Network protocol information retrieval method and device, computer device, readable storage medium and program product

Abstract

The application relates to a network protocol information retrieval method and device, computer equipment, a computer readable storage medium and a computer program product, which can improve network protocol information retrieval efficiency. The method comprises the following steps: in response to a network protocol information retrieval request, a retrieval keyword is acquired; a node index and a node relationship index corresponding to an information knowledge graph are acquired; the information knowledge graph comprises a first node corresponding to network protocol information and an association relationship between the nodes, the node index records the first node associated with the keyword, and the node relationship index records the association relationship between the first node and other first nodes associated with the first node; the retrieval keyword and the node index are matched to determine a second node associated with the retrieval keyword, the third node and the node association relationship between the second node and the third node are determined according to the node relationship index; a node connectivity graph is generated according to the second node, the third node and the node association relationship, and a retrieval result is obtained.

Description

Technical Field

[0001] This application relates to the field of network information management technology, and in particular to a method, apparatus, computer equipment, computer-readable storage medium, and computer program product for retrieving network protocol information. Background Technology

[0002] As networks expand and the number of network devices increases, network management systems need to process massive amounts of network information, such as SNMP (Simple Network Management Protocol) messages containing status and performance data of network devices. Effectively storing, indexing, and retrieving this data is crucial for network health monitoring, fault diagnosis, and performance optimization. However, due to the sheer volume of data and the complexity of query requirements, traditional network protocol information retrieval methods suffer from low retrieval efficiency. Summary of the Invention

[0003] Therefore, it is necessary to provide a network protocol information retrieval method, apparatus, computer equipment, computer-readable storage medium, and computer program product that can improve the efficiency of network protocol information retrieval in response to the above-mentioned technical problems.

[0004] Firstly, this application provides a method for retrieving network protocol information, including:

[0005] In response to a network protocol information retrieval request, retrieve search keywords;

[0006] Obtain the node index and node relationship index corresponding to the information knowledge graph; the information knowledge graph includes a first node corresponding to each of multiple network protocol information, and the association relationship between multiple first nodes. The node index records the first node associated with each keyword, and the node relationship index records the association relationship between each first node and other associated first nodes.

[0007] The search keywords and the node index are matched, and a second node associated with the search keywords is determined based on the matching results. A third node associated with the second node and the node association relationship between the second node and the third node are determined according to the node relationship index.

[0008] A node connectivity graph is generated based on the second node, the third node, and the node association relationships, and search results are obtained based on the node connectivity graph.

[0009] In one embodiment, before obtaining the search keywords in response to the network protocol information retrieval request, the method further includes:

[0010] Extract the keywords corresponding to each of the relationships in the information knowledge graph;

[0011] For each keyword, identify the first nodes associated with the keyword, and merge the first nodes associated with the keyword to obtain a node list;

[0012] The node index is generated based on the node list corresponding to each of the multiple keywords.

[0013] In one embodiment, before obtaining the search keywords in response to the network protocol information retrieval request, the method further includes:

[0014] The information knowledge graph is determined by identifying multiple first nodes, and relation description information corresponding to each first node in the information knowledge graph is obtained; the relation description information records the node relationship between each first node and other first nodes.

[0015] For each first node, based on the relationship description information, determine the edges connecting the first node and the other first nodes and the type of each edge, and associate each type of edge with the first node as the first node relationship information corresponding to the first node;

[0016] The node relationship index is obtained based on the first node relationship information corresponding to each of the multiple first nodes.

[0017] In one embodiment, the step of obtaining search keywords in response to a network protocol information retrieval request includes:

[0018] In response to the received network protocol information retrieval request, a query pattern diagram corresponding to the network protocol information request is obtained, and one or more of the search keywords are extracted from the query pattern diagram;

[0019] And / or,

[0020] If the network protocol information retrieval request carries keywords, then the keywords carried in the network protocol information retrieval request shall be used as the retrieval keywords.

[0021] In one embodiment, matching the search keywords and the node index, and determining the second node associated with the search keywords based on the matching results, includes:

[0022] The first node matching the search keyword is determined based on the node index;

[0023] The search keywords are sorted in ascending order according to the number of first nodes matched by the keywords, and the order of each search keyword is obtained. The target first node corresponding to the search keyword whose order satisfies the order condition is determined.

[0024] When the search keywords are determined based on the query pattern graph, the first adjacency relationship corresponding to the target first node is determined from the preset adjacency relationship index file, and the target first node whose first adjacency relationship matches the second adjacency relationship of the target first node in the query pattern graph is retained;

[0025] Based on the retained first target node, a second node associated with the search keyword is obtained.

[0026] In one embodiment, the association between the plurality of first nodes includes edges between the plurality of first nodes; the adjacency index file is constructed through the following steps:

[0027] Determine the upper limit for the number of edges processed in a single operation;

[0028] Based on the upper limit of the number of edges, the multiple edges in the information knowledge graph are divided into at least one list of edge tuples; the number of edges contained in each list of edge tuples is less than or equal to the upper limit of the number of edges;

[0029] The at least one edge tuple list is processed sequentially to obtain an ordered edge tuple list corresponding to each edge tuple list; wherein, each processing step is used to sort the edges in the edge tuple list according to the first node.

[0030] The ordered edge tuple lists are merged to obtain the adjacency index file.

[0031] In one embodiment, determining the third node associated with the second node and the node association relationship between the second node and the third node based on the node relationship index includes:

[0032] Obtain the weight of the second node, adjust the parameters according to the preset number of nodes, and filter out the target second node that meets the weight condition from multiple second nodes for book search.

[0033] Obtain the second node relationship information corresponding to each target second node from the node relationship index, and determine the third node associated with the target second node and the node association relationship between the target second node and the third node based on the second node relationship information.

[0034] In one embodiment, step A includes: generating a node connectivity graph based on the second node, the third node, and the node association relationships, including:

[0035] If the execution space of the retrieval task is less than the space threshold for batch processing, the second node, the third node, and the node association relationship are stored in other storage spaces outside the execution space of the retrieval task.

[0036] The second node, the third node, and the node associations are read in batches from the other storage spaces and processed in the retrieval task execution space to generate the node connectivity graph.

[0037] Secondly, this application also provides a network protocol information retrieval device, comprising:

[0038] The retrieval request acquisition module is used to retrieve retrieval keywords in response to network protocol information retrieval requests;

[0039] An index building module is used to obtain the node index and node relationship index corresponding to the information knowledge graph; the information knowledge graph includes a first node corresponding to each of multiple network protocol information, and the association relationship between multiple first nodes; the node index records the first node associated with each keyword, and the node relationship index records the association relationship between each first node and other associated first nodes;

[0040] The retrieval execution module is used to match the retrieval keywords and the node index, determine the second node associated with the retrieval keywords based on the matching results, and determine the third node associated with the second node and the node association relationship between the second node and the third node according to the node relationship index.

[0041] The retrieval result generation module is used to generate a node connectivity graph based on the second node, the third node, and the node association relationship, and to obtain retrieval results based on the node connectivity graph.

[0042] Thirdly, this application also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to perform the following steps:

[0043] In response to a network protocol information retrieval request, retrieve search keywords;

[0044] Obtain the node index and node relationship index corresponding to the information knowledge graph; the information knowledge graph includes a first node corresponding to each of multiple network protocol information, and the association relationship between multiple first nodes. The node index records the first node associated with each keyword, and the node relationship index records the association relationship between each first node and other associated first nodes.

[0045] The search keywords and the node index are matched, and a second node associated with the search keywords is determined based on the matching results. A third node associated with the second node and the node association relationship between the second node and the third node are determined according to the node relationship index.

[0046] A node connectivity graph is generated based on the second node, the third node, and the node association relationships, and search results are obtained based on the node connectivity graph.

[0047] Fourthly, this application also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, performs the following steps:

[0048] In response to a network protocol information retrieval request, retrieve search keywords;

[0049] Obtain the node index and node relationship index corresponding to the information knowledge graph; the information knowledge graph includes a first node corresponding to each of multiple network protocol information, and the association relationship between multiple first nodes. The node index records the first node associated with each keyword, and the node relationship index records the association relationship between each first node and other associated first nodes.

[0050] The search keywords and the node index are matched, and a second node associated with the search keywords is determined based on the matching results. A third node associated with the second node and the node association relationship between the second node and the third node are determined according to the node relationship index.

[0051] A node connectivity graph is generated based on the second node, the third node, and the node association relationships, and search results are obtained based on the node connectivity graph.

[0052] Fifthly, this application also provides a computer program product, including a computer program that, when executed by a processor, performs the following steps:

[0053] In response to a network protocol information retrieval request, retrieve search keywords;

[0054] Obtain the node index and node relationship index corresponding to the information knowledge graph; the information knowledge graph includes a first node corresponding to each of multiple network protocol information, and the association relationship between multiple first nodes. The node index records the first node associated with each keyword, and the node relationship index records the association relationship between each first node and other associated first nodes.

[0055] The search keywords and the node index are matched, and a second node associated with the search keywords is determined based on the matching results. A third node associated with the second node and the node association relationship between the second node and the third node are determined according to the node relationship index.

[0056] A node connectivity graph is generated based on the second node, the third node, and the node association relationships, and search results are obtained based on the node connectivity graph.

[0057] In the aforementioned network protocol information retrieval method, in response to a network protocol information retrieval request, search keywords are obtained; the node index and node relationship index corresponding to the information knowledge graph are obtained; the information knowledge graph includes first nodes corresponding to multiple network protocol information and the association relationships between multiple first nodes; the node index records the first nodes associated with each keyword, and the node relationship index records the association relationships between each first node and other associated first nodes; the search keywords and node indexes are matched, and the second node associated with the search keywords is determined based on the matching results; the third node associated with the second node and the node association relationships between the second and third nodes are determined based on the node relationship index; a node connectivity graph is generated based on the second node, the third node, and the node association relationships, and the retrieval results are obtained based on the node connectivity graph. By matching search keywords with node indexes and further utilizing node relationship indexes to determine the relationships between nodes, efficient network protocol information retrieval can be achieved. This not only speeds up the retrieval process but also handles complex relationships, ensuring the integrity and accuracy of retrieved network protocol information data. Furthermore, by displaying the relationships between nodes using a node connectivity graph, the deep-seated relationships of network protocol information are visually presented, thus providing a more flexible and efficient data retrieval solution and significantly improving the efficiency of network protocol information retrieval. Attached Figure Description

[0058] To more clearly illustrate the technical solutions in the embodiments of this application or related technologies, the drawings used in the description of the embodiments of this application or related technologies will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0059] Figure 1 This is an application environment diagram of a network protocol information retrieval method in one embodiment;

[0060] Figure 2 This is a flowchart illustrating a network protocol information retrieval method in one embodiment;

[0061] Figure 3 This is a schematic diagram of the node index generation process in one embodiment;

[0062] Figure 4 This is a schematic diagram of the structure of an information knowledge graph in one embodiment;

[0063] Figure 5 This is a schematic diagram illustrating the process of generating a node relationship index in one embodiment;

[0064] Figure 6 This is a schematic diagram illustrating the construction process of an adjacency index file in one embodiment;

[0065] Figure 7 This is a flowchart illustrating a network protocol information retrieval method in another embodiment;

[0066] Figure 8 This is a structural block diagram of a network protocol information retrieval device in one embodiment;

[0067] Figure 9 This is an internal structural diagram of a computer device in one embodiment. Detailed Implementation

[0068] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0069] The network protocol information retrieval method provided in this application embodiment, in one embodiment, can be applied to, for example... Figure 1 The application environment shown. Figure 1 An exemplary network deployment diagram in a cluster environment is illustrated. A cluster environment typically includes the deployment of various devices / clients, a network management system (NMS), and a backup and disaster recovery system. In an enterprise's cluster network environment, one or more network devices 102a and 102b are connected to one or more servers 104a-104c. Servers 104a-104c contain various cloud services such as computing services, virtualization services, and storage services. These business servers construct the enterprise's productivity environment, and are connected via network devices 102a-102b to achieve management, monitoring, data backup, and disaster recovery.

[0070] The system may include multiple logically grouped servers 106. In some scenarios, the logical group of servers may be referred to as a customer's cluster environment. This environment includes multiple server clusters 104a-104c, each server cluster representing one or more enterprise cloud services.

[0071] For example, server 104a can be referred to as a file server, application server, web server, proxy server, or gateway server. In some cases, server 104a can act as an application server or main application server. In one possible configuration, server 104a may include Active Directory. In some embodiments, server 104a can act as a client node accessing server applications and provide other clients (104a-104n) with access to hosted applications. In some embodiments, server 104b provides cloud services for enterprise virtualization, including applications such as computing, containers, and desktops. In another configuration, enterprises use storage centers 104c to store and manage critical business data, and these storage centers use protocols such as SAN (Storage Area Network), DAS (Direct-Attached Storage), or NFS (Network File System) for data communication. In some embodiments, storage center 104c provides storage services including structured data, semi-structured data (JSON / XML), and unstructured objects (such as txt, doc, SNMP).

[0072] In one embodiment, backup and disaster recovery servers 101a-104c operate within the same physical host or network environment, thereby providing data monitoring and protection services. In some implementations, data monitoring and protection services are supported through agentless methods via interfaces of relevant applications on enterprise cloud services 104a-104c, such as RESTful and WebHook, providing HTTP (Hypertext Transfer Protocol) protocol interfaces.

[0073] The topology of network devices 102a and / or 102b can be bus, star, or ring network topology, and they are connected via network 103. In some possible implementations, network 103 includes any of the following: point-to-point network, broadcast network, wide area network, local area network, telecommunications network, data communication network, computer network, ATM (Asynchronous Transfer Mode) network, SONET (Synchronous Fiber Network), SDH (Synchronous Digital Hierarchy) network, wireless network, and wired network. In different application scenarios, network 103 may also include wireless links, such as infrared channels or satellite bands.

[0074] Figure 1 The diagram shows the Network Management System (NMS) 100, Backup and Disaster Recovery Systems 101a-101c, and Enterprise Servers 104a-104c located on different networks 102a and 102b, respectively. However, in actual deployment, these networks can be local area networks (LANs), metropolitan area networks (MANs), or wide area networks (WANs). In practice, network 102a can be a private network, and network 102b can be a public network; conversely, network 102b can be a private network, and network 102a can be a public network; alternatively, both networks 102a and 102b can be private networks, such as an enterprise's internal private network. In actual deployment, the NMS and backup service center can be located in a management center within the enterprise, communicating with the company's data center or a third-party cloud server 104 via a WAN (Wide Area Network), such as network 103, connected through network 102b.

[0075] In one exemplary embodiment, such as Figure 2 As shown, a network protocol information retrieval method is provided, which can be applied to... Figure 1 The following explanation uses the Network Management System (NMS) system 100 as an example, including the following steps S202 to S208. Wherein:

[0076] Step S202: In response to the network protocol information retrieval request, obtain the search keywords.

[0077] Among them, a network protocol information retrieval request is a request sent by a user or application to the system to obtain the required network protocol information; search keywords are keywords or phrases used to search and match network protocol information, which may include protocol names, device identifiers, message types, etc.

[0078] Optionally, when the Network Management System (NMS) receives a network protocol information retrieval request from a user or application, it begins to parse the request content and extracts the search keywords from the parsed content.

[0079] Step S204: Obtain the node index and node relationship index corresponding to the information knowledge graph; the information knowledge graph includes the first node corresponding to each of the multiple network protocol information, and the association relationship between the multiple first nodes. The node index records the first node associated with each keyword, and the node relationship index records the association relationship between each first node and other associated first nodes.

[0080] Information knowledge graphs, by organizing data into a graphical structure—nodes (entities) and edges (relationships between entities)—can provide in-depth data insights and complex query capabilities. Entities represent specific information items; an SNMP message can be considered a node. Node attributes can include various metadata about the message, such as timestamps, device IDs, and message types.

[0081] Edges represent relationships between nodes. In the context of SNMP messages, edges can represent various types of relationships, such as: message flow, where one message is a response to another, and an edge can be created to represent this request-response relationship; device relationship, where two messages come from the same or related devices (such as a router and a connected terminal), and similarity, where two messages have similar content or attributes (such as warnings of the same type), and an edge can be created to represent this similarity.

[0082] To distinguish it from other nodes in the following text, the nodes in the information knowledge graph will be referred to as the first node. The first node is the basic unit in the information knowledge graph, representing network protocol information, which can be an SNMP message from a network device.

[0083] The node index records the mapping relationship between keywords and first nodes, enabling the system to quickly locate the first node related to the keyword. The node relationship index records the association relationship between each first node and other related first nodes, supporting the querying and analysis of complex relationships.

[0084] Optionally, the network management system first initializes and maintains an information knowledge graph containing multiple first nodes. Keywords are associated with these first nodes to build a node index. The relationships between each first node and other first nodes are recorded to build a node relationship index. When a search is needed, the constructed node index and node relationship index are retrieved. The node index is used to quickly locate first nodes related to keywords, while the node relationship index is used to determine the relationships between these nodes, supporting complex queries.

[0085] Step S206: Match the search keywords and node indexes, determine the second node associated with the search keywords based on the matching results, and determine the third node associated with the second node and the node association relationship between the second node and the third node according to the node relationship index.

[0086] To distinguish other nodes in this application, nodes identified from the node index through keyword matching are designated as second nodes. These serve as intermediate search results and are further used to determine relationships, connecting search keywords with more in-depth network protocol information analysis. Nodes associated with the second nodes, determined based on the node relationship index, are designated as third nodes. These are part of the search results, further revealing the connections between network protocol information, expanding the search scope, and providing more comprehensive network protocol information. Node relationships are the relationships between nodes, which can include one or more types such as direct association and dependency. These relationships can be used to construct a node connectivity graph, where edges connect related nodes.

[0087] Optionally, when the Network Management System (NMS) receives a network protocol information retrieval request and extracts keywords, it matches the keywords in the node index. The matched second nodes can correspond to multiple related network protocol information items. Subsequently, using the node relationship index, it further searches for node associations between these second and third nodes, constructing a more complete network of keyword association information.

[0088] Step S208: Generate a node connectivity graph based on the second node, the third node, and the node association relationships, and obtain the search results based on the node connectivity graph.

[0089] A node connectivity graph is a graph composed of edges between nodes, representing nodes and their relationships. It provides an overall view that enables comprehensive analysis of network protocol information.

[0090] Optionally, after the Network Management System (NMS) receives a search request and identifies the relevant second and third nodes, it constructs a node connectivity graph of these nodes and their relationships. This graph displays the nodes related to the search request and their relationships. Based on the analysis of this graph, comprehensive search results are ultimately generated.

[0091] In this embodiment, in response to a network protocol information retrieval request, search keywords are obtained; the node index and node relationship index corresponding to the information knowledge graph are obtained; the information knowledge graph includes first nodes corresponding to multiple network protocol information and the association relationships between multiple first nodes. The node index records the first nodes associated with each keyword, and the node relationship index records the association relationships between each first node and other associated first nodes; the search keywords and node indexes are matched, and the second node associated with the search keywords is determined based on the matching results. According to the node relationship index, the third node associated with the second node and the node association relationship between the second and third nodes are determined; a node connectivity graph is generated based on the second node, the third node, and the node association relationship, and the retrieval results are obtained based on the node connectivity graph. This achieves efficient network protocol information retrieval by matching search keywords with node indexes and further using the node relationship index to determine the relationships between nodes. It not only speeds up the retrieval process but also handles complex association relationships, ensuring the integrity and accuracy of the retrieved network protocol information data. At the same time, by displaying the relationships between nodes using a node connectivity graph, the deep-level relationships of network protocol information are intuitively shown, thus providing a more flexible and efficient data retrieval solution and significantly improving the efficiency of network protocol information retrieval.

[0092] In one possible implementation, before obtaining the search keywords in response to a network protocol information retrieval request, the process includes: extracting keywords corresponding to each relationship in the information knowledge graph; for each keyword, determining each first node associated with the keyword, and merging the first nodes associated with the keyword to obtain a node list; and generating a node index based on the node lists corresponding to multiple keywords.

[0093] Specifically, the node index I-index is generated as follows: Figure 3 As shown, the network management system performs a sequential scan of the knowledge information graph G.

[0094] In one embodiment, SNMP messages are represented as a knowledge information graph, such as... Figure 4 As shown, the MIB (Management Information Base) in the message refers to the MIB file, which is an information database associated with various elements in the network, used to describe relevant device parameters. MIB files are written in the standard format of Abstract Syntax Notation One. HOST-RESOURCES-MIB (Network Resource Management Information Base) is also a standard MIB, used in the context of SNMP. Scutech-MIB provides a custom MIB management information base for backup appliances, defining specific message types.

[0095] Object Identifiers (OIDs) are an identity type used to uniquely identify SNMP messages within a network. For example, "1.3.6.1.2.1.25.2.3.1.6" is a specific OID defined in the HOST-RESOURCES-MIB, representing the used space of a storage device. This OID is a partial path; a complete OID will include the base path of the OID, plus one or more additional numbers used to describe specific storage device instances. "1.3.6.1.4.1.58021.1.1" is defined in the Scutech-MIB, indicating that the device is performing traffic balancing operations.

[0096] Node SNMP1 represents message 1, with attributes including timestamp, device ID A, message type CPU warning, content, MIB, OID, and value. Node NMP2 represents message 2, with attributes including timestamp, device ID B, message type traffic balancing operation, content, MIB, OID, and value. Node SNMP3 represents message 3, with attributes including timestamp, device ID A, message type memory warning, content, MIB, OID, and value.

[0097] In the information knowledge graph, edge 1 (vertical forward edge) connects node 1 and node 2, representing a message flow; that is, a CPU warning from node 1 leads to a traffic balancing operation on node 2. Vertical forward edges connect nodes of different levels or types, representing a relationship from a lower-level node to a higher-level node. Edge 3 (horizontal edge) connects node 1 and node 3, representing their similarity because they both originate from device A and both are resource utilization warnings. Horizontal edges connect nodes of the same level or type; in a sibling graph, this represents a sibling or brother relationship between nodes. The direction of this connection is defined based on the chronological order of occurrence. Edge 2 (vertical forward edge) connects node 2 and node 3, indicating a relationship between devices if the traffic balancing operation on device B is in response to multiple warnings from device A. This relationship needs to be determined through additional logic. Edge 4 (vertical backward edge) is used for post-mortem analysis, such as when the backup task slows down and it is necessary to understand the operation that caused the failure. The vertical backward edge is designed for the purpose of pre-defined analysis, rather than to represent the actual event flow or logic flow.

[0098] It should be noted that in actual model construction, the graph should be further refined based on the relationships between devices, the direction of message flow, and the context. In most practical applications, the relationships between events and messages are usually constructed in chronological order and causal logic in a forward direction. After scanning the graph G, for each edge... , Represents an edge. and This represents the two nodes that connect the edge. For edge labels, extract a series of keywords related to the edges. These keywords can be extracted from edge labels, attributes of nodes connected by the edges, or other edge-related information. In the context of SNMP messages, keywords can be determined using methods such as IRI (Information Retrieval Interface), predicates, and SPARQL (SPARQL Protocol and RDF Query Language) strings. Keywords can be message types, opcodes, or specific field values ​​contained in the message.

[0099] For the extracted keywords Generate two key-value pairs and And add these key-value pairs to the collection. In this way, each keyword is associated with its corresponding node, making it easy to quickly find relevant nodes using keywords.

[0100] Starting from set M1, construct set M2. In set M2, each key is a keyword, and each value is a list of UNI identifiers of all nodes containing that keyword. In this way, each keyword is associated with a UNI list containing all nodes that contain that keyword.

[0101] Sort each value (UNI list) in set M2 and remove duplicate elements. By sorting and deduplicating the UNI list, not only is storage space optimized, but query efficiency is also improved.

[0102] In this embodiment, by effectively extracting keywords corresponding to the relationships from the information knowledge graph and generating a complete node index, not only is the efficiency and accuracy of the system in processing network protocol information improved, but the comprehensiveness and relevance of the search results are also ensured.

[0103] In one embodiment, before obtaining search keywords in response to a network protocol information retrieval request, the method further includes: determining multiple first nodes in an information knowledge graph, and obtaining relation description information corresponding to each first node in the information knowledge graph; the relation description information records the node relationships between each first node and other first nodes; for each first node, based on the relation description information, determining the edges connecting the first node to other first nodes and the type of each edge, and associating each type of edge with the first node as the first node relation information corresponding to the first node; and obtaining a node relation index based on the first node relation information corresponding to each of the multiple first nodes.

[0104] Specifically, the steps for generating the node relationship index M-index are as follows: Figure 5 The network management system initializes the set M to contain an initial node UNI value and three empty sets. Each node can be considered isolated, without any associated edges, and can be represented as follows: Scan the set of RFDS (Resource Framework Data Set) triples. The result is split into a set U of nodes (UNI) without vertical back edges and a set of RDFS triples. Nodes without vertical back edges refer to those nodes in the graph that have no edges pointing to more general concepts or parent nodes. These nodes can be considered "leaf nodes" in the graph. In the context of a knowledge graph, this means that the concepts represented by these nodes have no further generalization or higher-level classification; they are at the very bottom of the information hierarchy in the graph. The purpose of this step is to read the contents of the RFDS dataset and prepare it for further processing. Each element in the list is processed in a loop until the loop continues. Empty. Specifically, for each triple, determine the relationships between its associated nodes and edges, and obtain the edge set from graph G. and ,in, It contains all the UNI corresponding nodes in U. Vertical forward edges in the graph connect nodes of different levels or types; It contains all the UNI corresponding nodes in U. The horizontal edge in the diagram connects nodes of the same level or type. and Edge data is removed from graph G and added to the corresponding quadruples of nodes in M, updating the processing result. After each iteration, the result is updated. and U. Updated to the processed set of triples, the updated set An RDFS triple containing all nodes whose UNI values ​​have never been placed in set U is used to update set U to a set containing the new node relationships. The UNI used as the target node, i.e. The updated set U and set Re-iterate, satisfying set U and set After all sets are empty, the loop is exited. This step ensures that the relationships between all nodes and edges are fully processed and recorded. Finally, the updated set M is hashed to generate the final node relationship index M-Index.

[0105] In this embodiment, the node relationships in the information knowledge graph are fully captured and recorded, generating an efficient node relationship index. This not only improves the system's query and analysis efficiency but also ensures the accuracy and comprehensiveness of information processing, providing a solid foundation for subsequent retrieval and analysis.

[0106] In one embodiment, in response to a network protocol information retrieval request, obtaining search keywords includes: in response to a received network protocol information retrieval request, obtaining a query pattern diagram corresponding to the network protocol information request, and extracting one or more search keywords from the query pattern diagram; and / or, if the network protocol information retrieval request carries keywords, then using the keywords carried in the network protocol information retrieval request as search keywords.

[0107] The query pattern graph is a structured graph that represents the user's query intent. It can include several nodes and edges, enabling the system to more accurately interpret the user's search requirements. Search keywords are keywords or phrases used to search and match network protocol information. They are extracted from the query pattern graph or obtained directly from the search request, guiding the system in information retrieval.

[0108] Optionally, upon receiving a network protocol information retrieval request, a query pattern graph is constructed based on the retrieval request, the nodes and edges in the query pattern graph are parsed, and one or more search keywords are extracted.

[0109] Optionally, upon receiving a network protocol information retrieval request, which carries search keywords such as "memory warning" or "traffic balancing operation," these keywords can be used directly as search keywords.

[0110] In this embodiment, by extracting keywords from the query pattern diagram or directly using keywords carried in the network protocol information request, it is possible to flexibly respond to different types of search requests, provide comprehensive and accurate search results, and ensure the accuracy and efficiency of the search.

[0111] In some embodiments, matching search keywords and node indexes, and determining a second node associated with the search keywords based on the matching results, includes: determining a first node matching the search keywords based on the node index; sorting multiple search keywords in ascending order according to the number of first nodes matching the keywords to obtain the order of each search keyword, and determining the target first node corresponding to the search keywords whose order satisfies the order condition; when the search keywords are determined based on the query pattern graph, determining the first adjacency relationship corresponding to the target first node from a preset adjacency relationship index file, and retaining the target first node whose first adjacency relationship matches the second adjacency relationship of the target first node in the query pattern graph; and obtaining the second node associated with the search keywords based on the retained target first nodes.

[0112] Optionally, a network protocol information retrieval request is received, search keywords are extracted, and the keywords are matched according to the node index, which can return at least one first node for each keyword. Multiple search keywords are sorted in ascending order based on the number of first nodes matched by each search keyword, resulting in the order of each keyword. Based on the order condition, the target first node corresponding to each search keyword is determined, with keywords having fewer matched first nodes being processed first. Matching and retention of target first nodes are then performed. When the search keywords are determined based on the query pattern graph, the first adjacency relationships corresponding to the target first nodes are determined from a preset adjacency relationship index file, and these adjacency relationships are matched with the second adjacency relationships of the target first nodes in the query pattern graph, retaining the target first nodes that meet the conditions. Based on the retained target first nodes, the second nodes associated with the search keywords are determined.

[0113] For example, if keyword X matches F first nodes and keyword Y matches E first nodes, and F is greater than E, the first nodes related to the search keyword Y are identified as target first nodes. The target first node matching keyword Y is then retained if it has a matching adjacency relationship with other nodes in the query pattern graph, based on the adjacency index file. Based on the retained target first nodes matching keyword Y, second nodes related to keyword Y can be determined.

[0114] In this embodiment, matching search keywords using node indexes enables the rapid identification of the first node associated with the keyword, improving retrieval efficiency. Sorting and filtering multiple search keywords ensures the accuracy and relevance of the search results. Utilizing an adjacency index file allows for deeper node relationship matching, further identifying the second node associated with the search keyword, enhancing the depth and breadth of information retrieval.

[0115] In one embodiment, the association between multiple first nodes includes edges between multiple first nodes; the adjacency index file is constructed through the following steps: determining the upper limit of the number of edges in a single processing; dividing multiple edges in the information knowledge graph into at least one list of edge tuples according to the upper limit of the number of edges; the number of edges contained in each list of edge tuples is less than or equal to the upper limit of the number of edges; processing at least one list of edge tuples in sequence to obtain an ordered list of edge tuples corresponding to each list of edge tuples; wherein, each processing is used to sort the edges in the list of edge tuples according to the first node; merging the ordered lists of edge tuples to obtain the adjacency index file.

[0116] Among them, the edge tuple list is a data structure used to store edges in the information knowledge graph. Multiple edges are divided into multiple edge tuple lists according to a predetermined upper limit of the number of edges, which facilitates batch processing and sorting of a large number of edges and improves processing efficiency.

[0117] An ordered list of edge tuples is a sorted list of edge tuples. The edges in each edge tuple are ordered to ensure that they are arranged in a specific order, which helps with subsequent merging and index building.

[0118] The adjacency index file records the index of edges between multiple first nodes. The ordered list of edge tuples is merged to generate the final adjacency index file.

[0119] Optionally, firstly, SNMP messages in the network are collected and converted into a knowledge graph. SNMP messages in the knowledge graph are treated as nodes, and relationships between nodes are represented by edges, where each edge represents the sending and receiving relationship of an SNMP message. Based on the number of edges, the edge data in the knowledge graph is divided into multiple edge tuple lists, with the number of edges in each tuple list not exceeding a preset upper limit. The edges in each edge tuple list are then sorted, based on either the UNI value (unique identifier) ​​of the starting or ending node, thus generating an ordered list of edge tuples. The sorted ordered list of edge tuples is stored in external storage to provide data support for subsequent merging processing. Finally, the ordered list of edge tuples stored in external storage is merged into a complete ordered edge list, resulting in an adjacency index file.

[0120] For example, an adjacency index file is generated, such as Figure 3 As shown, SNMP messages in the network are collected and converted into a knowledge information graph G. Graph G contains... Strip edge System memory can at most maintain The edge data of G is divided into... In the list of edge tuples, that is For each list of edge tuples Sort them as .exist In graph G, edges originating from the same node are stored sequentially in the adjacent data storage locations, forming an adjacency list. For each node in graph G, the edges emanating from it are arranged in order in the list for easier subsequent processing. This results in an ordered list of edge tuples. Then, it is stored on disk, ensuring efficient management and access to large amounts of edge data even under memory constraints. The ordered list is then processed using an external storage merge method. Merge into the same file In the (adjacency index file). This is a disk adjacency list containing information about the edges in graph G. This edge information is ordered for easy access and processing. Finally, each node in graph G is assigned a UNI value. The read file pointer value is used as the UNI value of node V. This value is unique to each node, and all UNI values ​​have the same length and occupy the same storage space. This means that each SNMP message (node) will obtain a unique identifier, and the node's UNI can be used to directly jump to the node's outgoing edge data, thereby improving the efficiency of graph data processing.

[0121] In this embodiment, by effectively managing and processing a large amount of edge data, an efficient adjacency index file is constructed, which enables quick querying of the relationships between nodes. This not only improves the efficiency of retrieval processing but also ensures the accuracy and completeness of the query.

[0122] In one embodiment, determining the third node associated with the second node and the node association relationship between the second node and the third node based on the node relationship index includes: obtaining the weight of the second node, adjusting the parameters according to the preset number of nodes, and filtering out the target second node that meets the weight condition from multiple second nodes; obtaining the second node relationship information corresponding to each target second node from the node relationship index, and determining the third node associated with the target second node and the node association relationship between the target second node and the third node based on the second node relationship information.

[0123] Optionally, when determining the third nodes associated with the second node and their relationships, the weight of each second node is first calculated. Based on preset indicators, such as node connectivity or information importance, the weight value of each second node is calculated. Simultaneously, to control the number of nodes filtered, the system applies preset node number adjustment parameters to filter out target second nodes whose weights meet the criteria.

[0124] After identifying the target second nodes, the relationship information for each target second node is retrieved from the node relationship index. This information includes the connection information between the target second node and other nodes, as well as their relationship types. Based on this extracted relationship information, the third nodes associated with each target second node and their relationships can be determined.

[0125] In this embodiment, by calculating the weight of the second node and filtering out the target second node that meets the conditions, the relevance and importance of the selected node can be ensured, and the accuracy of the retrieval results can be improved. At the same time, by using the node relationship index, the third node associated with the target second node and the relationship between them can be efficiently determined, thereby improving the depth and breadth of information retrieval and analysis.

[0126] In one embodiment, generating a node connectivity graph based on the second node, the third node, and node associations includes: if the retrieval task execution space is less than the space threshold for batch processing, storing the second node, the third node, and node associations in other storage spaces outside the retrieval task execution space; and reading the second node, the third node, and node associations in batches from other storage spaces into the retrieval task execution space for processing to generate a node connectivity graph.

[0127] Optionally, when generating the node connectivity graph, it is determined whether the current retrieval task execution space is sufficient to handle all second nodes, third nodes, and their relationships. If it is found that the retrieval task execution space is less than the space threshold for batch processing, these nodes and their relationships are stored in other storage spaces outside the retrieval task execution space. These other storage spaces can be external storage devices such as hard disks or distributed storage systems.

[0128] Data is read in batches from external storage, and second and third nodes, along with their relationships, are gradually loaded into the retrieval task execution space for processing. After processing the nodes and relationships of the current batch, a node connectivity graph is gradually constructed. After each batch processing is complete, the corresponding memory space is released to prepare for reading and processing the next batch of data.

[0129] In this embodiment, by adopting a time-for-space strategy of storing data exceeding the execution space of the retrieval task in other storage spaces and reading and processing nodes and their relationships in batches, large-scale data can be effectively managed, memory overflow can be avoided, system stability can be improved, the generation efficiency and accuracy of the node connectivity graph can be ensured, it can adapt to datasets of different sizes and different memory capacities, it is suitable for various computing environments, and the overall performance of data retrieval is improved.

[0130] To enable those skilled in the art to better understand the above steps, the following example illustrates the embodiments of this application, but it should be understood that the embodiments of this application are not limited thereto.

[0131] Specifically, such as Figure 6 As shown, upon receiving a network protocol information retrieval request, the request is first parsed to extract one or more search keywords. Keywords can be obtained directly from the request or by parsing the query pattern diagram determined by the network protocol information retrieval request. These keywords can be SNMP OIDs, device IDs, message types, etc., and will be used to find relevant nodes in the I-Index.

[0132] An I-Index is a hashed index structure that records the UNI set of each keyword and its associated nodes. By querying the I-Index, the system can quickly locate the nodes related to the keyword and obtain their UNI values. The specific process includes querying the related nodes for each keyword in the I-Index; sorting the keywords according to the number of nodes to optimize the processing order; selecting the relevant nodes according to preset parameters; and storing the filtered node UNIs into a collection.

[0133] Based on the node UNI set matched by I-index, the system returns a result set containing detailed node information and relationships by querying the M-index. Target nodes that meet the criteria are filtered according to node weights and preset parameters to ensure the relevance and accuracy of the result set.

[0134] Finally, a connectivity graph CS is constructed based on the result set returned by the M-Index. The connectivity graph CS displays all relevant nodes and their relationships, providing a complete view of network protocol information. Simultaneously, if the retrieval task execution space is insufficient, the system will store the excess nodes and relationships in other storage spaces, and read and process this data in batches to determine the retrieval results.

[0135] For example, given a query pattern graph Q:

[0136] Extract the keyword set K, and use the I-Index to match nodes. , yes The hash value, then let Search for the set of keywords in the I-Index Each keyword and knowledge information graph Related nodes. According to... Sort Key Set For elements in the search, keywords with fewer related nodes are prioritized for processing, reducing the amount of data to be processed in subsequent steps. When memory space is limited during retrieval, users can set input parameters. (Maximum number of nodes), selectively matching some or all of the I-Index with the key set. The relevant node UNIs are imported into memory. Based on the adjacency index file, the imported node UNIs are compared with the input query pattern graph. The relationships between them are determined, and the query pattern graph is discarded. Unrelated nodes UNI are selected, and the retained node UNIs are stored in the node UNI set. middle.

[0137] Based on the obtained node UNI set Preset node quantity adjustment parameters ,when Eδ * Lrd ∈[ Min , Max ] At that time, the size of the result set ;when < hour, ;when > hour, . (Expansion factor < 1) is an important parameter for adjusting the size of the result set R. It can be defined by user input, system performance (actual response required, system load, etc.), or based on the task's complexity. When performing general queries that do not require a large amount of historical data, this parameter can be set to... =0.2; This can be set to 0.2 when high-quality results are required and query time can be sacrificed. =0.5 or greater. This determines the upper and lower limits of query quality.

[0138] Weighting rules are defined based on the characteristics of SNMP messages (such as timestamps, device IDs, message types, etc.) to assign weights to the UNI nodes in the set 𝐿𝑟d. In some embodiments, recent messages and special types of warnings (CPU alarms, memory alarms, etc.) may have higher weights. (From the set of UNI nodes) Select the one with the highest weight Each node UNI is added to the set. middle.

[0139] Traversing a collection The node relationship index of the middle node, when there is insufficient memory space to store the query pattern graph. Connected graph At that time, the collection will be processed in batches using disk space. The UNI node in the graph forms the query pattern diagram. Construct a connected graph Based on the obtained connected graph This allows us to determine the results of network protocol information retrieval.

[0140] In this embodiment, the system can efficiently extract keywords from network protocol information retrieval requests. By using both I-Index and M-Index, two lossless indexes, it can achieve fast indexing and querying of the SNMP information knowledge graph G, ensuring data integrity and query accuracy. This ensures that each returned result set R is a matching subgraph of the query pattern graph Q in the information knowledge graph G. This is achieved through index optimization and query parameter settings. Strategies such as minimizing the impact on user query response time are employed to reduce query latency while ensuring data integrity and accuracy, ultimately generating a comprehensive and accurate connectivity graph. It provides efficient and reliable network protocol information retrieval and analysis services.

[0141] It should be understood that although the steps in the flowcharts of the embodiments described above are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the embodiments described above may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages of other steps.

[0142] Based on the same inventive concept, this application also provides a network protocol information retrieval device for implementing the network protocol information retrieval method described above. The solution provided by this device is similar to the implementation described in the above method; therefore, the specific limitations in one or more network protocol information retrieval device embodiments provided below can be found in the limitations of the network protocol information retrieval method described above, and will not be repeated here.

[0143] In one exemplary embodiment, such as Figure 7As shown, a network protocol information retrieval device is provided, comprising: a retrieval request acquisition module 810, an index building module 820, a retrieval execution module 830, and a retrieval result generation module 840, wherein:

[0144] The retrieval request acquisition module 810 is used to obtain retrieval keywords in response to network protocol information retrieval requests.

[0145] The index building module 820 is used to obtain the node index and node relationship index corresponding to the information knowledge graph. The information knowledge graph includes the first node corresponding to each of the multiple network protocol information, as well as the association relationship between the multiple first nodes. The node index records the first node associated with each keyword, and the node relationship index records the association relationship between each first node and other associated first nodes.

[0146] The retrieval execution module 830 is used to match retrieval keywords and node indexes, determine the second node associated with the retrieval keywords based on the matching results, and determine the third node associated with the second node and the node association relationship between the second node and the third node according to the node relationship index.

[0147] The retrieval result generation module 840 is used to generate a node connectivity graph based on the second node, the third node, and the node association relationship, and obtain retrieval results based on the node connectivity graph.

[0148] In one embodiment, the index building module 820 is configured to:

[0149] Extract keywords corresponding to each relationship in the information knowledge graph; for each keyword, determine the first nodes associated with the keyword, and merge the first nodes associated with the keyword to obtain a node list; generate a node index based on the node lists corresponding to multiple keywords.

[0150] In one embodiment, the index building module 820 is configured to:

[0151] The process involves identifying multiple first nodes in an information knowledge graph and obtaining relational description information corresponding to each first node. The relational description information records the node relationships between each first node and other first nodes. For each first node, based on the relational description information, the process determines the edges connecting the first node to other first nodes and the type of each edge, and associates each type of edge with the first node as the first node relational information corresponding to the first node. Based on the first node relational information corresponding to each of the multiple first nodes, a node relation index is obtained.

[0152] In one embodiment, the retrieval request acquisition module 810 is configured to: in response to a received network protocol information retrieval request, acquire a query pattern diagram corresponding to the network protocol information request, and extract one or more retrieval keywords from the query pattern diagram; and / or, if the network protocol information retrieval request carries keywords, use the keywords carried in the network protocol information retrieval request as retrieval keywords.

[0153] In one embodiment, the retrieval execution module 830 is configured to: determine an upper limit for the number of edges processed in a single operation; divide multiple edges in the information knowledge graph into at least one list of edge tuples according to the upper limit for the number of edges; ensure that the number of edges contained in each list of edge tuples is less than or equal to the upper limit for the number of edges; process at least one list of edge tuples sequentially to obtain an ordered list of edge tuples corresponding to each list of edge tuples; wherein each operation is configured to sort the edges in the list of edge tuples according to the first node; and merge the ordered lists of edge tuples to obtain an adjacency index file.

[0154] In one embodiment, the retrieval execution module 830 is configured to:

[0155] The first node matching the search keyword is determined based on the node index; multiple search keywords are sorted in ascending order according to the number of first nodes matching the keyword to obtain the order of each search keyword, and the target first node corresponding to the search keyword whose order satisfies the order condition is determined; when the search keyword is determined based on the query pattern graph, the first adjacency relationship corresponding to the target first node is determined from the preset adjacency relationship index file, and the target first node that matches the first adjacency relationship with the second adjacency relationship of the target first node in the query pattern graph is retained; based on the retained target first node, the second node associated with the search keyword is obtained.

[0156] In one embodiment, the retrieval execution module 830 is configured to:

[0157] Obtain the weight of the second node and adjust the parameters according to the preset number of nodes to filter out the target second node that meets the weight condition from multiple second nodes; obtain the second node relationship information corresponding to each target second node from the node relationship index, and determine the third node associated with the target second node and the node relationship between the target second node and the third node based on the second node relationship information.

[0158] In one embodiment, the retrieval result generation module 840 is configured to:

[0159] If the execution space of the retrieval task is less than the space threshold for batch processing, the second node, the third node, and the node association relationship are stored in other storage spaces outside the execution space of the retrieval task. The second node, the third node, and the node association relationship are read in batches from other storage spaces and processed in the execution space of the retrieval task to generate a node connectivity graph.

[0160] Each module in the aforementioned network protocol information retrieval device can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in the processor of a computer device in hardware form or independent of it, or stored in the memory of the computer device in software form, so that the processor can call and execute the operations corresponding to each module.

[0161] In one exemplary embodiment, a computer device is provided, which may be a server, and its internal structure diagram may be as follows: Figure 8 As shown, this computer device includes a processor, memory, input / output interfaces (I / O), and a communication interface. The processor, memory, and I / O interfaces are connected via a system bus, and the communication interface is also connected to the system bus via the I / O interfaces. The processor provides computational and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and a database. The internal memory provides the environment for the operating system and computer programs stored in the non-volatile storage media. The database stores node indexes and node relationship indexes. The I / O interfaces are used for exchanging information between the processor and external devices. The communication interface is used for communicating with external terminals via a network connection. When executed by the processor, the computer program implements a network protocol information retrieval method.

[0162] Those skilled in the art will understand that Figure 8 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0163] In one embodiment, a computer device is provided, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps in the above-described method embodiments.

[0164] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon, which, when executed by a processor, implements the steps in the above method embodiments.

[0165] In one embodiment, a computer program product is provided, including a computer program that, when executed by a processor, implements the steps in the above method embodiments.

[0166] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of the relevant data must comply with relevant regulations.

[0167] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile memory and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM). The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, artificial intelligence (AI) processors, etc., and are not limited to these.

[0168] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this application.

[0169] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this application should be determined by the appended claims.

Claims

1. A method for retrieving network protocol information, characterized in that, The method includes: In response to a network protocol information retrieval request, retrieve search keywords; Obtain the node index and node relationship index corresponding to the information knowledge graph; the information knowledge graph includes a first node corresponding to each of multiple network protocol information, and the association relationship between multiple first nodes. The node index records the first node associated with each keyword, and the node relationship index records the association relationship between each first node and other associated first nodes. The first node matching the search keyword is determined based on the node index; The search keywords are sorted in ascending order according to the number of first nodes matched by the search keywords to obtain the order of each search keyword, and the target first node corresponding to the search keyword whose order satisfies the order condition is determined. When the search keywords are determined based on the query pattern graph, the first adjacency relationship corresponding to the target first node is determined from the preset adjacency relationship index file, and the target first node whose first adjacency relationship matches the second adjacency relationship of the target first node in the query pattern graph is retained; Based on the retained first target node, a second node associated with the search keyword is obtained. Obtain the weight of the second node, adjust the parameters according to the preset number of nodes, and filter out the target second node that meets the weight condition from multiple second nodes; Obtain the second node relationship information corresponding to each target second node from the node relationship index, and determine the third node associated with the target second node and the node association relationship between the target second node and the third node based on the second node relationship information; A node connectivity graph is generated based on the second node, the third node, and the node association relationships, and search results are obtained based on the node connectivity graph.

2. The method according to claim 1, characterized in that, Before obtaining search keywords in response to a network protocol information retrieval request, the method further includes: Extract the keywords corresponding to each of the relationships in the information knowledge graph; For each keyword, identify the first nodes associated with the keyword, and merge the first nodes associated with the keyword to obtain a node list; The node index is generated based on the node list corresponding to each of the multiple keywords.

3. The method according to claim 1, characterized in that, Before obtaining search keywords in response to a network protocol information retrieval request, the method further includes: The information knowledge graph is determined by identifying multiple first nodes, and relation description information corresponding to each first node in the information knowledge graph is obtained; the relation description information records the node relationship between each first node and other first nodes. For each first node, based on the relationship description information, determine the edges connecting the first node and the other first nodes and the type of each edge, and associate each type of edge with the first node as the first node relationship information corresponding to the first node; The node relationship index is obtained based on the first node relationship information corresponding to each of the multiple first nodes.

4. The method according to claim 1, characterized in that, The process of obtaining search keywords in response to a network protocol information retrieval request includes: In response to the received network protocol information retrieval request, a query pattern diagram corresponding to the network protocol information request is obtained, and one or more search keywords are extracted from the query pattern diagram; And / or, If the network protocol information retrieval request carries keywords, then the keywords carried in the network protocol information retrieval request shall be used as the retrieval keywords.

5. The method according to claim 1, characterized in that, The association relationships between multiple first nodes include edges between multiple first nodes; the adjacency index file is constructed through the following steps: Determine the upper limit for the number of edges processed in a single operation; Based on the upper limit of the number of edges, the multiple edges in the information knowledge graph are divided into at least one list of edge tuples; the number of edges contained in each list of edge tuples is less than or equal to the upper limit of the number of edges; The at least one edge tuple list is processed sequentially to obtain an ordered edge tuple list corresponding to each edge tuple list; wherein, each processing step is used to sort the edges in the edge tuple list according to the first node. The ordered edge tuple lists are merged to obtain the adjacency index file.

6. The method according to claim 1, characterized in that, The step of generating a node connectivity graph based on the second node, the third node, and the node association relationships includes: If the execution space of the retrieval task is less than the space threshold for batch processing, the second node, the third node, and the node association relationship are stored in other storage spaces outside the execution space of the retrieval task. The second node, the third node, and the node associations are read in batches from the other storage spaces and processed in the retrieval task execution space to generate the node connectivity graph.

7. A network protocol information retrieval device, characterized in that, The device includes: The retrieval request acquisition module is used to retrieve retrieval keywords in response to network protocol information retrieval requests; An index building module is used to obtain the node index and node relationship index corresponding to the information knowledge graph; the information knowledge graph includes a first node corresponding to each of multiple network protocol information, and the association relationship between multiple first nodes; the node index records the first node associated with each keyword, and the node relationship index records the association relationship between each first node and other associated first nodes; The retrieval execution module is used to match the retrieval keywords and the node index, determine the second node associated with the retrieval keywords based on the matching results, and determine the third node associated with the second node and the node association relationship between the second node and the third node according to the node relationship index. The retrieval result generation module is used to generate a node connectivity graph based on the second node, the third node, and the node association relationship, and to obtain retrieval results based on the node connectivity graph.

8. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 6.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 6.

10. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 6.

Images