Abnormal log processing method and device based on large language dialogue model

Abstract

The invention provides an abnormal log processing method and device based on a big language dialogue model. The method comprises the steps that after a pre-configured error code database and a code vector retrieval library are obtained, abnormal log content involved in abnormal execution of a target application service at the current moment is collected; determining model input data based on the error code database, the code vector retrieval library and the abnormal log content; and inputting the model input data into a pre-trained big language dialogue model to obtain an error code or an operation and maintenance solution corresponding to the abnormal log content output by the big language dialogue model. According to the method, errors can be quickly diagnosed through an automatic log analysis tool, intervention of developers is not needed, and the operation and maintenance efficiency and the response speed of customer problems are improved.

Description

Technical Field

[0001] The present application relates to the field of Internet technology, and in particular to an abnormal log processing method and device based on a large language dialogue model. Background Art

[0002] In the production environment of modern software systems, log analysis is a key means of problem diagnosis. Logs record various events, state changes, and error process information during system operation, which can help developers and operation and maintenance personnel to troubleshoot system anomalies. In a production environment, problems are often first discovered by the operation and maintenance team. Operation and maintenance personnel try to locate the cause of the problem by checking the system logs. At this point, they have to ask developers for help. Developers locate problems through in-depth analysis of logs combined with the specific implementation of the system, which prolongs the time to solve the problem. Summary of the invention

[0003] The purpose of the embodiments of the present application is to provide an exception log processing method and device based on a large language dialogue model, which can quickly diagnose errors through automated log parsing tools without relying on the intervention of developers, thereby improving operation and maintenance efficiency and the response speed to customer issues.

[0004] In a first aspect, a method for processing abnormal logs based on a large language dialogue model is provided, and the method may include:

[0005] Obtain a pre-configured error code database and a code vector retrieval library; the error code database includes error codes corresponding to error logs of different error types, corresponding error description information, and corresponding operation and maintenance solutions; the code vector retrieval library includes a correspondence between the code text of each log statement involved in the target application business and the corresponding code vector;

[0006] Collect the abnormal log content related to the abnormal execution of the target application business at the current moment;

[0007] Determining model input data based on the error code database, the code vector retrieval library and the abnormal log content;

[0008] The model input data is input into a pre-trained large language dialogue model to obtain an error code or operation and maintenance solution corresponding to the abnormal log content output by the large language dialogue model.

[0009] In a possible implementation, the configuration of the error code database includes:

[0010] Cluster the historical error logs generated by the target application business to obtain error log clusters of different error types;

[0011] Assign a unique error code to each error log cluster, and receive the input error description and corresponding operation and maintenance solution for any error log cluster;

[0012] An error code database is constructed based on the error codes, corresponding error description information and corresponding operation and maintenance solutions corresponding to the error log clusters of different error types.

[0013] In a possible implementation, the configuration of the code vector retrieval library includes:

[0014] Use the preset code structure parsing algorithm to cut the source code involved in the target application business and obtain the code text representing different log statements;

[0015] Perform vectorization processing on the code text of each log statement to obtain a code vector corresponding to the code text of each log statement;

[0016] Based on the code text of each log statement and the corresponding code vector, a code vector retrieval library is constructed.

[0017] In a possible implementation, determining model input data based on the error code database, the code vector retrieval library, and the abnormal log content includes:

[0018] Performing vectorization processing on the abnormal log content to obtain an abnormal log vector;

[0019] Obtaining vector distances between the log vector and each code vector in the code vector retrieval library in the vector space, and a plurality of target code vectors corresponding to the vector distances that meet a preset distance threshold;

[0020] Model input data is generated based on the code texts corresponding to the multiple target code vectors, the exception log content and the error code database.

[0021] In a possible implementation, after obtaining the error code or operation and maintenance solution corresponding to the abnormal log content output by the large language dialogue model, the method further includes:

[0022] If an error code corresponding to the abnormal log content output by the large language dialogue model is obtained, the error code is displayed to the operation and maintenance personnel, so that the operation and maintenance personnel can use the error code to find a solution in the error solution manual provided by the R&D department and provide it to the customer;

[0023] If an operation and maintenance solution corresponding to the abnormal log content output by the large language dialogue model is obtained, the operation and maintenance solution is displayed to the operation and maintenance personnel so that the operation and maintenance personnel can provide the operation and maintenance solution to the customer.

[0024] In one possible implementation, the method further includes:

[0025] Collect error logs of target application services in a preset historical time period at every preset time period; the preset historical time period is a time period with the current time as the end time;

[0026] If the error type corresponding to the collected error log does not exist in the current error code database, a specific error description and operation and maintenance solution are configured for the error type corresponding to the collected error log to update the current error code database.

[0027] In one possible implementation, the large language dialogue model is a generative language model based on a Transformer architecture.

[0028] In a second aspect, an abnormal log processing device based on a large language dialogue model is provided, and the device may include:

[0029] An acquisition unit, used to acquire a pre-configured error code database and a code vector retrieval library; the error code database includes error codes corresponding to error logs of different error types, corresponding error description information and corresponding operation and maintenance solutions, and the code vector retrieval library includes a correspondence between the code text of each log statement involved in the target application business and the corresponding code vector;

[0030] A collection unit, used to collect the abnormal log content related to the abnormal execution of the target application business at the current moment;

[0031] A determination unit, configured to determine model input data based on the error code database, the code vector retrieval library and the abnormal log content;

[0032] The input unit is used to input the model input data into the pre-trained large language dialogue model to obtain the error code or operation and maintenance solution corresponding to the abnormal log content output by the large language dialogue model.

[0033] In a third aspect, an electronic device is provided, the electronic device comprising a processor, a communication interface, a memory and a communication bus, wherein the processor, the communication interface and the memory communicate with each other via the communication bus;

[0034] Memory, used to store computer programs;

[0035] The processor is used to implement any method step described in the first aspect when executing the program stored in the memory.

[0036] In a fourth aspect, a computer-readable storage medium is provided, wherein a computer program is stored in the computer-readable storage medium, and when the computer program is executed by a processor, any method step described in the first aspect is implemented.

[0037] The exception log processing method based on the large language dialogue model provided by the present application obtains the pre-configured error code database and code vector retrieval library. The error code database includes the error codes, corresponding error description information and corresponding operation and maintenance solutions corresponding to the error logs of different error types. The code vector retrieval library includes the correspondence between the code text of each log statement involved in the target application business and the corresponding code vector, and collects the exception log content involved in the execution exception of the target application business at the current moment; based on the error code database, the code vector retrieval library and the exception log content, the model input data is determined; the model input data is input into the pre-trained large language dialogue model to obtain the error code or operation and maintenance solution corresponding to the exception log content output by the large language dialogue model. This method can quickly diagnose errors through automated log parsing tools without relying on the intervention of developers, thereby improving operation and maintenance efficiency and the response speed to customer issues. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for use in the embodiments of the present application will be briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present application and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without paying creative work.

[0039] Figure 1 A flowchart of an abnormal log processing method based on a large language dialogue model provided in an embodiment of the present application;

[0040] Figure 2 A schematic diagram of the structure of an abnormal log processing device based on a large language dialogue model provided in an embodiment of the present application;

[0041] Figure 3 A schematic diagram of the structure of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0042] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments of the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present application. Unless otherwise defined, the technical terms or scientific terms used in this application should be the common meanings understood by people with ordinary skills in the field to which the present invention belongs. The "first", "second" and similar words used in this application do not indicate any order, quantity or importance, but are only used to distinguish different components. "Including" or "including" and other similar words mean that the elements or objects appearing before the word include the elements or objects listed after the word and their equivalents, without excluding other elements or objects. "Connected", "coupled" or "connected" and other similar words are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. "Up", "down", "left", "right" and the like are only used to indicate relative positional relationships. When the absolute position of the described object changes, the relative positional relationship may also change accordingly.

[0043] As software systems become increasingly complex, log files have become an important tool for developers to debug and maintain systems. However, in production environments, log parsing and problem diagnosis still face many challenges:

[0044] 1. Disconnection between development and operation and maintenance: Although developers are able to analyze logs and find the root cause of system errors, in a production environment, customers need operation and maintenance personnel to quickly locate problems and provide solutions.

[0045] 2. Insufficient logs: After completing code development, developers are usually reluctant to add additional tracking points or refine error codes, because the newly added tracking logic may cause unnecessary bugs and affect the normal operation of the system. However, this will prevent the operation and maintenance team from quickly obtaining effective error information through logs when problems occur.

[0046] 3. Delayed operation and maintenance response: Operation and maintenance personnel must rely on developers’ log analysis, which prolongs the problem-solving time and affects customer experience and business continuity.

[0047] Therefore, the existing log parsing process, especially in emergency handling in a production environment, has obvious limitations. In order to solve the above problems, this application provides an exception log processing method based on a large language dialogue model, which can help operation and maintenance personnel automatically analyze logs and provide timely error diagnosis without the participation of developers.

[0048] The exception log processing method based on the large language dialogue model provided in the embodiment of the present application adopts the retrieval-augmented generation (RAG) technology, which combines the large language dialogue model with the plug-in code knowledge base. When an error occurs during system operation, the error log is collected in real time, and the context information of the log is input into the large language dialogue model. The large language dialogue model locates the code position corresponding to the log by analyzing the log content, and returns the correct error code or solution based on the code analysis. The large language dialogue model is a dialogue model trained on the basis of the large language model.

[0049] Among them, the large language model is a generative language model based on the Transformer architecture. It can be GPT, or it can be an open source llama, GLM, Qwen and other models. The training cost of the large language model is particularly high, so for vertical field users, the general method is to first select a base model (the above-mentioned GPT, GLM, etc.), use SFT (Supervised Fine-Tuning) to train the LoRA (Low-Rank Adaptation) plug-in on small-scale data, or the non-training method of the RAG (Retrieval-Augmented Generation) plug-in knowledge base used in this article.

[0050] The preferred embodiments of the present application are described below in conjunction with the drawings in the specification. It should be understood that the preferred embodiments described herein are only used to illustrate and explain the present application, and are not used to limit the present application. In addition, the embodiments and features in the embodiments of the present application may be combined with each other if there is no conflict.

[0051] Figure 1 The flowchart of an abnormal log processing method based on a large language dialogue model provided in an embodiment of the present application is as follows. Figure 1 As shown, the method may include:

[0052] Step S110: Obtain a pre-configured error code database and code vector retrieval library.

[0053] (1) Configuration of the error code database may include:

[0054] Cluster the historical error logs generated by the target application business to obtain error log clusters of different error types;

[0055] Assign a unique error code to each error log cluster, and receive the input error description and corresponding operation and maintenance solution for any error log cluster;

[0056] Based on the error codes, corresponding error description information and corresponding operation and maintenance solutions corresponding to error log clusters of different error types, an error code database is constructed, that is, the error code database may include error codes, corresponding error description information and corresponding operation and maintenance solutions corresponding to different types of error logs.

[0057] (2) The configuration of the code vector retrieval library may include:

[0058] Since a single paragraph may not be able to fully express the required information semantically, the most commonly used method is fixed-length segmentation with overlap, while limiting the destruction of the entire sentence or paragraph. For code segmentation, a preset code structure parsing algorithm can be used. Specifically, the preset code structure parsing algorithm is used to segment the source code involved in the target application business to obtain code texts representing different log statements;

[0059] The code text of each log statement is vectorized (for example, the code text is vectorized using an embedding model) to obtain the code vector corresponding to the code text of each log statement; the embedding model can be text-embedding-3 of OpenAI or BGE-Embedding of Wenxin Qianfan. Regardless of which model is selected, the purpose is to vectorize the text.

[0060] Based on the code text of each log statement and the corresponding code vector, a code vector retrieval library is constructed, that is, the code vector retrieval library may include the correspondence between the code text of each log statement involved in the target application business and the corresponding code vector.

[0061] Step S120: Collect the abnormal log content related to the abnormal execution of the target application service at the current moment.

[0062] Step S130: Determine model input data based on the error code database, the code vector retrieval library and the abnormal log content.

[0063] In the specific implementation, the exception log content is vectorized to obtain the exception log vector; the vector distances between the log vector and each code vector in the code vector retrieval library in the vector space, as well as multiple target code vectors corresponding to the vector distances that meet the preset distance threshold are obtained; then, based on the code texts corresponding to the multiple target code vectors, the exception log content and the error code database, the model input data prompt is generated.

[0064] Specifically, the log text is vectorized using the same model, and then the distance between the log vector and the code vector in the vector space is calculated (different embedding models support different distance calculation methods, such as Euclidean distance and cosine distance). Based on the vector distance, the code text corresponding to the top-N most relevant code vectors can be screened out.

[0065] In an example, "You are an operation and maintenance robot. The known error types and corresponding error codes are as follows {error code-key-value pair describing the error type}. The code implementation snippets are as follows: {code snippet 1},..., {code snippet N}. Now I have obtained an error log as follows {log text}. Please give the most appropriate error code. If you cannot find a particularly appropriate one, please answer that you cannot determine."

[0066] Step S140: input the model input data into the pre-trained large language dialogue model to obtain an error code or operation and maintenance solution corresponding to the abnormal log content output by the large language dialogue model.

[0067] If an error code corresponding to the abnormal log content output by the large language dialogue model is obtained, the error code is displayed to the operation and maintenance personnel so that the operation and maintenance personnel can use the error code to find a solution in the error resolution manual provided by the R&D department and provide it to the customer;

[0068] If an operation and maintenance solution corresponding to the abnormal log content output by the large language dialogue model is obtained, the operation and maintenance solution is displayed to the operation and maintenance personnel so that the operation and maintenance personnel can provide the operation and maintenance solution to the customer.

[0069] In some embodiments, the error log of the target application service in a preset historical time period is collected every preset time period; the preset historical time period is a time period with the current time as the end time;

[0070] If the error type corresponding to the collected error log does not exist in the current error code database, a specific error description and operation and maintenance solution are configured for the error type corresponding to the collected error log to update the current error code database.

[0071] The technical effects of this application include:

[0072] 1. Application of RAG technology: Through RAG technology, the generation capability of large models is combined with the retrieval capability of the plug-in code knowledge base, which greatly improves the accuracy of log analysis and problem location.

[0073] 2. Automatic matching of error codes: By automatically analyzing logs and code logic, the tool can automatically identify error codes without the developer having to embed additional log points, thus avoiding the limitations of relying on manual rules.

[0074] 3. Separation of development and operation and maintenance: The present invention allows the operation and maintenance team to independently obtain accurate error codes and solutions without the participation of developers, significantly improving operation and maintenance efficiency.

[0075] In some embodiments, if it is detected that within a time period of the target business application, the error codes or operation and maintenance solutions corresponding to multiple abnormal log contents obtained are the same, it indicates that multiple identical abnormal log contents are generated during this time period. At this time, the corresponding operation and maintenance personnel solutions in the error code database are updated, or an abnormal alarm is sent to the operation and maintenance personnel so that the operation and maintenance personnel can modify the source code corresponding to the abnormal log content.

[0076] Corresponding to the above method, the embodiment of the present application also provides an abnormal log processing device based on a large language dialogue model, such as Figure 2 As shown, the device comprises:

[0077] The acquisition unit 210 is used to acquire a pre-configured error code database and a code vector retrieval library; the error code database includes error codes corresponding to error logs of different error types, corresponding error description information and corresponding operation and maintenance solutions, and the code vector retrieval library includes a correspondence between the code text of each log statement involved in the target application business and the corresponding code vector;

[0078] The collection unit 220 is used to collect the abnormal log content related to the abnormal execution of the target application service at the current moment;

[0079] A determination unit 230, configured to determine model input data based on the error code database, the code vector retrieval library and the abnormal log content;

[0080] The input unit 240 is used to input the model input data into the pre-trained large language dialogue model to obtain the error code or operation and maintenance solution corresponding to the abnormal log content output by the large language dialogue model.

[0081] The functions of each functional unit of the abnormal log processing device based on the large language dialogue model provided in the above-mentioned embodiment of the present application can be realized through the above-mentioned method steps. Therefore, the specific working process and beneficial effects of each unit in the abnormal log processing device based on the large language dialogue model provided in the embodiment of the present application are not repeated here.

[0082] The present application also provides an electronic device, such as Figure 3 As shown, it includes a processor 310 , a communication interface 320 , a memory 330 and a communication bus 340 , wherein the processor 310 , the communication interface 320 , and the memory 330 communicate with each other via the communication bus 340 .

[0083] Memory 330, for storing computer programs;

[0084] The processor 310 is used to execute the program stored in the memory 330 to implement the following steps:

[0085] Obtain a pre-configured error code database and a code vector retrieval library; the error code database includes error codes corresponding to error logs of different error types, corresponding error description information, and corresponding operation and maintenance solutions; the code vector retrieval library includes a correspondence between the code text of each log statement involved in the target application business and the corresponding code vector;

[0086] Collect the abnormal log content related to the abnormal execution of the target application business at the current moment;

[0087] Determining model input data based on the error code database, the code vector retrieval library and the abnormal log content;

[0088] The model input data is input into a pre-trained large language dialogue model to obtain an error code or operation and maintenance solution corresponding to the abnormal log content output by the large language dialogue model.

[0089] The communication bus mentioned above can be a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus, etc. The communication bus can be divided into an address bus, a data bus, a control bus, etc. For ease of representation, only one thick line is used in the figure, but it does not mean that there is only one bus or one type of bus.

[0090] The communication interface is used for communication between the above electronic device and other devices.

[0091] The memory may include a random access memory (RAM) or a non-volatile memory (NVM), such as at least one disk memory. Optionally, the memory may also be at least one storage device located away from the aforementioned processor.

[0092] The above-mentioned processor can be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it can also be a digital signal processor (DSP), an application specific integrated circuit (ASIC), a field programmable gate array (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components.

[0093] The implementation methods and beneficial effects of the components of the electronic device in the above embodiments to solve the problems can be found in Figure 1 The various steps in the illustrated embodiment are implemented, therefore, the specific working process and beneficial effects of the electronic device provided by the embodiment of the present application are not repeated here.

[0094] In another embodiment provided in the present application, a computer-readable storage medium is also provided, in which instructions are stored. When the computer-readable storage medium is executed on a computer, the computer executes the exception log processing method based on the large language dialogue model described in any of the above embodiments.

[0095] In another embodiment provided by the present application, a computer program product including instructions is also provided. When the computer program product is executed on a computer, the computer executes the abnormal log processing method based on the large language dialogue model described in any of the above embodiments.

[0096] Those skilled in the art will appreciate that the embodiments in the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt a complete hardware embodiment, a complete software embodiment, or a form of an embodiment combining software and hardware. Moreover, the present application may adopt a form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0097] The embodiments of the present application are described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0098] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0099] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0100] Although the preferred embodiments of the present application have been described, those skilled in the art may make additional changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications falling within the scope of the embodiments of the present application.

[0101] Obviously, those skilled in the art can make various changes and modifications to the embodiments in the present application without departing from the spirit and scope of the embodiments in the present application. Thus, if these modifications and variations of the embodiments in the present application fall within the scope of the claims and their equivalents in the embodiments of the present application, the embodiments of the present application are also intended to include these modifications and variations.

Claims

1. A method for processing abnormal logs based on a large language dialogue model, characterized in that: The method comprises: Obtain a pre-configured error code database and a code vector retrieval library; the error code database includes error codes, corresponding error description information and corresponding operation and maintenance solutions corresponding to error logs of different error types, and the code vector retrieval library includes the correspondence between the code text of each log statement involved in the target application business and the corresponding code vector; Collect the abnormal log content related to the abnormal execution of the target application business at the current moment; Determining model input data based on the error code database, the code vector retrieval library and the abnormal log content; The model input data is input into a pre-trained large language dialogue model to obtain an error code or operation and maintenance solution corresponding to the abnormal log content output by the large language dialogue model.

2. The method according to claim 1, characterized in that The configuration of the error code database includes: Cluster the historical error logs generated by the target application business to obtain error log clusters of different error types; Assign a unique error code to each error log cluster, and receive the input error description and corresponding operation and maintenance solution for any error log cluster; An error code database is constructed based on the error codes, corresponding error description information and corresponding operation and maintenance solutions corresponding to the error log clusters of different error types.

3. The method according to claim 2, characterized in that The configuration of the code vector retrieval library includes: Use the preset code structure parsing algorithm to cut the source code involved in the target application business and obtain the code text representing different log statements; Perform vectorization processing on the code text of each log statement to obtain a code vector corresponding to the code text of each log statement; Based on the code text of each log statement and the corresponding code vector, a code vector retrieval library is constructed.

4. The method according to claim 3, characterized in that Determining model input data based on the error code database, the code vector retrieval library and the abnormal log content includes: Performing vectorization processing on the abnormal log content to obtain an abnormal log vector; Obtaining vector distances between the log vector and each code vector in the code vector retrieval library in the vector space, and a plurality of target code vectors corresponding to the vector distances that meet a preset distance threshold; Model input data is generated based on the code texts corresponding to the multiple target code vectors, the exception log content and the error code database.

5. The method according to claim 1, characterized in that After obtaining the error code or operation and maintenance solution corresponding to the abnormal log content output by the large language dialogue model, the method further includes: If an error code corresponding to the abnormal log content output by the large language dialogue model is obtained, the error code is displayed to the operation and maintenance personnel, so that the operation and maintenance personnel can use the error code to find a solution in the error solution manual provided by the R&D department and provide it to the customer; If an operation and maintenance solution corresponding to the abnormal log content output by the large language dialogue model is obtained, the operation and maintenance solution is displayed to the operation and maintenance personnel so that the operation and maintenance personnel can provide the operation and maintenance solution to the customer.

6. The method according to claim 1, characterized in that The method further comprises: Collect error logs of target application services in a preset historical time period at every preset time period; the preset historical time period is a time period with the current time as the end time; If the error type corresponding to the collected error log does not exist in the current error code database, a specific error description and operation and maintenance solution are configured for the error type corresponding to the collected error log to update the current error code database.

7. The method according to claim 1, characterized in that The large language dialogue model is a generative language model based on the Transformer architecture.

8. An abnormal log processing device based on a large language dialogue model, characterized in that: The device comprises: An acquisition unit, used to acquire a pre-configured error code database and a code vector retrieval library; the error code database includes error codes corresponding to error logs of different error types, corresponding error description information and corresponding operation and maintenance solutions, and the code vector retrieval library includes a correspondence between the code text of each log statement involved in the target application business and the corresponding code vector; A collection unit, used to collect the abnormal log content related to the abnormal execution of the target application business at the current moment; A determination unit, configured to determine model input data based on the error code database, the code vector retrieval library and the abnormal log content; The input unit is used to input the model input data into the pre-trained large language dialogue model to obtain the error code or operation and maintenance solution corresponding to the abnormal log content output by the large language dialogue model.

9. An electronic device, characterized in that: The electronic device comprises a processor, a communication interface, a memory and a communication bus, wherein the processor, the communication interface and the memory communicate with each other via the communication bus; Memory, used to store computer programs; A processor, for implementing any of the methods described in claims 1-7 when executing a program stored in a memory.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the method according to any one of claims 1 to 7 is implemented.

Images