Fault root cause determination method and device, electronic equipment and storage medium

Abstract

The invention provides a fault root cause determination method and device, electronic equipment and a storage medium, relates to the field of artificial intelligence, in particular to the technical fields of NLP, large language models, deep learning and the like, and can be applied to the application fields of generative search, document intelligent editing, intelligent assistants, intelligent e-commerce and the like. According to the specific implementation scheme, the method comprises the steps of obtaining a fault problem input in response to a fault of the distributed service system; classifying the fault problems to obtain target categories to which the fault problems belong; obtaining a target interaction topological graph matched with the target category; wherein nodes in the target interaction topological graph are used for indicating intelligent agents, edges between the nodes are used for indicating interaction relations between the intelligent agents, and the intelligent agents comprise tool type intelligent agents and / or non-tool type intelligent agents bound with tools; and sequentially calling the intelligent agents indicated by the nodes in the target interaction topological graph to process the fault problem so as to obtain the fault root cause of the distributed service system.

Description

Technical Field

[0001] The present disclosure relates to the field of AI (Artificial Intelligence), specifically to technical fields such as NLP (Natural Language Processing), LLM (Large Language Model), and deep learning, and can be used in application fields such as generative search, intelligent document editing, intelligent assistants, and intelligent e-commerce, and in particular to methods, devices, electronic devices, and storage media for determining root causes of faults. Background Art

[0002] For large-scale distributed business systems of enterprises, different business teams maintain their various components. Distributed business systems may face various types of stability failures in daily operation. In order to solve these stability failures, it is often necessary to analyze the root causes of the failures and then solve the stability failures based on the root causes of the failures. Therefore, it is very important to determine the root causes of the failures of distributed business systems. Summary of the invention

[0003] The present disclosure provides a method, device, electronic device and storage medium for determining a root cause of a fault.

[0004] According to one aspect of the present disclosure, a method for determining a root cause of a fault is provided, comprising:

[0005] Obtaining a fault problem input in response to a fault occurring in a distributed business system;

[0006] Classifying the fault problem to obtain a target category to which the fault problem belongs;

[0007] Acquire a target interaction topology map adapted to the target category; wherein the nodes in the target interaction topology map are used to indicate agents, the edges between the nodes are used to indicate the interaction relationship between the agents, and the agents include tool-type agents and / or non-tool-type agents bound to tools;

[0008] The intelligent agents indicated by the nodes in the target interaction topology diagram are called in sequence to process the fault problem, so as to obtain the root cause of the fault of the distributed business system.

[0009] According to another aspect of the present disclosure, a device for determining a root cause of a fault is provided, comprising:

[0010] A first acquisition module is used to acquire a fault problem input in response to a fault occurring in the distributed business system;

[0011] A classification module, used to classify the fault problem to obtain a target category to which the fault problem belongs;

[0012] A second acquisition module is used to acquire a target interaction topology map adapted to the target category; wherein the nodes in the target interaction topology map are used to indicate agents, and the edges between the nodes are used to indicate the interaction relationship between the agents, and the agents include tool-type agents and / or non-tool-type agents bound to tools;

[0013] The processing module is used to call the intelligent agent indicated by each node in the target interaction topology diagram in turn to process the fault problem so as to obtain the root cause of the fault of the distributed business system.

[0014] According to another aspect of the present disclosure, there is provided an electronic device, comprising:

[0015] at least one processor; and

[0016] a memory communicatively connected to the at least one processor; wherein,

[0017] The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the method for determining the root cause of the fault proposed in the above aspect of the present disclosure.

[0018] According to another aspect of the present disclosure, a non-transitory computer-readable storage medium of computer instructions is provided, wherein the computer instructions are used to enable the computer to execute the method for determining a root cause of a fault proposed in the above aspect of the present disclosure.

[0019] According to another aspect of the present disclosure, a computer program product is provided, including a computer program, and when the computer program is executed by a processor, the method for determining a root cause of a fault proposed in the above aspect of the present disclosure is implemented.

[0020] It should be understood that the content described in this section is not intended to identify the key or important features of the embodiments of the present disclosure, nor is it intended to limit the scope of the present disclosure. Other features of the present disclosure will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] The accompanying drawings are used to better understand the present solution and do not constitute a limitation of the present disclosure.

[0022] Figure 1 This is a diagram showing the correspondence between existing stability failures and business teams;

[0023] Figure 2 It is a schematic diagram of the correspondence between existing stability faults and tools;

[0024] Figure 3 A flowchart of a method for determining a root cause of a fault provided in the first embodiment of the present disclosure;

[0025] Figure 4 A flowchart of a method for determining a root cause of a fault provided in Embodiment 2 of the present disclosure;

[0026] Figure 5 A flowchart of a method for determining a root cause of a fault provided in Embodiment 3 of the present disclosure;

[0027] Figure 6 A flowchart of a method for determining a root cause of a fault provided in Embodiment 4 of the present disclosure;

[0028] Figure 7 A flowchart of a method for determining a root cause of a fault provided in Embodiment 5 of the present disclosure;

[0029] Figure 8 A flowchart of a method for determining a root cause of a fault provided in Embodiment 6 of the present disclosure;

[0030] Fig. 9 A schematic diagram of the structure of the low-code tool platform provided in the embodiment of the present disclosure;

[0031] Fig.10 A schematic diagram of the interaction process between a tool-type intelligent agent and a tool provided in an embodiment of the present disclosure;

[0032] Fig.11 A schematic diagram of the interaction relationship between intelligent agents provided in an embodiment of the present disclosure;

[0033] Fig.12 A schematic diagram of the interaction relationship between atomic tools provided in the embodiment of the present disclosure;

[0034] Fig.13 A schematic diagram of a combination method of atomic tools provided in an embodiment of the present disclosure;

[0035] Fig.14 A schematic diagram of the structure of a device for determining a root cause of a fault provided in Embodiment 7 of the present disclosure;

[0036] Fig.15 A schematic block diagram of an example electronic device that can be used to implement embodiments of the present disclosure is shown. DETAILED DESCRIPTION

[0037] The following is a description of exemplary embodiments of the present disclosure in conjunction with the accompanying drawings, including various details of the embodiments of the present disclosure to facilitate understanding, which should be considered as merely exemplary. Therefore, it should be recognized by those of ordinary skill in the art that various changes and modifications may be made to the embodiments described herein without departing from the scope and spirit of the present disclosure. Similarly, for the sake of clarity and conciseness, descriptions of well-known functions and structures are omitted in the following description.

[0038] For stability failures in distributed business systems, it is usually necessary for multiple business teams to work together, and in the process of solving different stability failures, business teams need to use different tools to solve stability failures. Among them, the efficiency of collaboration between multiple business teams and the completeness of tools directly determine the speed at which stability failures are solved (i.e. MTTR (Mean-Time-To-Resolution, the average time from the occurrence of a failure to its complete repair and restoration of service)), which in turn affects the availability of the company's distributed business systems and is related to the company's revenue and reputation.

[0039] As an example, the relationship between stability failures and business teams can be as follows: Figure 1 As shown, for example, stability failure 2 (referred to as failure 2 for short) can be coordinated and resolved by business team 1, business team 4, and business team 7.

[0040] In this case, the fault is relatively complex, and a stability fault requires different business teams to work together to analyze it. First, there must be a top-level organizer to determine the fault form, and then organize the business team leaders who need to participate in solving the stability fault to work together to complete the analysis and solve the stability fault. This process requires a high level of experience for the top-level organizers and business team leaders, and organizing and hosting the solution of stability faults requires a lot of communication costs, all of which affect the overall fault solving efficiency.

[0041] As an example, the relationship between stability failures and tools can be as follows Figure 2 As shown, for example, stability fault 2 (referred to as fault 2) requires calling tool 1, tool 3 and tool 6 in combination to solve it.

[0042] In summary, different types of stability failures require different combinations of tools to solve them. And different types of stability failures may require the use of the same tools, which are often script commands that need to be written temporarily. In this case, on the one hand, when solving different types of stability failures, the tools written temporarily are repeated, wasting manpower costs; on the other hand, because the tools need to be written temporarily, it takes a certain amount of time, which affects the efficiency of troubleshooting.

[0043] Currently, large models are booming, especially large language models (LLMs), which have shown strong capabilities. In terms of stability fault resolution of distributed business systems, the industry also has many intelligent solutions combined with LLMs, and has built intelligent resolution systems for stability faults. However, none of these solutions and intelligent resolution systems have the characteristics of "stability", "maintainability" and "low cost" at the same time.

[0044] Among them, stability: In the process of solving stability problems, each step needs to produce stable and reliable analysis results. However, a typical defect of LLM is the existence of "illusion", which brings challenges to the stability of analysis results.

[0045] Maintainability: When solving a stability failure requires the participation of different business parties (i.e. business teams), the business parties have the core experience of their business. If these experiences cannot be accumulated by the business parties themselves, then even if the overall intelligent solution system is well implemented, it is not maintainable. This means that when the business system changes, the business parties cannot update the knowledge or experience in the intelligent solution system in a timely manner, making the intelligent solution system no longer effective.

[0046] Low cost: When a new fault type appears, it is only necessary to build new tools or new intelligent agents in a low-cost manner, and no changes can be made to the original tools or intelligent agents. In other words, the new fault type must make the intelligent solution system stronger without any operation on the solution capabilities of the existing fault type.

[0047] In related technologies, the following solutions are mainly used to solve the failure problem of distributed business systems:

[0048] The first one is to use an intelligent agent platform to solve the failure problems of distributed business systems.

[0049] However, the intelligent agent platform is responded by a single intelligent agent, and the fault problem is only distributed to a single intelligent agent, which cannot solve the fault problems in complex scenarios.

[0050] The second method is to use multi-agent intelligent operation and maintenance to solve the failure problems of distributed business systems.

[0051] Although this approach uses the interaction mode between multiple intelligent agents to solve fault problems, when encountering new forms of fault problems, the solution logic of the old forms of fault problems may need to be modified, that is, it does not meet the maintainability requirements. At the same time, it does not mention how to enable different business parties to participate in a friendly manner, which does not reflect business friendliness.

[0052] The third method is to use multi-scenario intelligent operation and maintenance to implement an interactive mode between intelligent agents to solve fault problems.

[0053] However, this method uses function_call when calling the tool, and uses the "reflection" mechanism to solve the instability caused by the illusion of the model. At the same time, it does not mention how to make different business parties participate in a friendly manner, which does not reflect business friendliness.

[0054] Therefore, in response to at least one of the above-mentioned problems, the present disclosure proposes a method, device, electronic device and storage medium for determining the root cause of a fault to solve various stability failures while meeting the above-mentioned requirements of "stability", "maintainability" and "low cost".

[0055] The following describes the method, device, electronic device and storage medium for determining the root cause of a fault according to an embodiment of the present disclosure with reference to the accompanying drawings. Before describing the embodiment of the present disclosure in detail, for ease of understanding, common technical terms are first introduced:

[0056] A large model refers to a machine learning model with large-scale parameters and complex computational structures. It is usually built from deep neural networks and has billions or even hundreds of billions of parameters. The purpose of designing a large model is to improve the model's expressiveness and predictive performance, and to be able to handle more complex tasks and data. Large models are widely used in various fields, including natural language processing, computer vision, speech recognition, and recommendation systems.

[0057] The LLM in the big model is a type of natural language processing model based on deep learning. Its main features are huge model parameters and complex neural network structure, strong language understanding, context perception and language generation capabilities. It can automatically learn useful feature representations from input data and generate relevant text.

[0058] Figure 3 A flowchart of a method for determining a root cause of a fault provided in Embodiment 1 of the present disclosure.

[0059] The embodiment of the present disclosure takes the method for determining the root cause of a fault as an example in which the method is configured in a device for determining the root cause of a fault. The device for determining the root cause of a fault can be applied to any electronic device so that the electronic device can perform the function of determining the root cause of a fault.

[0060] Among them, the electronic device can be any device with computing capabilities, such as a personal computer, a mobile terminal, a server, etc. The mobile terminal can be, for example, a mobile phone, a tablet computer, a personal digital assistant, a wearable device, and other hardware devices with various operating systems, touch screens and / or display screens.

[0061] like Figure 3As shown, the method for determining the root cause of the fault may include the following steps S301 to S304:

[0062] Step S301: Obtain a fault problem input in response to a fault occurring in a distributed business system.

[0063] The fault problem may be a problem manually input by relevant personnel (such as users, business personnel, etc.).

[0064] Among them, the input method of the fault problem may include but is not limited to touch input (such as sliding, clicking, etc.), keyboard input, voice input, etc.

[0065] Step S302: classify the fault problem to obtain the target category to which the fault problem belongs.

[0066] Among them, the target categories include but are not limited to: query lossy problems (such as no search results for the user's input query, irrelevant search results, duplicate results (that is, there are a large number of duplicate entries in the search results)), long-tail problems, Core problems (core system problems), and other problem categories.

[0067] In the embodiment of the present disclosure, the fault problem may be classified based on text classification technology to obtain the problem category to which the fault problem belongs (referred to as the target category in the present disclosure).

[0068] Step S303, obtaining a target interaction topology map adapted to the target category; wherein the nodes in the target interaction topology map are used to indicate agents, and the edges between the nodes are used to indicate the interaction relationship between agents, and the agents include tool-type agents and / or non-tool-type agents bound to tools.

[0069] The tools include but are not limited to script commands, API (Application Programming Interface), functions, etc.

[0070] In an embodiment of the present disclosure, an agent interaction topology map adapted to the target category (referred to as a target interaction topology map in the present disclosure) can be obtained; wherein the nodes in the target interaction topology map are used to indicate agents, and the edges between the nodes are used to indicate the interaction relationship between agents, wherein the agents include tool-type agents and / or non-tool-type agents bound to tools.

[0071] It should be noted that for different problem categories, the adapted intelligent agent interaction topology diagram may be different to improve the pertinence and accuracy of handling different fault problems.

[0072] Step S304, calling the intelligent agent indicated by each node in the target interaction topology diagram in turn to process the fault problem, so as to obtain the root cause of the fault of the distributed business system.

[0073] The root cause of a fault refers to the fundamental or underlying cause of a fault in a distributed business system.

[0074] In the disclosed embodiment, the intelligent agents indicated by the nodes in the target interaction topology diagram may be called in sequence to process the fault problem, so as to obtain the root cause of the fault of the distributed business system.

[0075] The method for determining the root cause of a fault in the disclosed embodiment, when a fault occurs in a distributed business system, determines an agent interaction topology diagram for handling the fault problem based on the fault category to which the fault problem belongs, and collaboratively handles the fault problem based on the agents indicated by each node in the agent interaction topology diagram to obtain the root cause of the fault of the distributed business system, which can not only solve the fault problem in complex scenarios, but also improve the accuracy of determining the root cause of the fault. In addition, the agent can be a tool-type agent, which can improve the stability of the agent execution, thereby improving the accuracy of determining the root cause of the fault.

[0076] It should be noted that in the technical solution of the present disclosure, the collection, storage, use, processing, transmission, provision and disclosure of user personal information are all carried out with the user's consent, and are in compliance with relevant laws and regulations and do not violate public order and good morals.

[0077] In order to clearly explain how to determine the root cause of a fault in a distributed business system in any embodiment of the present disclosure, the present disclosure also proposes a method for determining the root cause of a fault.

[0078] Figure 4 This is a flowchart of a method for determining a root cause of a fault provided in Embodiment 2 of the present disclosure.

[0079] like Figure 4 As shown, the method for determining the root cause of the fault may include the following steps S401 to S405:

[0080] Step S401: obtaining a fault problem input in response to a fault occurring in a distributed business system.

[0081] Step S402: classify the fault problem to obtain the target category to which the fault problem belongs.

[0082] Step S403, obtaining a target interaction topology map adapted to the target category; wherein the nodes in the target interaction topology map are used to indicate agents, and the edges between the nodes are used to indicate the interaction relationship between agents, and the agents include tool-type agents and / or non-tool-type agents bound to tools.

[0083] For explanations of steps S401 to S403, reference may be made to the relevant descriptions in any embodiment of the present disclosure, and will not be repeated here.

[0084] Step S404, calling the intelligent agent indicated by each node in the target interaction topology diagram in turn to handle the fault problem.

[0085] In the disclosed embodiment, the intelligent agent indicated by each node in the target interaction topology diagram may be called in sequence to process the fault problem, and the execution result of the intelligent agent indicated by each node may be obtained.

[0086] In any embodiment of the present disclosure, the agents indicated by the nodes in the target interactive topology diagram can be called in sequence. When the currently called agent is the agent indicated by the root node in the target interactive topology diagram, it can be determined whether the agent indicated by the root node is a tool-type agent. If so, the fault problem can be processed by the agent indicated by the root node (i.e., the tool-type agent) and the tool bound to the agent indicated by the root node to obtain the processing result of the agent indicated by the root node; if not, the fault problem can be directly processed by the agent indicated by the root node (i.e., the non-tool-type agent) to obtain the processing result of the agent indicated by the root node. Therefore, different processing strategies are used for tool-type agents and non-tool-type agents to process fault problems, which can improve the pertinence, rationality and accuracy of agent processing.

[0087] When the currently called agent is the agent indicated by the non-root node in the target interaction topology diagram, it can be determined whether the agent indicated by the non-root node is a tool-type agent. If so, the processing result of the agent indicated by the parent node of the non-root node is processed through the agent indicated by the non-root node (i.e., the tool-type agent) and the tool bound to the agent indicated by the non-root node to obtain the processing result of the agent indicated by the non-root node; if not, the processing result of the agent indicated by the parent node of the non-root node is directly processed through the agent indicated by the non-root node (i.e., the non-tool-type agent) to obtain the processing result of the agent indicated by the non-root node.

[0088] In summary, the processing results of the agent indicated by the child node in the agent interaction topology diagram are dependent on the processing results of the agent indicated by its parent node. Through clear dependencies or interactions, the logic of the processing flow can be ensured to be clear. Each agent works only after its parent node has completed processing, which reduces error processing caused by sequence errors or data inconsistencies, allowing each agent to operate in the correct context, thereby improving the reliability and accuracy of fault problem processing. In addition, the dependencies between agents can be flexibly configured, which makes it easier for the system to add new agents or modify the dependencies between existing agents when facing new fault types or processing requirements, thereby improving the flexibility and scalability of the system.

[0089] Step S405, determining the root cause of the failure of the distributed business system according to the processing result of the intelligent agent indicated by the leaf node in the target interaction topology diagram.

[0090] In the disclosed embodiment, the root cause of the failure of the distributed business system can be determined by comprehensively analyzing the processing results of the intelligent agents indicated by the leaf nodes in the target interaction topology diagram.

[0091] The method for determining the root cause of a fault in the disclosed embodiment is that since the leaf node is located at the very end of the agent interaction topology diagram, its processing result is often based on the accumulated information and processing logic of all previous nodes (including its parent node), and has high accuracy and reliability. Therefore, the root cause of the fault of the distributed business system is determined according to the processing result of the agent indicated by the leaf node, which can improve the accuracy and reliability of determining the root cause of the fault. In addition, in complex systems, faults are often accompanied by a variety of appearances and interference factors. Determining the root cause of the fault according to the processing result of the agent indicated by the leaf node can avoid being misled by these appearances and interference factors, thereby further improving the accuracy and reliability of determining the root cause of the fault.

[0092] In order to clearly explain how the above embodiments sequentially call the intelligent agents indicated by the nodes in the target interaction topology diagram to handle the fault problem, the present disclosure also proposes a method for determining the root cause of the fault.

[0093] Figure 5 This is a flowchart of a method for determining a root cause of a fault provided in Embodiment 3 of the present disclosure.

[0094] like Figure 5 As shown, the method for determining the root cause of the fault may include the following steps S501 to S511:

[0095] Step S501: obtaining a fault problem input in response to a fault occurring in a distributed business system.

[0096] Step S502: classify the fault problem to obtain the target category to which the fault problem belongs.

[0097] Step S503: Acquire a target interaction topology map adapted to the target category.

[0098] Among them, the nodes in the target interaction topology diagram are used to indicate intelligent agents, and the edges between the nodes are used to indicate the interaction relationship between the intelligent agents. The intelligent agents include tool-type intelligent agents and / or non-tool-type intelligent agents bound to tools.

[0099] Step S504, calling the intelligent agent indicated by each node in the target interaction topology diagram in sequence.

[0100] For explanations of steps S501 to S504, reference may be made to the relevant descriptions in any embodiment of the present disclosure, and will not be repeated here.

[0101] Step S505, in response to the currently called agent being the agent indicated by the root node in the target interaction topology diagram, determine whether the agent indicated by the root node is a tool-type agent, if so, execute step S506, if not, execute step S507.

[0102] It should be noted that step S506 and step S507 are two parallel implementation methods. In actual application, only one needs to be executed.

[0103] Step S506, processing the fault problem through the agent indicated by the root node and the tool bound to the agent indicated by the root node to obtain the processing result of the agent indicated by the root node.

[0104] In an embodiment of the present disclosure, when the currently called agent is the agent indicated by the root node in the target interaction topology diagram, and the agent indicated by the root node is a tool-type agent, the fault problem can be collaboratively processed through the agent indicated by the root node and the tool bound to the agent indicated by the root node to obtain the processing result of the agent indicated by the root node.

[0105] In any embodiment of the present disclosure, the following steps A to E may be used to process the fault problem and obtain the processing result of the agent indicated by the root node:

[0106] Step A: Get the input parameter description and function description associated with the tool bound to the agent indicated by the root node.

[0107] The input parameter description is used to describe the input parameter description of the tool. For example, the input parameter description of the tool can be described in markdown format. For example, the input parameter description of the tool can be marked as input_param_description.

[0108] The function description is used to describe the function of the tool. For example, the function description of the tool can be marked as description.

[0109] As an example, different interfaces may be constructed in the tool, and the input parameter description and function description of the tool may be obtained by calling the above interfaces.

[0110] For example, the following interface can be constructed in the tool:

[0111] Interface name: name

[0112] Input: None.

[0113] Output: Tool name.

[0114] Interface name: description / / Function description of the tool

[0115] Input: None.

[0116] Output: Description of the tool's functionality.

[0117] Interface name: input_param_description / / tool input parameter description

[0118] Input: None.

[0119] Output: Description of tool input parameters in markdown format.

[0120] Interface name: run_with_params / / Tool execution results

[0121] Input: parameter list in json format.

[0122] Output: tool execution results.

[0123] Step B: Determine tool call parameters based on the fault problem, input parameter description and function description.

[0124] In the embodiment of the present disclosure, the fault problem, input parameter description and function description can be comprehensively considered to determine the calling parameters of the tool bound to the agent indicated by the root node, which are referred to as tool calling parameters in the present disclosure.

[0125] As a possible implementation method, a large language model can be used to process the fault problem, input parameter description and function description to obtain tool calling parameters.

[0126] As an example, first, at least one first reference example associated with the agent indicated by the root node can be obtained or queried, wherein the first reference example is used to indicate a conversion method for converting natural language information into input parameters (or parameter list) of a tool bound to the agent indicated by the root node. Thereafter, a first prompt message (Prompt) can be generated based on each first reference example, input parameter description, function description, and fault problem. For example, each first reference example, input parameter description, function description, and fault problem can be assembled to obtain a first prompt message, wherein the first prompt message is used to prompt the large language model to perform a calling parameter generation task, and then the large language model can be called to process the first prompt message to obtain the tool calling parameters.

[0127] Exemplarily, each first reference example associated with the agent indicated by the root node can be obtained by calling the agent interface.

[0128] For example, an agent interface may include:

[0129] Interface name: name

[0130] Input: None.

[0131] Output: Agent name.

[0132] Interface name: description / / Functional description of the agent

[0133] Input: None.

[0134] Output: A description of the agent’s functionality.

[0135] Interface name: binding_tool_name / / Tool name bound to the agent

[0136] Input: None.

[0137] Output: The name of the bound tool.

[0138] Interface name: set_tool_input_examples / / Set tool input parameter examples (referred to as the first reference example in this disclosure)

[0139] Input: string. A collection of reference examples showing how to convert natural language information into a list of input parameters for tools bound to an agent.

[0140] Output: None.

[0141] Interface name: tool_input_examples / / Tool input parameter examples (referred to as the first reference example in this disclosure)

[0142] Input: None.

[0143] Output: The reference examples corresponding to the set_tool_input_examples interface settings are returned.

[0144] In summary, the first prompt information can be used as prior information or task information to indicate the task information to be performed by the large language model, which can improve the prediction accuracy of the large language model. In addition, a large number of reference examples can be listed in the first prompt information to indicate how to convert natural language information (such as fault problems in colloquial form) into input parameters of the tool, which can improve the stability of the large language model generation result process.

[0145] Step C: Based on the tool calling parameters, call the tool bound to the agent indicated by the root node to obtain the execution result (or output result) of the tool, which is recorded as the tool execution result in this disclosure.

[0146] Step D: Convert the tool execution results to obtain natural language text that matches the input format required by the agent indicated by the root node.

[0147] In the disclosed embodiment, the tool execution result may be converted to obtain a natural language text (or colloquial text, colloquial result) that matches the input format required by the agent indicated by the root node.

[0148] As a possible implementation method, a large language model can be used to convert the tool execution results to obtain the natural language text required by the agent indicated by the root node.

[0149] As an example, first, the output format description associated with the tool bound to the agent indicated by the root node can be obtained, and at least one second reference example associated with the agent indicated by the root node can be obtained, wherein the second reference example is used to indicate the processing method of the output result of the tool bound to the agent indicated by the root node to the input data of the agent indicated by the root node. Afterwards, the second prompt information can be generated according to each second reference example, tool execution result, output format description and function description. For example, the second reference examples, tool execution results, output format description and function description can be assembled to obtain the second prompt information, wherein the second prompt information is used to prompt the large language model to perform a text conversion task, and then the large language model can be called to process the second prompt information to obtain the natural language text required by the agent indicated by the root node.

[0150] Exemplarily, the output format description associated with the tool can be obtained through an interface constructed in the tool. For example, the following interface can also be constructed in the tool:

[0151] Interface name: output_param_description / / Output format description of the tool

[0152] Input: None.

[0153] Output: tool output parameter description in markdown format.

[0154] Exemplarily, each second reference example associated with the agent indicated by the root node can be obtained by calling the agent interface. For example, the agent interface can also include:

[0155] Interface name: set_tool_output_examples / / Example of setting tool output format (referred to as the second reference example in this disclosure)

[0156] Input: string. A series of reference examples showing how to convert the output of the tool into the natural language text (also known as spoken text, spoken results) required by the agent.

[0157] Output: None.

[0158] Interface name: tool_output_examples / / Tool output format example (referred to as the second reference example in this disclosure)

[0159] Input: None.

[0160] Output: Corresponds to the reference examples set by the set_tool_output_examples interface, and returns these reference examples.

[0161] Interface name: query / / Failure problem in the colloquial form of user input

[0162] Input: None.

[0163] Output: Colloquial questions.

[0164] In summary, the second prompt information can be used as prior information or task information to indicate the task information to be performed by the large language model, which can improve the prediction accuracy of the large language model. In addition, a large number of reference examples can be listed in the second prompt information to indicate how to convert the output results of the tool (such as structured results in Json format) into the natural language text (such as spoken text) required by the intelligent agent, which can improve the stability of the large language model generation result process.

[0165] Step E: Use the agent indicated by the root node to process the natural language text to obtain the processing result of the agent indicated by the root node.

[0166] Step S507: Process the fault problem through the agent indicated by the root node to obtain the processing result of the agent indicated by the root node.

[0167] In an embodiment of the present disclosure, when the currently called agent is the agent indicated by the root node in the target interaction topology diagram, and the agent indicated by the root node is a non-tool-type agent, the fault problem can be directly processed through the agent indicated by the root node (i.e., the non-tool-type agent) to obtain the processing result of the agent indicated by the root node.

[0168] Step S508, in response to the currently called agent being the agent indicated by the non-root node in the target interaction topology diagram, determine whether the agent indicated by the non-root node is a tool-type agent, if so, execute step S509, if not, execute step S510.

[0169] It should be noted that step S509 and step S510 are two parallel implementation methods. In actual application, only one needs to be executed.

[0170] Step S509, the processing result of the agent indicated by the non-root node is processed by the agent indicated by the non-root node and the tool bound to the agent indicated by the non-root node to obtain the processing result of the agent indicated by the non-root node.

[0171] In the disclosed embodiment, when the currently called agent is the agent indicated by the non-root node in the target interaction topology diagram, and the agent indicated by the non-root node is a tool-type agent, the agent indicated by the non-root node and the tool bound to the agent indicated by the non-root node can be used to process the processing result of the agent indicated by the parent node of the non-root node to obtain the processing result of the agent indicated by the non-root node. The implementation principle is similar to step S506 and will not be repeated here.

[0172] As an example, the input parameter description and function description associated with the tool bound to the agent indicated by the non-root node can be obtained, and the tool calling parameters can be determined based on the processing results, input parameter description and function description of the agent indicated by the parent node of the non-root node. After that, the tool bound to the agent indicated by the non-root node can be called based on the tool calling parameters to obtain the tool execution result. Then, the tool execution result can be converted to obtain a natural language text that matches the input format required by the agent indicated by the non-root node. Finally, the natural language text can be processed by the agent indicated by the non-root node to obtain the processing result of the agent indicated by the non-root node.

[0173] Step S510, the processing result of the agent indicated by the parent node is processed by the agent indicated by the non-root node to obtain the processing result of the agent indicated by the non-root node.

[0174] In an embodiment of the present disclosure, when the currently called agent is the agent indicated by a non-root node in the target interaction topology diagram, and the agent indicated by the non-root node is a non-instrumental agent, the processing results of the agent indicated by the parent node of the non-root node can be directly processed by the agent indicated by the non-root node (i.e., the non-instrumental agent) to obtain the processing results of the agent indicated by the non-root node.

[0175] Step S511, determining the root cause of the failure of the distributed business system according to the processing result of the intelligent agent indicated by the leaf node in the target interaction topology diagram.

[0176] In the disclosed embodiment, the root cause of the failure of the distributed business system can be determined by comprehensively analyzing the processing results of the intelligent agents indicated by the leaf nodes in the target interaction topology diagram.

[0177] The method for determining the root cause of a fault in the disclosed embodiment, combined with a tool-type intelligent agent and the tools bound thereto, performs collaborative processing of fault problems, which can not only improve the stability of problem processing, but also improve the accuracy of problem processing.

[0178] In order to clearly illustrate any embodiment of the present disclosure, the present disclosure also proposes a method for determining a root cause of a fault.

[0179] Figure 6 This is a flow chart of a method for determining a root cause of a fault provided in Embodiment 4 of the present disclosure.

[0180] like Figure 6 As shown, the method for determining the root cause of the fault may include the following steps S601 to S607:

[0181] Step S601: Obtain a fault problem input in response to a fault occurring in a distributed business system.

[0182] Step S602: classify the fault problem to obtain the target category to which the fault problem belongs.

[0183] Step S603, obtaining multiple target interaction topology graphs adapted to the target category.

[0184] Among them, the nodes in the target interaction topology diagram are used to indicate intelligent agents, and the edges between the nodes are used to indicate the interaction relationship between the intelligent agents. The intelligent agents include tool-type intelligent agents and / or non-tool-type intelligent agents bound to tools.

[0185] For explanations of steps S601 to S603, reference may be made to the relevant descriptions in any embodiment of the present disclosure, and will not be repeated here.

[0186] Step S604, determining the interaction order between the multiple target interaction topology graphs, and based on the interaction order, calling the multiple target interaction topology graphs in sequence.

[0187] Among them, the interaction order of multiple target interaction topology diagrams adapted to the target category can be pre-specified, or can also be manually specified by relevant personnel, and the embodiments of the present disclosure do not limit this.

[0188] In the embodiment of the present disclosure, multiple target interaction topology graphs may be called in sequence based on the interaction order between the multiple target interaction topology graphs.

[0189] Step S605, in response to the currently called target interaction topology diagram being the first called target interaction topology diagram, the intelligent agents indicated by the nodes in the first target interaction topology diagram are called in sequence to process the fault problem to obtain the processing result of the first target interaction topology diagram.

[0190] In the disclosed embodiment, when the currently called target interaction topology diagram is the first called target interaction topology diagram, the agents indicated by the nodes in the first target interaction topology diagram can be called in sequence to process the fault problem to obtain the processing result of the first target interaction topology diagram; wherein the processing result of the first target interaction topology diagram can be: the processing result of the agent indicated by the leaf node in the first target interaction topology diagram. Its implementation principle is similar to steps S504 to S510, which will not be described in detail here.

[0191] Step S606, in response to the fact that the target interaction topology map currently called is not the first target interaction topology map called, the intelligent agents indicated by each node in the non-first target interaction topology map are called in sequence, and the processing results of the target interaction topology map previously called are processed to obtain the processing results of the non-first target interaction topology map.

[0192] In the embodiment of the present disclosure, when the target interaction topology diagram currently called is not the target interaction topology diagram called for the first time, the agents indicated by the nodes in the non-first target interaction topology diagram can be called in sequence, and the processing result of the target interaction topology diagram called for the previous time can be processed to obtain the processing result of the non-first target interaction topology diagram. The processing result of the non-first target interaction topology diagram can be: the processing result of the agent indicated by the leaf node in the non-first target interaction topology diagram.

[0193] Step S607: determine the root cause of the failure of the distributed business system according to the processing result of the last called target interaction topology graph.

[0194] In the disclosed embodiment, the root cause of the failure of the distributed business system can be determined by integrating the processing results of the intelligent agents indicated by the leaf nodes in the target interaction topology diagram of the last call.

[0195] The method for determining the root cause of a fault in the disclosed embodiment is based on the interaction sequence between multiple target interaction topology graphs, and sequentially calls the agents in multiple target interaction topology graphs to collaboratively process the fault problem, which can ensure that the processing of each agent is based on the processing result of the previous agent, forming a complete logical chain. This sequence ensures the consistency and accuracy of the root cause analysis of the fault. In addition, since each agent is responsible for processing a specific part of the problem and passes the processing result to the next agent, it can ensure that all relevant fault factors are fully considered, avoiding misdiagnosis caused by omitting certain factors, thereby further improving the accuracy of determining the root cause of the fault.

[0196] In order to clearly explain how to obtain a target interaction topology map adapted to a target category in the above embodiment, the present disclosure also proposes a method for determining a root cause of a fault.

[0197] Figure 7 This is a flowchart of a method for determining a root cause of a fault provided in Embodiment 5 of the present disclosure.

[0198] like Figure 7 As shown, the method for determining the root cause of the fault may include the following steps S701 to S709:

[0199] Step S701: Obtain a fault problem input in response to a fault occurring in a distributed business system.

[0200] Step S702: classify the fault problem to obtain the target category to which the fault problem belongs.

[0201] For explanations of steps S701 to S702, reference may be made to the relevant descriptions in any embodiment of the present disclosure and will not be repeated here.

[0202] Step S703, determining whether the target category is a newly added problem category in the distributed business system, if so, executing steps S706 to S708, if not, executing steps S704 to S705.

[0203] It should be noted that steps S706 to S708 and steps S704 to S705 are two parallel implementation methods. In actual application, only one needs to be executed.

[0204] Step S704 : Based on the correspondence between the problem category and the business object, determine a first business object corresponding to the target category from a plurality of business objects maintaining the distributed business system.

[0205] Among them, the number of first business objects (such as business teams, business parties) can be one, or can be multiple, which is not limited in this embodiment of the present disclosure.

[0206] In an embodiment of the present disclosure, when the target category is an existing problem category in a distributed business system, the first business object corresponding to the target category can be determined from multiple business objects that maintain the distributed business system based on the correspondence between the created problem category and the business object (such as a business team, a business party).

[0207] Step S705, determining a target interaction topology map adapted to the target category according to the agent interaction topology map maintained by the first business object.

[0208] Among them, the nodes in the target interaction topology diagram are used to indicate intelligent agents, and the edges between the nodes are used to indicate the interaction relationship between the intelligent agents. The intelligent agents include tool-type intelligent agents and / or non-tool-type intelligent agents bound to tools.

[0209] As an example, the agent interaction topology map maintained by the first business object can be used as the target interaction topology map adapted to the target category.

[0210] Step S706: In response to the first configuration instruction, determine a second business object that is compatible with the target category from the multiple business objects, and establish a corresponding relationship between the target category and the second business object.

[0211] Among them, the number of second business objects (such as business teams, business parties) can be one, or can be multiple, which is not limited in the embodiments of the present disclosure.

[0212] In an embodiment of the present disclosure, when the target category is a newly added problem category in a distributed business system, a second business object that is compatible with the target category can be determined from multiple business objects in response to a first configuration instruction triggered by relevant personnel (such as an administrator), that is, the second business object that is compatible with the target category is manually selected by the relevant personnel, and a corresponding relationship between the target category and the second business object is established.

[0213] Step S707, dispatching the fault problem to the second business object so that the second business object updates the corresponding maintained intelligent agent interaction topology map.

[0214] In the disclosed embodiment, the fault problem may be dispatched to the second business object so that the second business object can update the agent interaction topology map maintained by itself based on existing business experience.

[0215] As a possible implementation method, a scheduling task can be sent to the second business object according to the fault problem and the target category, so that the second business object executes the scheduling task and updates the agent interaction topology map maintained by the second business object.

[0216] The scheduling task can be performed by following steps a to c:

[0217] Step a: Determine whether there is an agent in the agent library that is compatible with the target category. If so, execute step c; if not, execute steps b to c.

[0218] Step b: If there is no agent adapted to the target category in the agent library, then in response to the second configuration instruction triggered by the second business object, an agent adapted to the target category is created based on the fault problem.

[0219] As an example, the second business object may manually determine whether there is an agent adapted to the target category in the agent library. If not, the second business object may manually create an agent adapted to the target category based on the fault problem.

[0220] Step c: Based on the agents that are adapted to the target category, the agent interaction topology map maintained by the second business object is updated. For example, a new node is added to the agent interaction topology map maintained by the second business object, and the new node is used to indicate the agents that are adapted to the target category, or the node positions, dependencies or interactions between nodes in the agent interaction topology map maintained by the second business object can also be updated.

[0221] As an example, the agent interaction topology diagram maintained by the second business object can be updated based on the agent adapted to the target category by dragging and dropping.

[0222] It should be noted that when the resolution of a stability failure requires the participation of different business objects (or business parties), the business objects hold the most core experience of their business. If these experiences cannot be accumulated by the business objects themselves, then even if the intelligent solution system for solving stability failures is implemented well, it is not maintainable, which means that when the distributed business system changes, the business objects cannot update the knowledge or experience in the intelligent solution system in a timely manner, making the intelligent solution system no longer effective. In the present disclosure, for new stability failures in the distributed business system, new logic or intelligent entities can be introduced in an “additional” manner to achieve business maintainability. Moreover, the new logic or intelligent entity is introduced by the business objects (or business parties) that maintain the distributed business system, which can reflect business friendliness.

[0223] As another possible implementation, the scheduling task can also be performed by following steps d to f:

[0224] Step d: Determine whether there is an atomic tool associated with an intelligent agent adapted to the target category among the multiple atomic tools in the tool library. If so, execute step f; if not, execute steps e to f.

[0225] Step e: If there is no atomic tool associated with the intelligent agent adapted to the target category in the tool library, then according to the third configuration instruction, create at least one atomic tool associated with the intelligent agent adapted to the target category, and add the at least one atomic tool to the tool library.

[0226] As an example, the second business object can manually determine whether there is an atomic tool associated with an intelligent agent adapted to the target category in the tool library. If not, the second business object can manually create at least one atomic tool associated with an intelligent agent adapted to the target category based on the fault problem, and add the created atomic tools to the tool library.

[0227] Step f: In response to the fourth configuration instruction, assemble at least one atomic tool associated with the agent adapted to the target category to obtain a tool bound to the adapted agent.

[0228] As an example, at least one atomic tool associated with an agent adapted to the target category may be assembled by dragging and dropping or the like to obtain a tool bound to the agent adapted to the target category.

[0229] In summary, when a new stability failure occurs in a distributed business system, it is only necessary to build new tools or new intelligent agents in a low-cost manner. At the same time, no changes are made to the original tools or intelligent agents. This can make the intelligent solution system for solving stability failures more powerful without any operation on the solution capabilities of existing fault types.

[0230] Step S708, determining a target interaction topology map adapted to the target category based on the updated agent interaction topology map maintained by the second business object.

[0231] As an example, the updated agent interaction topology map maintained by the second business object can be used as the target interaction topology map adapted to the target category.

[0232] Step S709, calling the intelligent agent indicated by each node in the target interaction topology diagram in turn to process the fault problem, so as to obtain the root cause of the fault of the distributed business system.

[0233] For the explanation of step S709, reference may be made to the relevant description in any embodiment of the present disclosure, and no further details will be given here.

[0234] The method for determining the root cause of a fault in an embodiment of the present disclosure, when the target category is an existing problem category in a distributed business system, directly determines the first business object corresponding to the target category from multiple business objects that maintain the distributed business system based on the correspondence between the problem category and the business object, and uses the intelligent agent interaction topology map maintained by the first business object as the target interaction topology map adapted to the target category. This can achieve rapid resolution of fault problems of known problem categories based on existing business experience, thereby improving the efficiency of handling fault problems.

[0235] When the target category is a newly added problem category in the distributed business system, a second business object that is adapted to the newly added problem category is determined from multiple business objects that maintain the distributed business system through a visual configuration method, and the intelligent agent interaction topology map maintained by itself is updated based on the visual configuration method through the second business object, so as to solve the fault problem of the newly added problem category based on the updated intelligent agent interaction topology map, which can improve the accuracy and effectiveness of fault problem handling.

[0236] In order to clearly explain how to classify the fault problems in the above embodiments and obtain the target category to which the fault problems belong, the present disclosure also proposes a method for determining the root cause of the fault.

[0237] Figure 8 This is a flowchart of a method for determining a root cause of a fault provided in Embodiment 6 of the present disclosure.

[0238] like Figure 8 As shown, the method for determining the root cause of the fault may include the following steps S801 to S806:

[0239] Step S801: Obtain a fault problem input in response to a fault occurring in a distributed business system.

[0240] For the explanation of step S801, please refer to the relevant description in any embodiment of the present disclosure, which will not be repeated here.

[0241] Step S802, obtaining multiple third reference examples; wherein the third reference examples include reference questions and question categories to which the reference questions belong.

[0242] The problem categories to which the reference problems belong include, but are not limited to: query lossy problems, long-tail problems, Core problems (core system problems), etc.

[0243] As an example, the problem category to which the reference problem belongs may include a problem category that has occurred or existed in the distributed business system.

[0244] Among them, the third reference example is preset.

[0245] Step S803: Generate third prompt information based on multiple third reference examples and fault problems; wherein the third prompt information is used to prompt the large language model to perform the text classification task.

[0246] In the embodiment of the present disclosure, multiple third reference examples and fault problems can be assembled to obtain third prompt information, wherein the third prompt information is used to prompt the large language model to perform a text classification task.

[0247] Step S804: calling the large language model to process the third prompt information to obtain the target category to which the fault problem belongs.

[0248] In the disclosed embodiment, a large language model may be called to process the third prompt information to obtain the target category to which the fault problem belongs.

[0249] Step S805: Acquire a target interaction topology map adapted to the target category.

[0250] Among them, the nodes in the target interaction topology diagram are used to indicate intelligent agents, and the edges between the nodes are used to indicate the interaction relationship between the intelligent agents. The intelligent agents include tool-type intelligent agents and / or non-tool-type intelligent agents bound to tools.

[0251] Step S806, calling the intelligent agent indicated by each node in the target interaction topology diagram in turn to process the fault problem, so as to obtain the root cause of the fault of the distributed business system.

[0252] For explanations of steps S805 to S806, reference may be made to the relevant descriptions in any embodiment of the present disclosure, and will not be repeated here.

[0253] In the method for determining the root cause of a fault in the disclosed embodiment, the third prompt information can be used as prior information or task information to indicate the task information to be performed by the large language model, which can improve the prediction accuracy of the large language model. In addition, a large number of reference examples can be listed in the third prompt information to indicate the existing problem categories of the distributed business system, so that the large language model can only classify within the problem category range indicated by the third prompt information, without having to classify in all problem categories, which can eliminate classification illusions, thereby improving the reliability of the classification results, that is, improving the stability of the large language model generation result process.

[0254] In any embodiment of the present disclosure, the following may be used: Fig. 9 The low-code tool platform shown is used to implement the above-mentioned method embodiments, wherein the first layer (tool layer) and the second layer (scenario layer) in the low-code tool platform are the core two layers of the present disclosure.

[0255] The first layer (tool layer, also known as tool library, tool market) completes the construction of tools through low-code methods. Among them, each tool in the tool layer is also called atomic tool.

[0256] Fig. 9 Clickhouse in the tool layer is a completely column-based distributed database management system; the callgraph platform is a distributed service tracking system used to monitor and track the call relationships between components in a distributed business system; the tower platform is a project management tool designed for business team collaboration, which provides task management, project progress tracking, team collaboration and other functions to help business teams advance projects more efficiently; NL2SQL is a technology that converts natural language into SQL (Structured Query Language) queries (Natural Language to Structured Query Language).

[0257] The second layer (scenario layer, also called assembly layer) assembles the atomic tools of the first layer to form larger tools to solve relatively complex fault problems. In addition, intelligent agents can be created in the second layer, and the interaction topology diagram between intelligent agents can be constructed to achieve the purpose of solving complex problems through natural language.

[0258] The low-code tool platform has at least the following advantages:

[0259] Part one: stable execution capability.

[0260] 1.1. Fine-grained agent.

[0261] The simpler the responsibility of an intelligent agent, the more stable its execution is. Therefore, in the present disclosure, when constructing an intelligent agent, it is avoided to be too large, and an intelligent agent can only complete one task. If the intelligent agent has to complete too many things, the intelligent agent is split.

[0262] 1.2. Tool-type intelligent agent.

[0263] The reason why the intelligent agent has the problem of unstable execution is that it relies on LLM. However, because the tool is hard-coded logic, there is no problem of execution stability. In other words, the execution of the tool is the most stable, because in the present disclosure, the tool can also "play" the role of the intelligent agent and have the ability to interact with the outside world through natural language. This kind of intelligent agent can be called a tool-type intelligent agent.

[0264] The specific steps are as follows:

[0265] (1) Interfaces built in tools

[0266] Interface name: name / / name of the tool

[0267] Input: None.

[0268] Output: Tool name.

[0269] Interface name: description / / Function description of the tool

[0270] Input: None.

[0271] Output: Description of the tool's functionality.

[0272] Interface name: input_param_description / / tool input parameter description

[0273] Input: None.

[0274] Output: Description of tool input parameters in markdown format.

[0275] Interface name: output_param_description / / Output format description of the tool

[0276] Input: None.

[0277] Output: tool output parameter description in markdown format.

[0278] Interface name: run_with_params / / Call the tool based on the parameters, the execution result of the tool

[0279] Input: parameter list in json format.

[0280] Output: tool execution results.

[0281] (2) Agent interface / / Agent name

[0282] Interface name: name

[0283] Input: None.

[0284] Output: Agent name.

[0285] Interface name: description / / Functional description of the agent

[0286] Input: None.

[0287] Output: A description of the agent’s functionality.

[0288] Interface name: binding_tool_name / / The name of the tool bound to the agent

[0289] Input: None.

[0290] Output: The name of the bound tool.

[0291] Interface name: set_tool_input_examples / / Set tool input parameter examples

[0292] Input: string, i.e., a list of examples showing how to convert natural language information into a list of input parameters for a tool bound to an agent, i.e., a string indicating how to convert natural language information into input parameters for a tool;

[0293] Output: None.

[0294] Interface name: tool_input_examples / / Tool input parameter examples

[0295] Input: None.

[0296] Output: Examples corresponding to the set_tool_input_examples interface setting, which are returned.

[0297] Interface name: set_tool_output_examples / / Set tool output format examples

[0298] Input: a string, i.e., a series of examples showing how to convert the output of the tool into the spoken input data of the agent, i.e., indicating the specific processing method of the output of the tool into the spoken input data;

[0299] Output: None.

[0300] Interface name: tool_output_examples / / Tool output format examples

[0301] Input: None.

[0302] Output: Examples corresponding to the set_tool_output_examples interface setting are returned.

[0303] Interface name: query / / Failure problem of spoken language input by the user

[0304] Input: None.

[0305] Output: Colloquial questions.

[0306] As an example, you can use Fig.10 The interactive process shown is used to solve the query input by the user, where: Fig.10 The middle gray area refers to the tool execution engine, which is responsible for coordinating function calls between tools and agents and interacting with LLM. Fig.10 The area to the left of is the tool-type agent. Fig.10 The area to the right of is the tool, and the two are bound through the agent's binding_tool_name (binding tool name) interface.

[0307] Fig.10 It consists of two parts, the upper and lower parts. Fig.10 The first half of the main task is to convert the fault problem (query) described in colloquial language into tool call parameters. Specifically, the tool execution engine calls the tool-type agent and tool-related functions (such as tool_input_examples of the tool-type agent, description of the tool, input_param_description of the tool) to obtain the corresponding information and assemble it to obtain the prompt. Then, the tool execution engine uses the prompt to request the LLM and obtains the various parameters of the tool described in Json format returned by the LLM (i.e., tool call parameters). Finally, the tool execution engine can call the run_with_param interface of the tool based on these parameters, execute the tool, and obtain the execution result of the tool (structured result in Json format).

[0308] Fig.10The lower part mainly completes the conversion of the tool's execution result (structured result) into the spoken result required by the tool-type agent, that is, the conversion of the tool's execution result into a spoken result that matches the input format required by the tool-type agent. Specifically, first, the tool execution engine calls the tool-type agent and tool-related functions (such as the tool's description, the tool's output_param_description, and the tool_output_examples of the tool-type agent) to obtain the corresponding information and assemble it to obtain the prompt. Then, the tool execution engine uses the prompt to request the LLM to obtain the spoken result required by the tool-type agent returned by the LLM. Finally, the tool execution engine returns the spoken result to the agent for processing to obtain the processed result output by the agent.

[0309] 1.3. Enhance stability by conveniently introducing examples.

[0310] During the execution of the above-mentioned tool-based intelligent agent, a large number of examples were listed when constructing the two prompts to further improve the stability of the LLM result generation process.

[0311] The second part is maintainability.

[0312] 2.1. Align organizational structure.

[0313] The interactive architecture of tool-based agents follows the division of business teams. First, there is a general entry agent at the top level, and under the general entry agent, there are more agents divided according to each business team. For each business team, the business team has an entry agent that connects with the general entry agent, and the entry agent of the business team can include more sub-agents for division of labor and collaboration, such as Fig.11 shown.

[0314] exist Fig.11 In the architecture shown, each business team maintains the agent within its own business scope. In addition, the granularity of the agent can be split according to the granularity that the business considers appropriate, and different agents interact with each other in natural language.

[0315] The general entry agent is responsible for understanding the original query input by the user and organizing the entry agents of each business team to complete the task.

[0316] 2.2. Query routing for new fault types.

[0317] When a query of a new fault type appears, there is no need to modify the existing logic of the agent, but only to add new logic. Specifically:

[0318] (1) Add an example for the query of the new fault type in the main entry agent, indicating which business team the query corresponding to this fault type should be routed to, or through which business teams' collaboration to complete it.

[0319] (2) If there is a lack of an agent necessary to complete the query within the agent maintained by the business team, a new agent is created.

[0320] The third part is low cost.

[0321] 3.1. Low-cost tool assembly.

[0322] There are many kinds of atomic tools in the tool layer, which are the basis for building scenes. In order to support the ability of "low-cost scene building", each atomic tool can be designed as follows:

[0323] (1) The input format and output format are fixed, both are a key-value map.

[0324] (2) Multiple atomic tools can be connected through Fig.12 The "data whiteboard" shown transfers data.

[0325] The "data whiteboard" can be regarded as a central data storage or message queue for transferring data between different atomic tools. The data whiteboard acts as a neutral and accessible data exchange center, so that any atomic tool can read or write data on it. Transferring data through the "data whiteboard" has at least the following advantages:

[0326] Decoupling: Atomic tools do not need to communicate directly, but communicate indirectly through the "data whiteboard", which reduces the coupling between atomic tools and improves the maintainability and scalability of the system.

[0327] Asynchronous processing: Asynchronous data processing is supported. Atomic tools can publish results to the "data whiteboard" after completing the current task without having to wait for responses from other atomic tools.

[0328] Concurrent processing: Multiple atomic tools can read and process data from the "data whiteboard" at the same time, improving overall processing efficiency.

[0329] After supporting the above two capabilities, atomic tools can be assembled through the data whiteboard to form more complex tools.

[0330] Corresponding to specific problem scenarios, there are many types of problems. In order to support these scenarios, the combination of tools mainly includes the following: Fig.13 (a) Fig.13 (b) and Fig.13(c) shows three types, among which, Fig.13 (a) refers to the scenario where multiple atomic tools are used serially. Fig.13 (b) refers to the scenario of selecting which atomic tool to trigger based on the conditions. Fig.13 (c) refers to the scenario where multiple atomic tools are executed in parallel.

[0331] In terms of implementation, the assembly of atomic tools can be completed through the "drag and drop" method of the UI (User Interface).

[0332] 3.2. Low-cost assembly of intelligent agents.

[0333] Similar to tools, agents can also be assembled by "dragging" on the UI interface. Unlike tool assembly, the assembled tools are called through deterministic APIs, while the assembled agents interact with each other through natural language.

[0334] In summary, the solution provided by the present disclosure has at least the following advantages:

[0335] 1. Stability: By maintaining the fine-grained nature of the agent, implementing a tool-type agent, and using an example-based approach (i.e., setting examples), the stability requirements of the agent execution process are ensured;

[0336] 2. Maintainability: Align the agent architecture with the organizational architecture and introduce new logic by “increasing” the processing capabilities of new fault types to achieve business maintainability;

[0337] 3. Low cost: A “data whiteboard” for exchanging data between tools is implemented, and the function of supporting three assembly modes is implemented to achieve drag-and-drop assembly of tools; the assembly method of intelligent agents is similar. The logic in the intelligent agent is completely input through the natural language of the business. Through these methods, low-cost creation of tools and intelligent agents is achieved.

[0338] With the above Figures 3 to 8 Corresponding to the method for determining the root cause of a fault provided in the embodiment, the present disclosure further provides a device for determining the root cause of a fault. Figures 3 to 8 The method for determining the root cause of a fault provided in the embodiment corresponds to the method, so the implementation of the method for determining the root cause of a fault is also applicable to the device for determining the root cause of a fault provided in the embodiment of the present disclosure, and will not be described in detail in the embodiment of the present disclosure.

[0339] Fig.14 This is a schematic diagram of the structure of the device for determining the root cause of a fault provided in Embodiment 7 of the present disclosure.

[0340] like Fig.14As shown, the device 1400 for determining a root cause of a fault may include: a first acquisition module 1410 , a classification module 1420 , a second acquisition module 1430 , and a processing module 1440 .

[0341] The first acquisition module 1410 is used to acquire a fault problem input in response to a fault occurring in the distributed business system;

[0342] The classification module 1420 is used to classify the fault problem to obtain the target category to which the fault problem belongs;

[0343] The second acquisition module 1430 is used to acquire a target interaction topology map adapted to the target category; wherein the nodes in the target interaction topology map are used to indicate agents, and the edges between the nodes are used to indicate the interaction relationship between the agents, and the agents include tool-type agents and / or non-tool-type agents bound to the tools;

[0344] The processing module 1440 is used to sequentially call the intelligent agent indicated by each node in the target interaction topology diagram to process the fault problem so as to obtain the root cause of the fault of the distributed business system.

[0345] In a possible implementation of the embodiment of the present disclosure, the processing module 1440 is used to: sequentially call the intelligent agent indicated by each node in the target interaction topology diagram to handle the fault problem; and determine the root cause of the fault of the distributed business system based on the processing result of the intelligent agent indicated by the leaf node in the target interaction topology diagram.

[0346] In a possible implementation of the embodiments of the present disclosure, the processing module 1440 is used to: sequentially call the agents indicated by each node in the target interaction topology diagram; in response to the currently called agent being the agent indicated by the root node in the target interaction topology diagram, and the agent indicated by the root node being a tool-type agent, process the fault problem through the agent indicated by the root node and the tool bound to the agent indicated by the root node, and obtain the processing result of the agent indicated by the root node; in response to the agent indicated by the root node being a non-tool-type agent, process the fault problem through the agent indicated by the root node, and obtain the processing result of the agent indicated by the root node.

[0347] In a possible implementation of the embodiments of the present disclosure, the processing module 1440 is also used to: in response to the currently called agent being the agent indicated by the non-root node in the target interaction topology diagram, and the agent indicated by the non-root node being a tool-type agent, the processing result of the agent indicated by the parent node of the non-root node is processed through the agent indicated by the non-root node and the tool bound to the agent indicated by the non-root node to obtain the processing result of the agent indicated by the non-root node; in response to the agent indicated by the non-root node being a non-tool-type agent, the processing result of the agent indicated by the parent node is processed through the agent indicated by the non-root node to obtain the processing result of the agent indicated by the non-root node.

[0348] In a possible implementation of the embodiments of the present disclosure, the processing module 1440 is used to: determine the tool calling parameters according to the fault problem and the input parameter description and function description associated with the tool bound to the agent indicated by the root node; based on the tool calling parameters, call the tool bound to the agent indicated by the root node to obtain the tool execution result; convert the tool execution result to obtain a natural language text that matches the input format required by the agent indicated by the root node; and use the agent indicated by the root node to process the natural language text to obtain the processing result of the agent indicated by the root node.

[0349] In a possible implementation of the embodiments of the present disclosure, the processing module 1440 is used to: obtain at least one first reference example associated with the agent indicated by the root node; wherein the first reference example is used to indicate a conversion method for converting natural language information into input parameters of a tool bound to the agent indicated by the root node; generate first prompt information according to each first reference example, input parameter description, function description and fault problem; wherein the first prompt information is used to prompt the large language model to execute a call parameter generation task; and use the large language model to process the first prompt information to obtain tool call parameters.

[0350] In a possible implementation of the embodiments of the present disclosure, the processing module 1440 is used to: obtain an output format description associated with a tool bound to an agent indicated by a root node; obtain at least one second reference example associated with the agent indicated by the root node; wherein the second reference example is used to indicate a processing method for the output result of the tool bound to the agent indicated by the root node to the input data of the agent indicated by the root node; generate a second prompt message according to each second reference example, tool execution result, output format description and function description; wherein the second prompt message is used to prompt a large language model to perform a text conversion task; and the second prompt message is processed using a large language model to obtain the natural language text required by the agent indicated by the root node.

[0351] In a possible implementation of the embodiment of the present disclosure, there are multiple target interaction topology graphs, and the processing module 1440 is used to: determine the interaction order between the multiple target interaction topology graphs, and based on the interaction order, call the multiple target interaction topology graphs in sequence; in response to the target interaction topology graph currently called being the first target interaction topology graph called, the intelligent agents indicated by the nodes in the first target interaction topology graph are called in sequence to process the fault problem to obtain the processing result of the first target interaction topology graph; in response to the target interaction topology graph currently called being a target interaction topology graph not called for the first time, the intelligent agents indicated by the nodes in the non-first target interaction topology graph are called in sequence to process the processing result of the target interaction topology graph previously called to obtain the processing result of the non-first target interaction topology graph; determine the root cause of the fault of the distributed business system according to the processing result of the target interaction topology graph called for the last time.

[0352] In a possible implementation of the embodiment of the present disclosure, the second acquisition module 1430 is used to: determine whether the target category is a newly added problem category in the distributed business system; if not, based on the correspondence between the problem category and the business object, determine the first business object corresponding to the target category from multiple business objects that maintain the distributed business system; and determine the target interaction topology map adapted to the target category based on the intelligent agent interaction topology map maintained by the first business object.

[0353] In a possible implementation of the embodiment of the present disclosure, the second acquisition module 1430 is also used to: if yes, in response to the first configuration instruction, determine a second business object that is adapted to the target category from multiple business objects, and establish a corresponding relationship between the target category and the second business object; schedule the fault problem to the second business object so that the second business object updates the corresponding maintained intelligent agent interaction topology map; determine the target interaction topology map that is adapted to the target category based on the updated intelligent agent interaction topology map maintained by the second business object.

[0354] In a possible implementation of the embodiment of the present disclosure, the second acquisition module 1430 is used to: send a scheduling task to the second business object according to the fault problem and the target category, so that the second business object executes the scheduling task and updates the corresponding maintained agent interaction topology map;

[0355] Among them, the scheduling task is used to: determine whether there is an agent adapted to the target category in the agent library; if not, create an agent adapted to the target category based on the fault problem in response to the second configuration instruction; based on the adapted agent, update the agent interaction topology map maintained by the second business object.

[0356] In a possible implementation of the embodiments of the present disclosure, the scheduling task is also used to: determine whether there is an atomic tool associated with the adapted intelligent agent among multiple atomic tools in the tool library; if not, create at least one atomic tool associated with the adapted intelligent agent according to a third configuration instruction, and add at least one atomic tool to the tool library; and in response to a fourth configuration instruction, assemble at least one atomic tool to obtain a tool bound to the adapted intelligent agent.

[0357] In a possible implementation of the embodiments of the present disclosure, the classification module 1420 is used to: obtain multiple third reference examples; wherein the third reference examples include reference questions and the question category to which the reference questions belong; generate third prompt information based on the multiple third reference examples and the fault problem; wherein the third prompt information is used to prompt the large language model to perform a text classification task; and call the large language model to process the third prompt information to obtain a target category.

[0358] The device for determining the root cause of a fault in the disclosed embodiment, when a fault occurs in a distributed business system, determines an agent interaction topology diagram for handling the fault problem based on the fault category to which the fault problem belongs, and collaboratively handles the fault problem based on the agents indicated by each node in the agent interaction topology diagram to obtain the root cause of the fault of the distributed business system, which can not only solve the fault problem in complex scenarios, but also improve the accuracy of determining the root cause of the fault. In addition, the agent can be a tool-type agent, which can improve the stability of the agent execution, thereby improving the accuracy of determining the root cause of the fault.

[0359] In order to implement the above embodiments, the present disclosure also provides an electronic device, which may include at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the method for determining the root cause of the fault proposed in any of the above embodiments of the present disclosure.

[0360] In order to implement the above embodiments, the present disclosure further provides a non-transitory computer-readable storage medium storing computer instructions, wherein the computer instructions are used to enable a computer to execute the method for determining the root cause of a fault proposed in any of the above embodiments of the present disclosure.

[0361] In order to implement the above embodiments, the present disclosure further provides a computer program product, which includes a computer program. When the computer program is executed by a processor, it implements the method for determining the root cause of a fault proposed in any of the above embodiments of the present disclosure.

[0362] According to an embodiment of the present disclosure, the present disclosure also provides an electronic device, a readable storage medium and a computer program product.

[0363] Fig.15 A schematic block diagram of an example electronic device that can be used to implement an embodiment of the present disclosure is shown. Among them, the electronic device may include the server and client in the above-mentioned embodiment. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workbenches, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present disclosure described and / or required herein.

[0364] like Fig.15 As shown, the electronic device 1500 includes a computing unit 1501, which can perform various appropriate actions and processes according to a computer program stored in a ROM (Read-Only Memory) 1502 or a computer program loaded from a storage unit 1507 to a RAM (Random Access Memory) 1503. In the RAM 1503, various programs and data required for the operation of the device 1500 can also be stored. The computing unit 1501, the ROM 1502, and the RAM 1503 are connected to each other via a bus 1504. An I / O (Input / Output) interface 1505 is also connected to the bus 1504.

[0365] A number of components in the device 1500 are connected to the I / O interface 1505, including: an input unit 1506, such as a keyboard, a mouse, etc.; an output unit 1507, such as various types of displays, speakers, etc.; a storage unit 1508, such as a disk, an optical disk, etc.; and a communication unit 1509, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 1509 allows the device 1500 to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.

[0366] The computing unit 1501 may be a variety of general and / or special processing components with processing and computing capabilities. Some examples of the computing unit 1501 include, but are not limited to, a CPU (Central Processing Unit), a GPU (Graphic Processing Units), various dedicated AI (Artificial Intelligence) computing chips, various computing units running machine learning model algorithms, a DSP (Digital Signal Processor), and any appropriate processor, controller, microcontroller, etc. The computing unit 1501 performs the various methods and processes described above, such as the above-mentioned method for determining the root cause of the fault. For example, in some embodiments, the above-mentioned method for determining the root cause of the fault may be implemented as a computer software program, which is tangibly contained in a machine-readable medium, such as a storage unit 1508. In some embodiments, part or all of the computer program may be loaded and / or installed on the device 1500 via the ROM 1502 and / or the communication unit 1509. When the computer program is loaded into the RAM 1503 and executed by the computing unit 1501, one or more steps of the method for determining the root cause of the fault described above may be performed. Alternatively, in other embodiments, the computing unit 1501 may be configured to execute the above-mentioned method for determining the root cause of the fault in any other appropriate manner (for example, by means of firmware).

[0367] Various embodiments of the systems and techniques described above herein may be implemented in digital electronic circuit systems, integrated circuit systems, FPGAs (Field Programmable Gate Arrays), ASICs (Application-Specific Integrated Circuits), ASSPs (Application Specific Standard Products), SOCs (System On Chips), CPLDs (Complex Programmable Logic Devices), computer hardware, firmware, software, and / or combinations thereof. These various embodiments may include: being implemented in one or more computer programs that may be executed and / or interpreted on a programmable system including at least one programmable processor that may be a dedicated or general-purpose programmable processor that may receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0368] The program code for implementing the method of the present disclosure may be written in any combination of one or more programming languages. These program codes may be provided to a processor or controller of a general-purpose computer, a special-purpose computer, or other programmable data processing device, so that the program code, when executed by the processor or controller, enables the functions / operations specified in the flow chart and / or block diagram to be implemented. The program code may be executed entirely on the machine, partially on the machine, partially on the machine and partially on a remote machine as a stand-alone software package, or entirely on a remote machine or server.

[0369] In the context of the present disclosure, a machine-readable medium may be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, device, or equipment. A machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium may include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, device, or device, or any suitable combination of the foregoing. More specific examples of machine-readable storage media may include electrical connections based on one or more lines, portable computer disks, hard disks, RAM, ROM, EPROM (Electrically Programmable Read-Only-Memory) or flash memory, optical fiber, CD-ROM (Compact Disc Read-Only Memory), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.

[0370] To provide interaction with a user, the systems and techniques described herein can be implemented on a computer having: a display device (e.g., a CRT (Cathode-Ray Tube) or LCD (Liquid Crystal Display) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user can provide input to the computer. Other types of devices can also be used to provide interaction with the user; for example, the feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including acoustic input, voice input, or tactile input).

[0371] The systems and techniques described herein may be implemented in a computing system that includes backend components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes frontend components (e.g., a user computer with a graphical user interface or a web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such backend components, middleware components, or frontend components. The components of the system may be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: LAN (Local Area Network), WAN (Wide Area Network), the Internet, and blockchain networks.

[0372] A computer system may include a client and a server. The client and the server are generally remote from each other and usually interact through a communication network. The relationship between the client and the server is generated by computer programs running on the corresponding computers and having a client-server relationship with each other. The server may be a cloud server, also known as a cloud computing server or a cloud host, which is a host product in the cloud computing service system to solve the defects of difficult management and weak business scalability in traditional physical hosts and VPS services (Virtual Private Server). The server may also be a server of a distributed system, or a server combined with a blockchain.

[0373] It should be noted that artificial intelligence is a discipline that studies how computers can simulate certain human thought processes and intelligent behaviors (such as learning, reasoning, thinking, planning, etc.), and includes both hardware-level and software-level technologies. Artificial intelligence hardware technologies generally include technologies such as sensors, dedicated artificial intelligence chips, cloud computing, distributed storage, and big data processing; artificial intelligence software technologies mainly include computer vision technology, speech recognition technology, natural language processing technology, as well as machine learning / deep learning, big data processing technology, knowledge graph technology, and other major directions.

[0374] According to the technical solution of the embodiment of the present disclosure, when a fault occurs in a distributed business system, based on the fault category to which the fault problem belongs, an agent interaction topology map for handling the fault problem is determined, and based on the agents indicated by each node in the agent interaction topology map, the fault problem is handled collaboratively to obtain the root cause of the fault of the distributed business system, which can not only solve the fault problem in complex scenarios, but also improve the accuracy of determining the root cause of the fault. In addition, the agent can be a tool-type agent, which can improve the stability of the agent execution and thus improve the accuracy of determining the root cause of the fault.

[0375] It should be understood that the various forms of processes shown above can be used to reorder, add or delete steps. For example, the steps recorded in this disclosure can be executed in parallel, sequentially or in different orders, as long as the desired results of the technical solutions disclosed in this disclosure can be achieved, and this document does not limit this.

[0376] The above specific implementations do not constitute a limitation on the protection scope of the present disclosure. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modification, equivalent substitution and improvement made within the spirit and principle of the present disclosure shall be included in the protection scope of the present disclosure.

Claims

1. A method for determining a root cause of a fault, comprising: Obtaining a fault problem input in response to a fault occurring in a distributed business system; Classifying the fault problem to obtain a target category to which the fault problem belongs; Acquire a target interaction topology map adapted to the target category; wherein the nodes in the target interaction topology map are used to indicate agents, the edges between the nodes are used to indicate the interaction relationship between the agents, and the agents include tool-type agents and / or non-tool-type agents bound to tools; The intelligent agents indicated by the nodes in the target interaction topology diagram are called in sequence to process the fault problem, so as to obtain the root cause of the fault of the distributed business system.

2. The method according to claim 1, wherein: The step of sequentially calling the agents indicated by the nodes in the target interaction topology diagram to process the fault problem to obtain the root cause of the fault of the distributed business system includes: Invoke the intelligent agents indicated by the nodes in the target interaction topology diagram in sequence to handle the fault problem; The root cause of the failure of the distributed business system is determined according to the processing result of the intelligent agent indicated by the leaf node in the target interaction topology diagram.

3. The method according to claim 2, wherein: The step of sequentially calling the intelligent agents indicated by the nodes in the target interaction topology diagram to process the fault problem includes: Calling the intelligent agents indicated by the nodes in the target interaction topology graph in sequence; In response to the currently called agent being the agent indicated by the root node in the target interaction topology diagram, and the agent indicated by the root node being a tool-type agent, the fault problem is processed by the agent indicated by the root node and the tool bound to the agent indicated by the root node to obtain a processing result of the agent indicated by the root node; In response to the agent indicated by the root node being a non-tool-type agent, the fault problem is processed by the agent indicated by the root node to obtain a processing result of the agent indicated by the root node.

4. The method according to claim 3, wherein: The step of sequentially calling the agents indicated by the nodes in the target interaction topology diagram to process the fault problem further includes: In response to the currently called agent being the agent indicated by the non-root node in the target interaction topology diagram, and the agent indicated by the non-root node being a tool-type agent, the processing result of the agent indicated by the parent node of the non-root node is processed by the agent indicated by the non-root node and the tool bound to the agent indicated by the non-root node to obtain the processing result of the agent indicated by the non-root node; In response to the agent indicated by the non-root node being a non-tool-type agent, the processing result of the agent indicated by the parent node is processed by the agent indicated by the non-root node to obtain the processing result of the agent indicated by the non-root node.

5. The method according to claim 3, wherein: The process of processing the fault problem through the agent indicated by the root node and the tool bound to the agent indicated by the root node to obtain the processing result of the agent indicated by the root node includes: Determine tool call parameters according to the fault problem and the input parameter description and function description associated with the tool bound to the agent indicated by the root node; Based on the tool calling parameters, calling the tool bound to the agent indicated by the root node to obtain the tool execution result; Converting the tool execution result to obtain a natural language text that matches the input format required by the agent indicated by the root node; The natural language text is processed using the agent indicated by the root node to obtain a processing result of the agent indicated by the root node.

6. The method according to claim 5, wherein: The determining of tool call parameters according to the fault problem and the input parameter description and function description associated with the tool bound to the agent indicated by the root node includes: Acquire at least one first reference example associated with the agent indicated by the root node; wherein the first reference example is used to indicate a conversion method for converting natural language information into input parameters of a tool bound to the agent indicated by the root node; Generate first prompt information according to each of the first reference examples, the input parameter description, the function description and the fault problem; wherein the first prompt information is used to prompt the large language model to execute a call parameter generation task; The first prompt information is processed using the large language model to obtain the tool calling parameter.

7. The method according to claim 5, wherein: The step of converting the tool execution result to obtain a natural language text matching the input format required by the agent indicated by the root node comprises: Obtain an output format description associated with a tool bound to the agent indicated by the root node; Acquire at least one second reference example associated with the agent indicated by the root node; wherein the second reference example is used to indicate a processing method of the output result of the tool bound to the agent indicated by the root node to the input data of the agent indicated by the root node; Generate second prompt information according to each of the second reference examples, the tool execution result, the output format description and the function description; wherein the second prompt information is used to prompt the large language model to perform a text conversion task; The second prompt information is processed using the large language model to obtain the natural language text required by the agent indicated by the root node.

8. The method according to claim 1, wherein: There are multiple target interaction topology graphs, and sequentially calling the agents indicated by the nodes in the target interaction topology graph to process the fault problem to obtain the root cause of the fault of the distributed business system includes: Determine an interaction order between a plurality of target interaction topology graphs, and based on the interaction order, sequentially call the plurality of target interaction topology graphs; In response to the currently called target interaction topology diagram being the first called target interaction topology diagram, the agents indicated by the nodes in the first target interaction topology diagram are called in sequence to process the fault problem, so as to obtain a processing result of the first target interaction topology diagram; In response to the target interaction topology map currently being called being a target interaction topology map that is not the first one being called, the agents indicated by the nodes in the target interaction topology map that is not the first one being called are called in sequence, and the processing result of the target interaction topology map previously being called is processed to obtain the processing result of the target interaction topology map that is not the first one being called; The root cause of the failure of the distributed business system is determined according to the processing result of the last called target interaction topology diagram.

9. The method according to any one of claims 1 to 8, wherein: The acquiring of a target interaction topology map adapted to the target category includes: Determining whether the target category is a newly added problem category in the distributed business system; If not, determining a first business object corresponding to the target category from a plurality of business objects maintaining the distributed business system based on a correspondence between the problem category and the business object; According to the agent interaction topology map maintained by the first business object, the target interaction topology map adapted to the target category is determined.

10. The method according to claim 9, wherein: The acquiring of the target interaction topology map adapted to the target category further includes: If so, in response to the first configuration instruction, determining a second business object adapted to the target category from the multiple business objects, and establishing a corresponding relationship between the target category and the second business object; Dispatching the fault problem to the second business object so that the second business object updates the corresponding maintained agent interaction topology graph; According to the updated agent interaction topology map maintained by the second business object, the target interaction topology map adapted to the target category is determined.

11. The method according to claim 10, wherein: The step of dispatching the fault problem to the second business object so that the second business object updates the corresponding agent interaction topology graph maintained includes: Sending a scheduling task to the second business object according to the fault problem and the target category, so that the second business object executes the scheduling task and updates the corresponding maintained agent interaction topology map; Wherein, the scheduling task is used to: Determine whether there is an agent in the agent library that is adapted to the target category; If not, in response to a second configuration instruction, create an agent adapted to the target category based on the fault problem; Based on the adapted agent, the agent interaction topology map maintained by the second business object is updated.

12. The method according to claim 11, wherein: The scheduling task is also used to: Determine whether there is an atomic tool associated with the adapted intelligent agent among multiple atomic tools in the tool library; If not, creating at least one atomic tool associated with the adapted agent according to a third configuration instruction, and adding the at least one atomic tool to the tool library; In response to the fourth configuration instruction, the at least one atomic tool is assembled to obtain a tool bound to the adapted agent.

13. The method according to any one of claims 1 to 8, wherein: The classifying the fault problem to obtain the target category to which the fault problem belongs includes: Acquire multiple third reference examples; wherein the third reference examples include reference questions and question categories to which the reference questions belong; Generate third prompt information according to the plurality of third reference examples and the fault problem; wherein the third prompt information is used to prompt the large language model to perform a text classification task; The large language model is called to process the third prompt information to obtain the target category.

14. A device for determining a root cause of a fault, comprising: A first acquisition module is used to acquire a fault problem input in response to a fault occurring in the distributed business system; A classification module, used to classify the fault problem to obtain a target category to which the fault problem belongs; A second acquisition module is used to acquire a target interaction topology map adapted to the target category; wherein the nodes in the target interaction topology map are used to indicate agents, and the edges between the nodes are used to indicate the interaction relationship between the agents, and the agents include tool-type agents and / or non-tool-type agents bound to tools; The processing module is used to call the intelligent agent indicated by each node in the target interaction topology diagram in turn to process the fault problem so as to obtain the root cause of the fault of the distributed business system.

15. The device according to claim 14, wherein: The processing module is used for: Invoke the intelligent agents indicated by the nodes in the target interaction topology diagram in sequence to handle the fault problem; The root cause of the failure of the distributed business system is determined according to the processing result of the intelligent agent indicated by the leaf node in the target interaction topology diagram.

16. The device according to claim 15, wherein: The processing module is used for: Calling the intelligent agents indicated by the nodes in the target interaction topology graph in sequence; In response to the currently called agent being the agent indicated by the root node in the target interaction topology diagram, and the agent indicated by the root node being a tool-type agent, the fault problem is processed by the agent indicated by the root node and the tool bound to the agent indicated by the root node to obtain a processing result of the agent indicated by the root node; In response to the agent indicated by the root node being a non-tool-type agent, the fault problem is processed by the agent indicated by the root node to obtain a processing result of the agent indicated by the root node.

17. The device according to claim 16, wherein: The processing module is further used for: In response to the currently called agent being the agent indicated by the non-root node in the target interaction topology diagram, and the agent indicated by the non-root node being a tool-type agent, the processing result of the agent indicated by the parent node of the non-root node is processed by the agent indicated by the non-root node and the tool bound to the agent indicated by the non-root node to obtain the processing result of the agent indicated by the non-root node; In response to the agent indicated by the non-root node being a non-tool-type agent, the processing result of the agent indicated by the parent node is processed by the agent indicated by the non-root node to obtain the processing result of the agent indicated by the non-root node.

18. The device according to claim 16, wherein: The processing module is used for: Determine tool call parameters according to the fault problem and the input parameter description and function description associated with the tool bound to the agent indicated by the root node; Based on the tool calling parameters, calling the tool bound to the agent indicated by the root node to obtain the tool execution result; Converting the tool execution result to obtain a natural language text that matches the input format required by the agent indicated by the root node; The natural language text is processed using the agent indicated by the root node to obtain a processing result of the agent indicated by the root node.

19. The device according to claim 18, wherein: The processing module is used for: Acquire at least one first reference example associated with the agent indicated by the root node; wherein the first reference example is used to indicate a conversion method for converting natural language information into input parameters of a tool bound to the agent indicated by the root node; Generate first prompt information according to each of the first reference examples, the input parameter description, the function description and the fault problem; wherein the first prompt information is used to prompt the large language model to execute a call parameter generation task; The first prompt information is processed using the large language model to obtain the tool calling parameter.

20. The device according to claim 18, wherein The processing module is used for: Obtain an output format description associated with a tool bound to the agent indicated by the root node; Acquire at least one second reference example associated with the agent indicated by the root node; wherein the second reference example is used to indicate a processing method of the output result of the tool bound to the agent indicated by the root node to the input data of the agent indicated by the root node; Generate second prompt information according to each of the second reference examples, the tool execution result, the output format description and the function description; wherein the second prompt information is used to prompt the large language model to perform a text conversion task; The second prompt information is processed using the large language model to obtain the natural language text required by the agent indicated by the root node.

21. The device according to claim 14, wherein: There are multiple target interaction topology graphs, and the processing module is used to: Determine an interaction order between a plurality of target interaction topology graphs, and based on the interaction order, sequentially call the plurality of target interaction topology graphs; In response to the currently called target interaction topology diagram being the first called target interaction topology diagram, the agents indicated by the nodes in the first target interaction topology diagram are called in sequence to process the fault problem, so as to obtain a processing result of the first target interaction topology diagram; In response to the target interaction topology map currently being called being a target interaction topology map that is not the first one being called, the agents indicated by the nodes in the target interaction topology map that is not the first one being called are called in sequence, and the processing result of the target interaction topology map previously being called is processed to obtain the processing result of the target interaction topology map that is not the first one being called; The root cause of the failure of the distributed business system is determined according to the processing result of the last called target interaction topology diagram.

22. The device according to any one of claims 14 to 21, wherein: The second acquisition module is used to: Determining whether the target category is a newly added problem category in the distributed business system; If not, determining a first business object corresponding to the target category from a plurality of business objects maintaining the distributed business system based on a correspondence between the problem category and the business object; According to the agent interaction topology map maintained by the first business object, the target interaction topology map adapted to the target category is determined.

23. The device according to claim 22, wherein: The second acquisition module is further used for: If so, in response to the first configuration instruction, determining a second business object adapted to the target category from the multiple business objects, and establishing a corresponding relationship between the target category and the second business object; Dispatching the fault problem to the second business object so that the second business object updates the corresponding maintained agent interaction topology graph; According to the updated agent interaction topology map maintained by the second business object, the target interaction topology map adapted to the target category is determined.

24. The device according to claim 23, wherein: The second acquisition module is used to: Sending a scheduling task to the second business object according to the fault problem and the target category, so that the second business object executes the scheduling task and updates the corresponding maintained agent interaction topology map; Wherein, the scheduling task is used to: Determine whether there is an agent in the agent library that is adapted to the target category; If not, in response to a second configuration instruction, create an agent adapted to the target category based on the fault problem; Based on the adapted agent, the agent interaction topology map maintained by the second business object is updated.

25. The device according to claim 24, wherein: The scheduling task is also used to: Determine whether there is an atomic tool associated with the adapted intelligent agent among multiple atomic tools in the tool library; If not, creating at least one atomic tool associated with the adapted agent according to a third configuration instruction, and adding the at least one atomic tool to the tool library; In response to the fourth configuration instruction, the at least one atomic tool is assembled to obtain a tool bound to the adapted agent.

26. The device according to any one of claims 14 to 21, wherein: The classification module is used to: Acquire multiple third reference examples; wherein the third reference examples include reference questions and question categories to which the reference questions belong; Generate third prompt information according to the plurality of third reference examples and the fault problem; wherein the third prompt information is used to prompt the large language model to perform a text classification task; The large language model is called to process the third prompt information to obtain the target category.

27. An electronic device comprising: at least one processor; as well as a memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the method for determining a root cause of a fault according to any one of claims 1 to 13.

28. A non-transitory computer-readable storage medium storing computer instructions, wherein: The computer instructions are used to cause the computer to execute the method for determining a root cause of a fault according to any one of claims 1-13.

29. A computer program product, comprising a computer program, wherein when the computer program is executed by a processor, the computer program implements the steps of the method for determining a root cause of a fault according to any one of claims 1 to 13.

Images