IT resource maintenance influence range intelligent analysis method and system

Abstract

The invention belongs to the technical field of enterprise IT resource operation and maintenance analysis, and discloses an IT resource maintenance influence range intelligent analysis method and system, and the method comprises the steps: retrieving a constructed IT resource RAG knowledge base according to maintenance plan information, and obtaining IT resource information related to the maintenance plan information; calling a large language model for analysis according to the IT resource information associated with the maintenance plan information to obtain an IT resource maintenance influence range; wherein the construction of the IT resource RAG knowledge base comprises the following steps: modeling IT resources to obtain an IT resource association relationship model; collecting IT resource data according to the IT resource association relationship model to obtain IT resource ledger data and IT resource association relationship data; and carrying out vectorization processing on the IT resource machine account data and the IT resource association relationship data, and constructing an IT resource RAG knowledge base. According to the invention, accurate analysis of the influence range of the IT resource maintenance plan is realized.

Description

Technical Field

[0001] This invention belongs to the field of enterprise IT resource operation and maintenance analysis technology, and in particular relates to an intelligent analysis method and system for the impact range of IT resource maintenance. Background Technology

[0002] The State Grid Corporation of China is responsible for power transmission, distribution, and sales in most parts of the country. To ensure stable operation, the company has built IT infrastructure, a cloud platform, a data middleware platform, a business middleware platform, and hundreds of various business systems. Provincial power companies, data centers in the three regions, and municipal companies use the State Grid Corporation's intelligent integrated operation and maintenance platform to operate and maintain the IT infrastructure, cloud platform, data middleware platform, business middleware platform, and business systems.

[0003] In operations and maintenance (O&M) work, maintenance plan management is involved. When filling out the maintenance plan application form, the scope of impact for the maintenance needs to be specified. Currently, O&M personnel can only analyze the scope of impact based on experience, which is not accurate enough. The current difficulties in analyzing the scope of impact for IT resource maintenance include: 1. Numerous integration relationships between information systems: Due to the large and complex nature of the State Grid Corporation's business, there are numerous internal business systems with extensive interface, page, and data integration relationships. Inter-system interactions are frequent. Furthermore, these business systems are constantly iterating and updating, leading to continuous updates in their integration relationships. When submitting an information system maintenance plan, O&M personnel can only analyze the scope of impact directly related to the system being maintained, unable to deeply analyze the indirect relationships between multiple systems, resulting in inaccurate impact analysis. 2. Numerous dependencies between microservices within information systems: Currently, the State Grid Corporation's internal information systems adopt a microservice architecture, with numerous interface call relationships between microservices. Simultaneously, these business systems are constantly iterating and updating, leading to a continuous increase in the number of microservices and updates in their interface call relationships. When submitting maintenance plans for microservices or functional modules of information systems, maintenance personnel can only analyze the impact range directly related to the microservices currently being maintained based on experience. They cannot deeply analyze the indirect relationships between multiple microservices, leading to inaccurate analysis of the maintenance impact range. 3. Complex cloud platform architecture: Cloud platforms are based on virtualization technology, enabling the sharing and reuse of computing, storage, and network resources of physical servers. This allows for rapid expansion and migration of virtual resources and rapid change control of virtual networks. As the current resource support platform for business system operation, cloud platform maintenance requires accurate assessment of the impact on cloud-based business systems; otherwise, inadequate consideration may lead to unexpected failures. 4. Numerous dependencies between physical servers, network devices, and security devices: The underlying IT infrastructure contains numerous physical servers with direct or indirect network connections. When a physical device needs maintenance, it is necessary to accurately assess whether it affects the operation of other devices and the operation of virtual resources on the upper-layer cloud platform; otherwise, inadequate consideration may also lead to unexpected failures. Summary of the Invention

[0004] The purpose of this invention is to overcome the shortcomings of the prior art and provide an intelligent analysis method and system for the impact range of IT resource maintenance, so as to achieve accurate analysis of the impact range of IT resource maintenance plans.

[0005] To solve the above-mentioned technical problems, the present invention is implemented using the following solution: This invention provides an intelligent analysis method for the impact range of IT resource maintenance, including: Based on the maintenance plan information, the IT resource RAG knowledge base that has been built is retrieved to obtain the IT resource information associated with the maintenance plan information; Based on the IT resource information associated with the maintenance plan information, the large language model is used for analysis to obtain the scope of the IT resource maintenance impact. The construction of the IT resource RAG knowledge base includes: By modeling IT resources, a model of the relationships between IT resources is obtained; IT resource data is collected based on the IT resource relationship model to obtain IT resource ledger data and IT resource relationship data. The IT resource ledger data and IT resource relationship data are vectorized to construct an IT resource RAG knowledge base.

[0006] Furthermore, based on the maintenance plan information retrieval and the constructed IT resource RAG knowledge base, the associated IT resource information is obtained, including: Obtain maintenance target information from maintenance plan information; Based on the IT resource RAG knowledge base built by retrieving information about the maintenance object, we can obtain IT resource information that is directly or indirectly related to the maintenance object.

[0007] Furthermore, the large language model is the DeepSeek R1 large language model. Based on the IT resource information associated with the maintenance plan information, the large language model is invoked for analysis to obtain the scope of the IT resource maintenance impact, including: Based on the IT resource information directly and indirectly related to the object under maintenance, DeepSeek R1 large language model is invoked for analysis to generate an IT resource maintenance impact scope analysis report.

[0008] Furthermore, IT resources are divided into three layers: information system layer, basic platform layer, and physical infrastructure layer. Modeling these IT resources yields an IT resource relationship model, including: Establish horizontal relationship models for IT resources at the information system layer, basic platform layer, and physical infrastructure layer, respectively; Establish a vertical relationship model for IT resources at the information system layer, basic platform layer, and physical infrastructure layer.

[0009] Furthermore, horizontal relationship models are established for IT resources at the information system layer, basic platform layer, and physical infrastructure layer, including: The information system layer constructs a model of the relationships between information systems, micro-applications, microservices, and interfaces; the basic platform layer constructs a model of the relationships between cloud platforms, cloud platform components, cloud resource clusters, and cloud resource instances; and the physical infrastructure layer constructs a model of the relationships between computer rooms, server racks, servers, network equipment, security equipment, power supply equipment, and environmental equipment. Establish a vertical relationship model for IT resources at the information system layer, basic platform layer, and physical infrastructure layer, including: Build a model of the relationships between microservices, PODs, ECS, and servers.

[0010] Furthermore, based on the IT resource relationship model, IT resource data is collected to obtain IT resource ledger data and IT resource relationship data, including: Use APM tools to collect data on information systems, micro-applications, microservices, microservice interface ledgers, and related relationship data; Use cloud platform management tools to collect cloud platform, cloud resource cluster, cloud resource instance ledger data and related relationship data; The State Grid Unified Agent is used to collect ledger data and related relationship data of servers, network devices, security devices, and storage devices; Use environmental management tools to collect ledger data and related relationship data for power supply equipment, air conditioners, and temperature and humidity sensors.

[0011] Furthermore, it also includes: The collected IT resource ledger data is stored in the ledger database, and the collected IT resource relationship data is stored in the relationship graph database.

[0012] This invention also provides an intelligent analysis system for the impact range of IT resource maintenance, comprising: The intelligent agent for analyzing the scope of maintenance impact is used to retrieve the built IT resource RAG knowledge base based on the maintenance plan information, obtain the IT resource information associated with the maintenance plan information, and call the large language model to analyze the IT resource maintenance impact based on the IT resource information associated with the maintenance plan information to obtain the scope of IT resource maintenance impact. The construction of the IT resource RAG knowledge base includes: By modeling IT resources, a model of the relationships between IT resources is obtained; IT resource data is collected based on the IT resource relationship model to obtain IT resource ledger data and IT resource relationship data. The IT resource ledger data and IT resource relationship data are vectorized to construct an IT resource RAG knowledge base.

[0013] Furthermore, it also includes: The Maintenance Impact Scope Analysis API is used to receive maintenance plan information. Beneficial effects

[0014] This invention models IT resources within the State Grid's operational scope, constructs a comprehensive IT resource topology diagram, automatically collects IT resource ledger data and relational data, and utilizes the intelligent analysis capabilities of the large model to intelligently analyze the impact range of IT resource maintenance, making adequate contingency plans for maintenance-related faults in advance, and reducing faults caused by insufficient maintenance considerations. Attached Figure Description

[0015] Figure 1 This is a flowchart of an intelligent analysis method for the impact range of IT resource maintenance provided in Embodiment 1 of the present invention; Figure 2 This is a flowchart illustrating an intelligent agent for analyzing the scope of inspection and maintenance impact provided in Embodiment 1 of the present invention; Figure 3 This is a schematic diagram of inputting maintenance plan information into an intelligent agent for analyzing the scope of maintenance impact, provided in Embodiment 1 of the present invention; Figure 4 This is a schematic diagram of an intelligent agent for analyzing the impact range of maintenance and repair, as provided in Embodiment 1 of the present invention, to achieve knowledge retrieval of IT resource association relationships. Figure 5 This is a schematic diagram of an intelligent agent for analyzing the impact range of maintenance provided in Embodiment 1 of the present invention, which realizes intelligent analysis of the impact range of maintenance. Figure 6 This is a schematic diagram of an intelligent agent for analyzing the impact range of maintenance and repair, as provided in Embodiment 1 of the present invention, which outputs an impact range analysis report. Figure 7 This is an architecture diagram of an intelligent analysis system for the impact range of IT resource maintenance provided in Embodiment 2 of the present invention; Figure 8 This is a topology diagram of the relationship between IT resources related to server maintenance provided in Embodiment 3 of the present invention. Detailed Implementation

[0016] The present invention will be further described below with reference to the accompanying drawings. The following embodiments are only used to more clearly illustrate the technical solution of the present invention, and should not be used to limit the scope of protection of the present invention. Example 1

[0017] like Figure 1 and Figure 2 As shown in the figure, this embodiment provides an intelligent analysis method for the impact range of IT resource maintenance. The specific implementation steps are as follows: Step 1: As Figure 3 As shown, the starting node receives basic maintenance plan information, obtains maintenance object information from the maintenance plan information, and then analyzes the impact range of subsequent maintenance based on the maintenance object information.

[0018] Step Two: As Figure 4As shown, based on the information of the object under maintenance, the IT resource information directly and indirectly related to the object under maintenance is retrieved from the constructed IT resource RAG knowledge base.

[0019] Step 3: As Figure 5 As shown, based on all relevant IT resource information retrieved from the constructed IT resource RAG knowledge base, the DeepSeek R1 large language model is called for analysis, and the analysis is performed according to the prompt words designed for this node, and the JSON results are output.

[0020] Step Four: As Figure 6 As shown, based on the text of the IT resource maintenance impact range analyzed by the DeepSeek R1 large language model, the text is organized according to user requirements, and an IT resource maintenance impact range analysis report is automatically generated and fed back into the current maintenance plan.

[0021] The construction process of the IT resource RAG knowledge base includes the following steps: Step 1: Model the IT resources to obtain a model of their relationships. The specific process is as follows: S1) Establish horizontal relationships All IT resources are divided into three layers: the information system layer, the infrastructure platform layer, and the physical infrastructure layer. A horizontal relationship model is established for each of these three layers. The information system layer establishes the relationships between information systems, micro-applications, microservices, and interfaces; the infrastructure platform layer establishes the relationships between the cloud platform, cloud platform components (such as RDS components), cloud resource clusters (such as Kubernetes clusters and Elasticsearch clusters), and cloud resource instances (such as Elasticsearch nodes); and the physical infrastructure layer establishes the relationships between data centers, server racks, servers, network equipment, security equipment, power supply equipment, and environmental protection equipment.

[0022] S2) Establish vertical relationships A vertical relationship model is established for IT resources at the information system layer, basic platform layer, and physical infrastructure layer, constructing the relationships between microservices, PODs, ECSs, and servers. Specific information is shown in Table 1. Table 1. IT Resource Relationship Table IT resource types Relationship IT resource types Relationship type Information systems Include Micro Applications Horizontal Micro Applications Include microservices Horizontal microservices supply interface Horizontal interface Call interface Horizontal cloud platform Include Cloud platform components Horizontal cloud platform Include Cloud resource cluster Horizontal Cloud resource instances belong Cloud resource cluster Horizontal Power supply equipment belong engine room Horizontal air conditioner belong engine room Horizontal Temperature and humidity sensor belong engine room Horizontal server rack Installed at engine room Horizontal server Installed at server rack Horizontal Network equipment Installed at engine room Horizontal Safety equipment Installed at engine room Horizontal storage devices Installed at engine room Horizontal microservices Deployed at POD Vertical ECS Running in server Vertical Step 2: Collect IT resource data based on the IT resource relationship model to obtain IT resource ledger data and IT resource relationship data. The specific process is as follows: S1) Use APM tools to collect data on information systems, micro-applications, microservices, microservice interface ledgers, and related relationship data.

[0023] S2) Use cloud platform management tools to collect cloud platform, cloud resource cluster, cloud resource instance ledger data and related relationship data.

[0024] S3) Use the State Grid Unified Agent to collect ledger data and related relationship data of servers, network devices, security devices, and storage devices.

[0025] S4) Use environmental management tools to collect ledger data and related relationship data of power equipment, air conditioners, temperature and humidity sensors.

[0026] S5) Store the collected IT resource ledger data into the ledger database, and store the collected IT resource relationship data into the relationship graph database.

[0027] Step 3: Vectorize the IT resource ledger data and IT resource relationship data to build an IT resource RAG knowledge base. The specific process is as follows: Build an IT resource RAG knowledge base, synchronize ledger data and related relationship data, vectorize IT resource ledger data and IT resource related relationship data, and build an IT resource RAG knowledge base. Example 2

[0028] like Figure 7 As shown, this embodiment provides an intelligent analysis system for the impact range of IT resource maintenance, used to implement the intelligent analysis method for the impact range of IT resource maintenance described in Embodiment 1. The core of the system architecture is an intelligent agent for analyzing the impact range of maintenance. Based on the open-source Dify platform, it uses the DeepSeek R1 large language model to retrieve the constructed IT resource RAG knowledge base according to the maintenance plan information, obtain the IT resource information associated with the maintenance plan information, and call the large language model to analyze the IT resource information associated with the maintenance plan information to obtain the impact range of IT resource maintenance.

[0029] In addition, the system also includes: an API for analyzing the scope of maintenance impact, used to receive maintenance plan information; an IT resource association database for resource management, enabling storage management of IT resource association model, IT resource ledger data, and IT resource association data, and connecting to the IT resource RAG knowledge base to support resource association queries; and a resource acquisition module for collecting IT resource data from the information system layer, basic platform layer, and physical infrastructure layer. Example 3

[0030] To verify and illustrate the technical effectiveness of the intelligent analysis method for the impact range of IT resource maintenance described in Example 1 and the intelligent analysis system for the impact range of IT resource maintenance described in Example 2, this example provides specific verification tests of the application of the intelligent analysis method for the impact range of IT resource maintenance described in Example 1 and the intelligent analysis system for the impact range of IT resource maintenance described in Example 2, as follows: S1) Taking the maintenance of physical servers at a provincial branch of the State Grid as an example, data was collected on physical servers, network equipment, security equipment, storage devices, ECS, RDS, Kubernetes, microservices, microapplications, information system ledgers, and related data. Specific IT resource ledger information and related data are shown in the table below. Due to the involvement of real production environment data, only some fields are listed. Furthermore, due to the large volume of data, this document only lists a portion of the data.

[0031] Table 2. IT Resource Ledger of a State Grid Company Serial Number Resource ID Resource Name IP address Resource types 1 5734D53F-FD7A-4CEF-B2E1-E78460E43E51 Jiangsu Company Jiangning Data Center A01 Rack PC Server-43E51 25.*.*.17 server 2 58E32BE8-5EF4-406A-A248-FF1DD7098DFE Jiangsu Company Jiangning Data Center A01 Rack PC Server - 98DFE 25.*.*.39 server 3 5E03E76F-5B59-43BB-AC8C-E22336EBB4A9 Jiangsu Company Jiangning Data Center A02 Rack PC Server - BB4A9 25.*.*.101 server 4 68B28433-1B2E-42A4-9FD5-B4D91364A269 Jiangsu Company Jiangning Data Center A02 Rack PC Server-4A269 25.*.*.103 server 5 de8044a36267b242d27991a5f4b3d43d Power Trading System ECS-3d43d 25.*.*.190 ECS 6 238299571258989aa60f017ba0d16c5b Power Trading System ECS-16c5b 25.*.*.247 ECS 7 279cbd12e87a9473a671db5f839d398e Power Trading System ECS-d398e 25.*.*.114 ECS 8 3eda543a9d5e33ef1d69cb1b7524b822 Power Trading System ECS-4b822 25.*.*.23 ECS 9 4d2e7369940f93f303a67c18dfc561cd Power Trading System ECS-561cd 25.*.*.246 ECS 10 016f67889ad645849010c937bbf71294 Power Trading System ECS-71294 25.*.*.177 ECS 11 a45ad49a3b817eec27eb5658a9709020 Power Trading System ECS-09020 25.*.*.175 ECS 12 645fcfbbd6690d85350a5ce9bc718d40 Power Trading System RDS-18d40 20.*.*.171 RDS 13 6c12ddcbcf0704f30266f902cceb2677 Power Trading System RDS-b2677 20.*.*.161 RDS 14 7359a4ce94362138b20050e4a54611d9 Power Trading System RDS-611d9 20.*.*.23 RDS 15 7c6152fbd58ac86fcbc8b5bfb6ff1932 Power Trading System RDS-f1932 20.*.*.130 RDS 16 26fde48ad0f434b3e87cdb15faceb184 Power Trading System K8S Cluster - eb184 / K8S 17 000234fa80b0ed00368956c8f9921fa9 Power Trading System K8S Cluster Node - 21fa9 25.*.*.246 K8S node 18 6f1648805d9a6f7e5fd77e4af52430ff Power Trading System K8S Cluster Node - 430ff 25.*.*.177 K8S node 19 b10672fbe9ceea2c461390291008ad74 Power Trading System K8S Cluster Node - 8ad74 25.*.*.175 K8S node 20 b5a047a7ea4d5d9e4e2faebd765358d2 Electricity Trading System K8S-POD-358d2 192.168.*.24 POD 21 b8a78df799d70d12545cc8449beb0b6e Electricity Trading System K8S-POD-b0b6e 192.168.*.25 POD 22 b8d797ccef304aec14317b7231fcb5a9 Electricity Trading System K8S-POD-cb5a9 192.168.*.26 POD 23 bf6c5acd7ed7e0544192e6162f8dcc10 Electricity Trading System K8S-POD-dcc10 192.168.*.27 POD 24 c1a2dfab165368a1bdcac9eaadeec6b2 K8S-POD-ec6b2 Power Trading System 192.168.*.28 POD 25 b8a329t799fb8712545cc8449betu873 Electricity Trading System K8S-POD-tu873 192.168.*.29 POD 26 70E3A260-4B32-4CAD-B468-F5C9629101C2 Electricity Trading System - Market Member Inquiry Service - 9101C2 / microservices 27 70E3A260-4B32-4CAD-B468-F2A1A3D9D093 Electricity Trading System - Market Member Inquiry Service - 3D9D0 / microservices 28 73C8F20B-9090-4BC6-9AFD-18194EF42708 Electricity Trading System - Spot Trading Inquiry Service - 42708 / microservices 29 73C8F20B-9090-4BC6-9AFD-92AEEF3D084E Electricity Trading System - Spot Trading Inquiry Service - D084E / microservices 30 773AD6CC-6E3A-4534-A3E2-814FAF4E3E92 Electricity Trading System - Market Member Query Interface - 9101C2 192.168.*.24 interface 31 773AD6CC-6E3A-4534-A3E2-814FAF4E3E90 Electricity Trading System - Market Member Query Interface - 9101C2 192.168.*.25 interface 32 72D0999F-06AD-4666-A004-F2A1A3D9D093 Electricity Trading System - Spot Trading Query Interface - 9D093 192.168.*.26 interface 33 72D0999F-06AD-4666-A004-F2A1A3D9D094 Electricity Trading System - Spot Trading Query Interface - 9D094 192.168.*.27 interface 34 7518E78F-2A9C-4E36-9322-92AEEF3D0840 Electricity Trading System - Market Member Management Micro-application / Micro Applications 35 7518E78F-2A9C-4E36-9322-92AEEF3D0842 Electricity Trading System - Spot Trading Management Micro-application / Micro Applications 36 7C36AA4A-E659-47BE-9D9B-814FAF4E3E90 Electricity Trading System / Information systems Table 3. IT Resource Relationships of a State Grid Company Serial Number Source Resource ID Target Resource ID Relationship Remark 1 de8044a36267b242d27991a5f4b3d43d 5734D53F-FD7A-4CEF-B2E1-E78460E43E51 Running in ECS runs on servers 2 238299571258989aa60f017ba0d16c5b 5734D53F-FD7A-4CEF-B2E1-E78460E43E51 Running in ECS runs on servers 3 279cbd12e87a9473a671db5f839d398e 58E32BE8-5EF4-406A-A248-FF1DD7098DFE Running in ECS runs on servers 4 3eda543a9d5e33ef1d69cb1b7524b822 58E32BE8-5EF4-406A-A248-FF1DD7098DFE Running in ECS runs on servers 5 4d2e7369940f93f303a67c18dfc561cd 5E03E76F-5B59-43BB-AC8C-E22336EBB4A9 Running in ECS runs on servers 6 016f67889ad645849010c937bbf71294 5E03E76F-5B59-43BB-AC8C-E22336EBB4A9 Running in ECS runs on servers 7 a45ad49a3b817eec27eb5658a9709020 68B28433-1B2E-42A4-9FD5-B4D91364A269 Running in ECS runs on servers 8 000234fa80b0ed00368956c8f9921fa9 de8044a36267b242d27991a5f4b3d43d use K8S cluster nodes use ECS 9 6f1648805d9a6f7e5fd77e4af52430ff 279cbd12e87a9473a671db5f839d398e use K8S cluster nodes use ECS 10 b10672fbe9ceea2c461390291008ad74 4d2e7369940f93f303a67c18dfc561cd use K8S cluster nodes use ECS 11 645fcfbbd6690d85350a5ce9bc718d40 238299571258989aa60f017ba0d16c5b use RDS uses ECS 12 6c12ddcbcf0704f30266f902cceb2677 3eda543a9d5e33ef1d69cb1b7524b822 use RDS uses ECS 13 7359a4ce94362138b20050e4a54611d9 016f67889ad645849010c937bbf71294 use RDS uses ECS 14 7c6152fbd58ac86fcbc8b5bfb6ff1932 a45ad49a3b817eec27eb5658a9709020 use RDS uses ECS 15 000234fa80b0ed00368956c8f9921fa9 26fde48ad0f434b3e87cdb15faceb184 belong K8S cluster nodes belong to the K8S cluster 16 6f1648805d9a6f7e5fd77e4af52430ff 26fde48ad0f434b3e87cdb15faceb184 belong K8S cluster nodes belong to the K8S cluster 17 b10672fbe9ceea2c461390291008ad74 26fde48ad0f434b3e87cdb15faceb184 belong K8S cluster nodes belong to the K8S cluster 18 b5a047a7ea4d5d9e4e2faebd765358d2 6f1648805d9a6f7e5fd77e4af52430ff Running in POD runs on a K8S cluster node. 19 b8a78df799d70d12545cc8449beb0b6e 6f1648805d9a6f7e5fd77e4af52430ff Running in POD runs on a K8S cluster node. 20 b8d797ccef304aec14317b7231fcb5a9 b10672fbe9ceea2c461390291008ad74 Running in POD runs on a K8S cluster node. 21 bf6c5acd7ed7e0544192e6162f8dcc10 b10672fbe9ceea2c461390291008ad74 Running in POD runs on a K8S cluster node. 22 c1a2dfab165368a1bdcac9eaadeec6b2 000234fa80b0ed00368956c8f9921fa9 Running in POD runs on a K8S cluster node. 23 b8a329t799fb8712545cc8449betu873 000234fa80b0ed00368956c8f9921fa9 Running in POD runs on a K8S cluster node. 24 70E3A260-4B32-4CAD-B468-F5C9629101C2 b5a047a7ea4d5d9e4e2faebd765358d2 Deployed at Microservices deployed on POD 25 70E3A260-4B32-4CAD-B468-F2A1A3D9D093 b8a78df799d70d12545cc8449beb0b6e Deployed at Microservices deployed on POD 26 73C8F20B-9090-4BC6-9AFD-18194EF42708 c1a2dfab165368a1bdcac9eaadeec6b2 Deployed at Microservices deployed on POD 27 73C8F20B-9090-4BC6-9AFD-92AEEF3D084E bf6c5acd7ed7e0544192e6162f8dcc10 Deployed at Microservices deployed on POD 28 70E3A260-4B32-4CAD-B468-F5C9629101C2 773AD6CC-6E3A-4534-A3E2-814FAF4E3E92 supply Microservices provide interfaces 29 70E3A260-4B32-4CAD-B468-F2A1A3D9D093 773AD6CC-6E3A-4534-A3E2-814FAF4E3E90 supply Microservices provide interfaces 30 73C8F20B-9090-4BC6-9AFD-18194EF42708 72D0999F-06AD-4666-A004-F2A1A3D9D093 supply Microservices provide interfaces 31 73C8F20B-9090-4BC6-9AFD-92AEEF3D084E 72D0999F-06AD-4666-A004-F2A1A3D9D094 supply Microservices provide interfaces 32 7518E78F-2A9C-4E36-9322-92AEEF3D0840 70E3A260-4B32-4CAD-B468-F5C9629101C2 Include Micro applications include microservices 33 7518E78F-2A9C-4E36-9322-92AEEF3D0842 73C8F20B-9090-4BC6-9AFD-18194EF42708 Include Micro applications include microservices 34 7518E78F-2A9C-4E36-9322-92AEEF3D0842 73C8F20B-9090-4BC6-9AFD-92AEEF3D084E Include Micro applications include microservices 35 7518E78F-2A9C-4E36-9322-92AEEF3D0840 70E3A260-4B32-4CAD-B468-F2A1A3D9D093 Include Micro applications include microservices 36 7C36AA4A-E659-47BE-9D9B-814FAF4E3E90 7518E78F-2A9C-4E36-9322-92AEEF3D0840 Include Information systems include micro-applications 37 7C36AA4A-E659-47BE-9D9B-814FAF4E3E90 7518E78F-2A9C-4E36-9322-92AEEF3D0842 Include Information systems include micro-applications S2) When the maintenance plan selects [Jiangsu Company Jiangning Data Center A01 Rack PC Server-43E51], the maintenance impact scope analysis agent is invoked. The agent queries all related IT resources based on the maintenance object [Jiangsu Company Jiangning Data Center A01 Rack PC Server-43E51] to form an IT resource relationship topology diagram, as shown in 8. S3) Input the topology data of the IT resources associated with the server into the DeepSeek R1 large language model. Based on the topology relationship, the DeepSeek R1 large language model intelligently analyzes the impact range of the server maintenance and outputs the analysis report content, as shown below: The maintenance of PC server-43E51 in rack A01 of the Jiangning data center in Jiangsu Province will directly affect the [Power Trading System ECS-3d43d] and [Power Trading System ECS-16c5b] running on this server. ECS-3d43d runs a node of the Power Trading System Kubernetes cluster, which deploys the spot trading management micro-application and spot trading query service. This maintenance will cause these services to become unusable. ECS-16c5b runs the Power Trading System RDS database, which will cause database access failures. In summary, the maintenance of PC server-43E51 in rack A01 of the Jiangning data center in Jiangsu Province will have a significant impact on the Power Trading System. It is necessary to prepare an emergency response plan for the Power Trading System in advance. After the server maintenance, it is necessary to simultaneously verify whether the Power Trading System functions have returned to normal.

[0032] In summary, this invention models IT resources within the scope of State Grid's operations, constructs a comprehensive IT resource topology diagram, automatically collects IT resource ledger data and relational data, and utilizes the intelligent analysis capabilities of the large model to intelligently analyze the impact range of IT resource maintenance, prepares contingency plans in advance for maintenance-related faults, and reduces faults caused by insufficient maintenance considerations.

[0033] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.

Claims

1. An intelligent analysis method for the influence range of IT resource maintenance, characterized in that, The method comprises the following steps: According to the maintenance plan information, the IT resource RAG knowledge base is retrieved to obtain the IT resource information associated with the maintenance plan information; According to the IT resource information associated with the maintenance plan information, a large language model is called to analyze and obtain the IT resource maintenance influence range; The construction of the IT resource RAG knowledge base comprises the following steps: Modeling the IT resource to obtain an IT resource association relationship model; According to the IT resource association relationship model, IT resource data is collected to obtain IT resource account data and IT resource association relationship data; The IT resource account data and the IT resource association relationship data are vectorized to construct the IT resource RAG knowledge base.

2. The intelligent analysis method for IT resource maintenance influence range according to claim 1, characterized in that, According to the maintenance plan information, the IT resource RAG knowledge base is retrieved to obtain the IT resource information associated with the maintenance plan information, which comprises the following steps: Obtain the maintenance object information from the maintenance plan information; According to the maintenance object information, the IT resource RAG knowledge base is retrieved to obtain the IT resource information directly associated and indirectly associated with the maintenance object.

3. The intelligent analysis method for IT resource maintenance influence range according to claim 1, characterized in that, The large language model is DeepSeek R1 large language model, and according to the IT resource information associated with the maintenance plan information, the large language model is called to analyze and obtain the IT resource maintenance influence range, which comprises the following steps: According to the IT resource information directly associated and indirectly associated with the maintenance object, the DeepSeek R1 large language model is called to analyze and generate an IT resource maintenance influence range analysis report.

4. The intelligent analysis method for IT resource maintenance influence range according to claim 1, characterized in that, The IT resource comprises an information system layer, a basic platform layer and a physical facility layer; the IT resource is modeled to obtain an IT resource association relationship model, which comprises the following steps: A horizontal association relationship model of the IT resource of the information system layer, the basic platform layer and the physical facility layer is established respectively; A vertical association relationship model of the IT resource of the information system layer, the basic platform layer and the physical facility layer is established.

5. The IT resource maintenance influence range intelligent analysis method according to claim 4, characterized in that, The horizontal association relationship model of the IT resource of the information system layer, the basic platform layer and the physical facility layer comprises the following steps: The information system layer constructs an association relationship model among information systems, micro applications, micro services and interfaces, the basic platform layer constructs an association relationship model among cloud platforms, cloud platform components, cloud resource clusters and cloud resource instances, and the physical facility layer constructs an association relationship model among machine rooms, cabinets, servers, network devices, security devices, power supply devices and environmental control devices; The vertical association relationship model of the IT resource of the information system layer, the basic platform layer and the physical facility layer comprises the following steps: An association relationship model among micro services, PODs, ECSs and servers is constructed.

6. The intelligent analysis method for IT resource maintenance influence range according to claim 4, characterized in that, According to the IT resource association relationship model, the IT resource data is collected to obtain the IT resource account data and the IT resource association relationship data, which comprises the following steps: Using an APM tool to collect information system, micro application, micro service, micro service interface account data and association relationship data; Using a cloud platform management tool to collect cloud platform, cloud resource cluster, cloud resource instance account data and association relationship data; Using a State Grid unified agent to collect server, network device, security device, storage device account data and association relationship data; Using an environmental control management tool to collect power supply device, air conditioner, temperature and humidity sensor account data and association relationship data.

7. The IT resource maintenance influence range intelligent analysis method according to claim 1, characterized in that, Also comprising: The collected IT resource account data is stored in the account database, and the collected IT resource association data is stored in the association graph database.

8. An IT resource maintenance influence range intelligent analysis system, characterized in that, Comprising: The overhaul influence range analysis intelligent agent is used to retrieve the constructed IT resource RAG knowledge base according to the overhaul plan information, obtain the IT resource information associated with the overhaul plan information, call the large language model for analysis according to the IT resource information associated with the overhaul plan information, and obtain the IT resource overhaul influence range; The construction of the IT resource RAG knowledge base comprises: Modeling the IT resource to obtain an IT resource association relationship model; Collecting IT resource data according to the IT resource association relationship model to obtain IT resource account data and IT resource association relationship data; The IT resource account data and the IT resource association relationship data are vectorized to construct the IT resource RAG knowledge base.

9. The IT resource maintenance influence range intelligent analysis system according to claim 8, wherein, Also comprising: The overhaul influence range analysis API is used to receive the overhaul plan information.

Images