to obtain the difference values;

[0114] 6. If the difference value meets the preset adjustment conditions, readjust the fault atoms and adaptation parameters, regenerate the fault template, and perform simulation exercises until the difference value no longer meets the preset adjustment conditions, thus obtaining the target fault template.

[0115] The method provided in the above embodiments automatically generates and optimizes fault templates, ensuring that the generated fault templates can highly reproduce actual fault situations, thereby improving the accuracy and efficiency of fault diagnosis and handling. Specifically, it can automatically adjust fault atoms and adaptation parameters, making the simulation results as consistent as possible with the original fault data. The final target fault template can be used more reliably for fault reproduction and analysis, improving the effectiveness and efficiency of fault drills. In this way, it solves the problems of unreliable basis, lack of unified standards, and inconsistent quality caused by relying on manual experience to generate fault templates, effectively ensuring the quality of fault templates. Moreover, this method can generate fault templates based on fault classification, thereby increasing the number of fault templates and covering as many fault phenomena as possible.

[0116] See Figure 3 The diagram illustrates an optional fault template generation method according to an embodiment of the present invention, comprising the following steps:

[0117] S301: Determine the fault impact score based on the fault information;

[0118] S302: Determine the category to which the fault information belongs, and count the number of times the same category of faults occurs within a preset time period;

[0119] S303: Determine the number of business users for the product;

[0120] S304: Determine the additional attribute score for the fault based on the number of occurrences, business information, number of business users, and fault information;

[0121] S305: Determine the score of the target fault template based on the fault impact score and the fault additional attribute score; wherein, the score is used to measure the importance of the target fault template.

[0122] In the above implementation, after generating a fault template, it is necessary to measure the importance of the template, thereby solving the problem of "lack of a unified measurement method for fault templates". During the fault template generation process, the importance of the template is affected by factors such as the fault's business attributes, the number of business users, the scope of business impact, and the blast radius. Therefore, calculating the importance of the template is particularly important for business personnel when selecting templates for drills.

[0123] Based on data such as the business attributes of the fault phenomenon, the number of business users, the scope of impact, the blast radius, and the fault recovery MTTR, combined with data such as the historical occurrence frequency of faults, this solution proposes a weighting and scoring mechanism for fault templates to uniformly measure and evaluate them. The fault template weighting and scoring mechanism is as follows:

[0124] For step S301, the fault impact score is calculated by weighting intuitive data such as the fault impact range, fault blast radius, and fault level. It should be noted that the fault impact range and the fault blast radius are two different concepts. The fault impact range refers to the area or system part actually affected by the fault, including the number of affected users and business functions. The fault blast radius, on the other hand, refers to other related areas or system parts that the fault may spread to and indirectly affect; it not only includes the directly affected parts but also considers potential chain reactions.

[0125] If an e-commerce website's payment system malfunctions, preventing payment transactions from being completed, the impact will be felt by all users attempting to make payments and the entire transaction process. The blast radius of this malfunction could include order processing delays, inventory lock-up timeouts, a surge in customer service inquiries caused by payment failures, and even broader issues such as affecting supplier settlement cycles.

[0126] When there are multiple faults in a fault information group, the fault impact score is calculated by first considering the score of each dimension, and then summing them in a weighted manner to obtain the overall fault impact score. Assume there are three faults in the fault information group: payment system fault, user login fault, and inventory management system fault. Each fault can be scored in three dimensions: "fault impact scope," "fault blast radius," and "fault level," and then the overall fault impact score is obtained by weighted summation.

[0127] Payment system failure: Impact range: High (8 / 10), Explosion radius: High (8 / 10), Failure severity: Severe (9 / 10). User login failure: Impact range: Very high (10 / 10), Explosion radius: Medium (6 / 10), Failure severity: Important (7 / 10). Inventory management system failure: Impact range: Low (3 / 10), Explosion radius: Low (4 / 10), Failure severity: Moderate (5 / 10).

[0128] The impact range of the fault is 8 + 10 + 3 = 21, the blast radius is 8 + 6 + 4 = 18, and the fault level is 9 + 7 + 5 = 21. Next, weights are assigned according to the importance of each dimension; for example, the impact range is weighted at 40%, the blast radius at 30%, and the fault level at 30%. Then, a weighted sum is calculated for each fault: 21 * 0.4 + 18 * 0.3 + 21 * 0.3 = 20.1. This method yields the overall impact score for each fault information group.

[0129] For steps S302 to S304, the additional attribute score of the fault phenomenon is calculated by combining the MTTR required for fault recovery, the number of similar faults that have actually occurred in the past, the business attributes of the product to which the fault occurred, and the number of business users (referring to the number of users using the product) through feature analysis and weighting.

[0130] For example, the transaction module of an online payment system experienced five service outages in the past year, with an average recovery time of 2 hours (MTTR = 2 hours). This module is a core business component, directly impacting the transaction experience of millions of users. Based on this information, each factor can be weighted and scored: a longer MTTR results in a greater impact (8 points), higher outage frequency (7 points), core business attributes (9 points), and a large user base (10 points). Finally, the scores for each factor are summed or calculated using a specific formula to determine the additional attribute score for the outage.

[0131] It's important to note that the weights are determined based on the business type and company requirements. For example, the weights required for core and non-core business operations, as well as business operations used by external users and back-end support operations, will differ. Weighted calculations are used to assign more reasonable scores to faults with different impact levels. For instance, a fault with a very short MTTR but a very high frequency of occurrence will be given a recommended weight in the simulation compared to a fault that occurs only once but has a very long MTTR.

[0132] For example, the transaction module of an online payment system experienced five service outages in the past year, with an average recovery time (MTTR) of 2 hours per outage. This module is part of the company's core business and directly impacts the transaction experience of millions of users. Meanwhile, another back-end support module experienced 10 outages during the same period, but each had an MTTR of only 15 minutes. According to the company's requirements, core business operations and those used by external users have higher weighting, while non-core business operations and back-end support operations have lower weighting.

[0133] Therefore, in the weighted calculation, faults with longer MTTRs or higher frequency in core business operations will receive higher scores to reflect their greater impact. For example, a core business operation with a longer MTTR receives 9 points, a higher fault frequency receives 8 points, a critical business attribute receives 10 points, and a large user base receives 10 points; while a non-core business operation with a shorter MTTR receives 6 points, a higher fault frequency receives 7 points, a less important business attribute receives 5 points, and a smaller user base receives 4 points. This method allows for more reasonable scoring of faults with different impacts, ensuring that faults with a greater impact on business operations and user experience are prioritized.

[0134] The above explanation uses a single fault information as an example. In actual operation, when the fault information group includes multiple fault information, assuming there are 10 fault information, the additional fault attribute score can be determined based on these 10 fault information, the number of occurrences, the business attributes, and the total number of business users. The calculation method is similar to the calculation method in step S301.

[0135] For step S305, the score of the target fault template is determined based on the previously calculated fault impact score and fault additional attribute score. Specifically, different weights are pre-assigned to the fault impact score and fault additional attribute score, such as 0.7 and 0.3. The score of the target fault template is obtained through a weighted summation. Furthermore, these two weights, along with the score, can be used as the weight and score of the target fault template, for example, 0.7, 0.3, and 20.

[0136] By attaching weights to the target fault template, when the weights are modified, such as from 0.7 to 0.6, or from 0.3 to 0.4, the corresponding score of the target fault template also needs to be adjusted. After the target fault template is generated, when it is added to the fault template library, it needs to be comprehensively calculated in conjunction with the existing templates in the library to better evaluate the importance and ranking of the target fault template compared with existing fault templates. Therefore, the score calculated by this scheme can measure the importance of the target fault template.

[0137] See the overall process Figure 4 As shown:

[0138] 1. The impact score of a fault is calculated by weighting intuitive data such as the fault's impact range, blast radius, and fault level.

[0139] 2. By combining the MTTR required for fault recovery, the number of similar faults that have occurred in the past, and the business attributes and user volume of the product to which the fault occurred, the additional attribute score of the fault is calculated through feature analysis and weighting.

[0140] 3. Based on the impact score and additional attribute score after the fault occurs, calculate the weight and feature score of the fault template.

[0141] 4. Receive information on weight modifications, and based on the modified weights, redetermine the score and weight of the target fault template.

[0142] The method provided in the above embodiments comprehensively considers information such as the scope of the fault's impact and its frequency of occurrence, as well as the product's business attributes and the number of business users, to quantify the fault impact score and additional attribute score, thereby enabling a more accurate measurement of the importance of the target fault template. In this way, the importance and ranking of fault templates can be compared. Subsequently, when a user requests to find multiple fault templates, the more important fault templates can be prioritized and displayed at the top or in a prominent position, thus improving the overall efficiency and effectiveness of fault response and handling.

[0143] See Figure 5 The diagram illustrates another optional fault template generation method according to an embodiment of the present invention, including the following steps:

[0144] S501: Use text analysis to extract business text features from the business information;

[0145] S502: Determine the category to which the fault information belongs, and use text analysis to determine the fault text features based on the category information, fault information, and fault atoms;

[0146] S503: Determine the classification of the target fault template based on the characteristics of the business text and the fault text.

[0147] In the above implementation, during product iteration and fault platform development, a large number of faults and simulation methods are inevitably generated. After automatically sorting all the fault information to generate fault templates, the accumulated number of fault templates will reach a high level, and there will be multiple simulation schemes for similar fault phenomena. At this point, a robust classification and accurate recommendation mechanism is inevitably needed in the selection of fault templates. To address the challenge of fault template selection, this solution proposes a classification mechanism for fault templates based on a large model. Based on the importance and business adaptability of the fault templates, it accurately recommends the business scenarios that will use the templates for simulation, ensuring efficient and accurate template selection and adaptation.

[0148] It's important to note that fault classification and fault template classification are not the same. Fault classification is based on the nature, cause, and impact of the fault, such as network faults, hardware faults, and software faults. Determining the category to which fault information belongs is the existing method. For example, when a server becomes inaccessible, initial troubleshooting may reveal that it's due to a network configuration error. In this case, the fault is classified as a "network fault."

[0149] Fault template classification is designed to simulate faults, enabling rapid problem location and resolution. Fault template classification includes, but is not limited to: redundancy faults, disaster recovery faults, network faults, security faults, software faults, hardware faults, compatibility faults, and configuration faults. Fault template classification accurately describes fault types and enables efficient fault handling.

[0150] For steps S501-S503, the existing technology suffers from a chaotic classification of fault templates, relying mainly on subjective judgment without a unified standard. This makes it difficult for relevant personnel to choose templates, further affecting the efficiency of template management and use. To address this issue, this embodiment provides a corresponding solution, see [link to solution]. Figure 6 As shown.

[0151] First, identify the product corresponding to the fault information, obtain the business attributes and business architecture of each product, and automatically extract key business features using text data analysis techniques. For example, by analyzing product identifiers in fault reports or logs, detailed business attributes and architectural information for each product can be obtained. Using text data analysis techniques, business text features related to the product can be automatically extracted, such as core modules, user interaction paths, and transaction processes. These features will help understand the specific impact of the fault on the business.

[0152] Next, the fault information groups are categorized. Based on the fault category, the scope of the fault's impact, and the previously identified fault atoms or fault phenomena composed of multiple fault atoms, text data analysis techniques are further applied to automatically extract key fault features related to the fault. This step aims to identify the root cause of the fault and its impact on business processes, such as whether it involves core business, user volume, or transaction interruption.

[0153] Finally, key business features and key fault features are aggregated using text pattern mining algorithms to automatically analyze and integrate these features, generating fault template classifications. This process considers not only the technical details of the faults but also the business context and impact, ensuring that the fault templates accurately reflect different types of faults and their actual impact on the business. The final output fault template classifications can be used to quickly locate and handle similar faults, improving operational efficiency and response speed.

[0154] In addition, this solution can also store the generated fault template in a fault template library. The fault template library may already contain fault templates categorized across multiple types; typically, only one template is stored per category. Therefore, it's necessary to set the organization method according to actual needs. For example, a merging method can be used to integrate existing fault templates and the target fault template, updating the template information based on the integrated content. Another method is a replacement method, which directly deletes existing fault templates, retaining only the target fault template generated this time.

[0155] The method provided in the above embodiments aims to improve the quality of fault templates and provide a reliable basis and standard for system stability evaluation. By extracting text features from business information and fault information through text analysis technology and classifying them in conjunction with fault atoms, the classification of target fault templates can be determined more accurately. This approach not only improves the accuracy of fault template classification but also enhances the automation and intelligence of fault handling, thereby accelerating fault response speed and improving problem-solving efficiency.

[0156] See Figure 7 The diagram illustrates a specific fault template generation method according to an embodiment of the present invention, including the following steps:

[0157] 1. Historical Fault Management Platform: Collects a list of resolved faults. The fault list includes multiple fault information, each of which includes original fault data, fault root cause record, fault host change information, fault solution record, fault impact range, fault MTTR, fault level, fault blast radius, fault classification, etc. (Due to limited space in the figure, some content is not marked).

[0158] 2. Determine the category of each fault information in the fault list to group fault information of the same category together; or obtain fault records from each fault information and group fault information of the same category with the same or similar fault records together.

[0159] 3. Automatically generate fault templates: Based on fault information groups, the system automatically analyzes fault information and combines it with AI big data model recommendation algorithms to summarize more information on the causes and architecture of such faults from big data. It automatically matches fault atoms and adaptation parameters that can cause this fault and automatically arranges and generates fault templates.

[0160] 4. Chaos Engineering Platform: Transmits fault templates and fault information to the Chaos Engineering Platform so that the fault information can be used to simulate and exercise the fault templates.

[0161] By combining detailed historical fault data, such as fault level, fault impact range, blast radius, and fault MTTR, with the fault occurrence patterns in the large model, a comprehensive analysis and recommendation is made. A weighted overlay algorithm is used to automatically calculate the score and weight of the fault template to uniformly measure the template quality and the priority recommendation weight of the template.

[0162] By combining data such as product characteristics, business attributes, business architecture, and fault classification of historical faults with the applicable business and architecture suggestions recommended by the big model, the fault templates are identified through intelligent aggregation algorithms to uniformly label the fault templates.

[0163] 5. Optimized Simulation Exercise Template:

[0164] By conducting actual drills and analysis of fault templates, intelligent optimization of the templates is derived. The fault platform automatically recommends drill scenarios and solutions based on the occurrence of the fault. For example, by comparing the simulated drill results with the original fault results in the original fault data, the difference value is obtained. If the difference value meets preset adjustment conditions, the fault atoms and adaptation parameters are readjusted, the fault template is regenerated, and a simulated drill is performed until the difference value no longer meets the preset adjustment conditions, thus obtaining the target fault template. For example, during the simulation drill, staff can score the template and provide optimization suggestions to achieve template optimization.

[0165] 6. Fault Monitoring and Contingency Plan Optimization: By analyzing the handling methods of historical faults and the iterative upgrade process of monitoring, and with the addition of large model recommendation algorithms to provide differentiated suggestions for various faults in different types of business and architecture modes, fault monitoring strategies (such as improving business monitoring and business alarms) and contingency plan suggestions (such as fault contingency plan optimization and MTTR improvement) are automatically matched to achieve a more intelligent and efficient fault management and response mechanism.

[0166] The method provided in this invention automatically generates fault templates, monitoring suggestions, and solution suggestions through a fault management platform combined with a large-scale model recommendation algorithm. Furthermore, it optimizes the automatically generated fault templates by analyzing the results of fault simulation exercises and simultaneously recommends contingency plans to the fault management platform, forming a complete closed loop for the generation and training of chaotic engineering templates based on an AI large-scale model. This solution achieves significant optimization through the following methods:

[0167] First, historical faults from various systems are synchronized to the chaos engineering platform using a fault management platform, and then a large number of realistic fault templates are automatically generated using AI large-scale model recommendation algorithms. These fault templates not only cover the fault points identified from various systems, but also have important reference value for other systems, thereby effectively improving the reliability and practicality of the fault templates.

[0168] Secondly, based on historical fault information such as severity level, impact scope, MTTI (Mean Time To Interest), and classification, combined with the frequency of occurrence, the business attributes of the product to which the fault belongs, and the number of business users, a weighted score and weight are generated for the fault template. This method can standardize the measurement of template value and ensure the quality of chaos engineering fault templates.

[0169] In addition, by combining fault information and business information of the product to which the fault belongs with large model training, the system automatically identifies and classifies the fault templates of the chaos platform, providing relevant personnel with effective usage suggestions and guidance.

[0170] In summary, by integrating historical fault information with AI large-scale model technology, this invention achieves high-quality generation, standardized measurement, intelligent classification and recommendation, and continuous optimization of fault templates. This further enhances the automation and intelligence of fault template generation, ensuring that templates can continuously improve and adapt to the ever-changing system environment, thereby significantly improving system stability and fault handling efficiency.

[0171] See Figure 8 The diagram shows the main modules of a fault template generation device 800 provided in an embodiment of the present invention, including:

[0172] Analysis module 801 is used to acquire fault information, determine the product corresponding to the fault information, and determine the fault atom and the parameters that are compatible with the fault atom based on the fault information and the business information of the product.

[0173] The exercise module 802 is used to generate a fault template based on fault information, fault atoms and adaptation parameters, and to conduct simulation exercises on the fault template using the original fault data in the fault information.

[0174] The adjustment module 803 is used to compare the simulation results with the fault results in the original fault data to obtain the difference value; in response to the difference value meeting the preset adjustment conditions, the fault atoms and adaptation parameters are readjusted, the fault template is regenerated, and the simulation operation is performed until the difference value no longer meets the preset adjustment conditions, thus obtaining the target fault template.

[0175] The apparatus of the present invention further includes determining a fault information group; wherein, the process of determining the fault information group includes:

[0176] By using the historical fault management platform, a list of resolved faults is collected, and the category of each fault information in the fault list is determined so that fault information of the same category is grouped together.

[0177] Fault records are obtained from each fault information, and fault information in the same category and with the same or similar fault records are grouped together; wherein, fault records include one or more of the following: fault root cause records and fault solution records.

[0178] In the apparatus of this invention, the analysis module 801 is used for:

[0179] Using feature analysis, business characteristics are extracted from the business information; and fault characteristics are extracted from the fault information.

[0180] Based on the fault characteristics and business characteristics, and combined with the fault atom characteristics of the chaos platform, the fault atom and the parameters adapted to the fault atom are determined.

[0181] The apparatus for implementing this invention further includes a scoring module, used for:

[0182] Determine the fault impact score based on the fault information;

[0183] Determine the category to which the fault information belongs, count the number of faults of the same category occurring within a preset time period, and determine the number of business users of the product;

[0184] The additional attribute score for the fault is determined based on the number of occurrences, business information, number of business users, and fault information.

[0185] The score of the target fault template is determined based on the fault impact score and the fault additional attribute score; the score is used to measure the importance of the target fault template.

[0186] In the apparatus of this invention, the scoring module is further used for:

[0187] The pre-configured weights for the fault impact score and the fault additional attribute score are added to the usage weight information of the target fault template.

[0188] Receive information on weight modifications, and based on the modified weights, redetermine the score of the target fault template and the weight information to be used.

[0189] The apparatus for implementing this invention further includes a classification module, used for:

[0190] Using text analysis methods, business text features are extracted from the business information;

[0191] Determine the category to which the fault information belongs, and use text analysis to determine the fault text features based on the category information, fault information, and fault atoms;

[0192] Based on the characteristics of business text and fault text, the classification of target fault templates is determined.

[0193] In the apparatus of this invention, the classification module is further used for:

[0194] Based on the classification of the target fault template, query the fault module library to see if there is an existing fault template of the same classification.

[0195] If the query result indicates that the existing fault template and the target fault template exist, the existing fault template and the target fault template will be integrated and the template information will be updated based on the integrated content; or the existing fault template will be replaced with the target fault template.

[0196] Furthermore, the specific implementation details of the device described in the embodiments of the present invention have been described in detail in the above-described method, so the details will not be repeated here.

[0197] Figure 9An exemplary system architecture 900 to which embodiments of the present invention can be applied is shown, including terminal devices 901, 902, 903, network 904, and server 905 (only an example).

[0198] Terminal devices 901, 902, and 903 can be various electronic devices with displays and web browsing capabilities, and have various communication client applications installed. Users can use terminal devices 901, 902, and 903 to interact with server 905 via network 904 to receive or send messages, etc.

[0199] Network 904 is a medium used to provide a communication link between terminal devices 901, 902, 903 and server 905. Network 904 can include various connection types, such as wired or wireless communication links or fiber optic cables, etc.

[0200] Server 905 can be a server providing various services, such as a backend management server supporting shopping websites browsed by users using terminal devices 901, 902, and 903 (this is just an example). The backend management server can analyze and process data such as received product information query requests, and feed back the processing results (such as target push information, product information—this is just an example) to the terminal devices. It should be noted that the method provided in this embodiment of the invention is generally executed by server 905, and correspondingly, the apparatus is generally set in server 905.

[0201] It should be understood that Figure 9 The number of terminal devices, networks, and servers shown is merely illustrative. Depending on implementation needs, any number of terminal devices, networks, and servers can be included.

[0202] The following is for reference. Figure 10 It shows a schematic diagram of the structure of a computer system 1000 suitable for implementing a terminal device of the present invention. Figure 10 The terminal device shown is merely an example and should not impose any limitations on the functionality and scope of use of the embodiments of the present invention.

[0203] like Figure 10 As shown, the computer system 1000 includes a central processing unit (CPU) 1001, which can perform various appropriate actions and processes based on programs stored in read-only memory (ROM) 1002 or programs loaded from storage section 1008 into random access memory (RAM) 1003. The RAM 1003 also stores various programs and data required for the operation of the system 1000. The CPU 1001, ROM 1002, and RAM 1003 are interconnected via a bus 1004. An input / output (I / O) interface 1005 is also connected to the bus 1004.

[0204] The following components are connected to I / O interface 1005: an input section 1006 including a keyboard, mouse, etc.; an output section 1007 including a cathode ray tube (CRT), liquid crystal display (LCD), etc., and speakers, etc.; a storage section 1008 including a hard disk, etc.; and a communication section 1009 including a network interface card such as a LAN card, modem, etc. The communication section 1009 performs communication processing via a network such as the Internet. A drive 1010 is also connected to I / O interface 1005 as needed. A removable medium 1011, such as a disk, optical disk, magneto-optical disk, semiconductor memory, etc., is installed on drive 1010 as needed so that computer programs read from it can be installed into storage section 1008 as needed.

[0205] In particular, according to the embodiments disclosed in this invention, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments disclosed in this invention include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via communication section 1009, and / or installed from removable medium 1011. When the computer program is executed by central processing unit (CPU) 1001, it performs the functions defined above in the system of this invention.

[0206] It should be noted that the computer-readable medium shown in this invention can be a computer-readable signal medium or a computer-readable storage medium, or any combination thereof. A computer-readable storage medium can be, for example,—but not limited to—an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of a computer-readable storage medium may include, but are not limited to: an electrical connection having one or more wires, a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination thereof. In this invention, a computer-readable storage medium can be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device. In this invention, a computer-readable signal medium can include a data signal propagated in baseband or as part of a carrier wave, carrying computer-readable program code. Such propagated data signals can take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. Computer-readable signal media can also be any computer-readable medium other than computer-readable storage media, which can send, propagate, or transmit a program for use by or in connection with an instruction execution system, apparatus, or device. The program code contained on the computer-readable medium can be transmitted using any suitable medium, including but not limited to: wireless, wire, optical fiber, RF, etc., or any suitable combination thereof.

[0207] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of the present invention. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in a block diagram or flowchart, and combinations of blocks in a block diagram or flowchart, may be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0208] The modules described in the embodiments of the present invention can be implemented in software or hardware. The described modules can also be housed in a processor; for example, a processor can be described as including an analysis module, a training module, and an adjustment module. The names of these modules do not necessarily limit the module itself; for example, the adjustment module can also be described as a "retraining module."

[0209] In another aspect, the present invention also provides a computer-readable medium, which may be included in the device described in the above embodiments; or it may exist independently and not assembled into the device. The computer-readable medium carries one or more programs that, when executed by the device, cause the device to perform any of the fault template generation methods described above.

[0210] The computer program product of the present invention includes a computer program that, when executed by a processor, implements the fault template generation method in the embodiments of the present invention.

[0211] The specific embodiments described above do not constitute a limitation on the scope of protection of this invention. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can occur depending on design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this invention should be included within the scope of protection of this invention.

Claims

1. A method for generating a fault template, characterized in that, include: Obtain fault information, determine the product corresponding to the fault information, and determine the fault atom and the parameters that are compatible with the fault atom based on the fault information and the product's business information; Based on the fault information, fault atoms, and adaptation parameters, a fault template is generated, and the original fault data in the fault information is used to simulate and exercise the fault template. By comparing the simulation results with the fault results in the original fault data, the difference values ​​are obtained; In response to the difference value meeting the preset adjustment conditions, the fault atoms and adaptation parameters are readjusted, the fault template is regenerated, and a simulation exercise is performed until the difference value no longer meets the preset adjustment conditions, thus obtaining the target fault template.

2. The method according to claim 1, characterized in that, The method further includes determining a fault information group; wherein the process of determining the fault information group includes: By using the historical fault management platform, a list of resolved faults is collected, and the category of each fault information in the fault list is determined so that fault information of the same category is grouped together. Fault records are obtained from each fault information, and fault information in the same category and with the same or similar fault records are grouped together; wherein, fault records include one or more of the following: fault root cause records and fault solution records.

3. The method according to claim 1 or 2, characterized in that, The step of determining the fault atom and the parameters compatible with the fault atom based on fault information and product business information includes: Using feature analysis, business characteristics are extracted from the business information; and fault characteristics are extracted from the fault information. Based on the fault characteristics and business characteristics, and combined with the fault atom characteristics of the chaos platform, the fault atom and the parameters adapted to the fault atom are determined.

4. The method according to claim 1, characterized in that, After obtaining the target fault template, the method further includes: Determine the fault impact score based on the fault information; Determine the category to which the fault information belongs, count the number of faults of the same category occurring within a preset time period, and determine the number of business users of the product; The additional attribute score for the fault is determined based on the number of occurrences, business information, number of business users, and fault information. The score of the target fault template is determined based on the fault impact score and the fault additional attribute score; the score is used to measure the importance of the target fault template.

5. The method according to claim 4, characterized in that, The method further includes: The pre-configured weights for the fault impact score and the fault additional attribute score are added to the usage weight information of the target fault template. Receive information on weight modifications, and based on the modified weights, redetermine the score of the target fault template and the weight information to be used.

6. The method according to claim 1 or 4, characterized in that, After obtaining the target fault template, the method further includes: Using text analysis methods, business text features are extracted from the business information; Determine the category to which the fault information belongs, and use text analysis to determine the fault text features based on the category information, fault information, and fault atoms; Based on the characteristics of business text and fault text, the classification of target fault templates is determined.

7. The method according to claim 6, characterized in that, The method further includes: Based on the classification of the target fault template, query the fault module library to see if there is an existing fault template of the same classification. If the query result indicates that the existing fault template and the target fault template exist, the existing fault template and the target fault template will be integrated and the template information will be updated based on the integrated content; or the existing fault template will be replaced with the target fault template.

8. A fault template generation device, characterized in that, include: The analysis module is used to obtain fault information, determine the product corresponding to the fault information, and determine the fault atom and the parameters that are compatible with the fault atom based on the fault information and the product's business information. The exercise module is used to generate a fault template based on fault information, fault atoms, and adaptation parameters, and to conduct simulation exercises on the fault template using the original fault data in the fault information. The adjustment module is used to compare the simulation results with the fault results in the original fault data to obtain the difference value; In response to the difference value meeting the preset adjustment conditions, the fault atoms and adaptation parameters are readjusted, the fault template is regenerated, and a simulation exercise is performed until the difference value no longer meets the preset adjustment conditions, thus obtaining the target fault template.

9. An electronic device, characterized in that, include: One or more processors; Storage device for storing one or more programs. When the one or more programs are executed by the one or more processors, the one or more processors implement the method as described in any one of claims 1-7.

10. A computer-readable medium having a computer program stored thereon, characterized in that, When the program is executed by the processor, it implements the method as described in any one of claims 1-7.

11. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by a processor, it implements the method as described in any one of claims 1-7.

Images