A network fault tracing method and system

By building complex networks and calculating connectivity, and automatically judging the fault period and root fault nodes, the problem of inaccurate network failure analysis in the existing technology is solved, and the accuracy and applicability of fault traceability in complex network environments is improved.

CN116232850BActive Publication Date: 2024-11-12SINA TECH (CHINA) CO LTD
View PDF 5 Cites 0 Cited by

Patent Information

Application Number
CN202211664502.5
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2022-12-23
Publication Date
2024-11-12
Estimated Expiration
2042-12-23

AI Technical Summary

Technical Problem

The existing technology has problems with low accuracy in network failure analysis and location, especially in complex network environments, where it is impossible to quickly and accurately determine the root cause of the failure, and rely on human settings and historical data to cope with architectural changes.

Method used

By summarizing full-link logs, building complex networks, calculating network connectivity, automatically judging fault periods and root fault nodes, using complex network technology to trace faults, reducing dependence on historical data, improving applicability, and visualizing fault development and scope of impact .

Benefits of technology

The accuracy of fault propagation analysis and cascading fault root cause analysis in complex network environments is achieved, reducing human intervention, and the results are more objective, improving the accuracy and applicability of fault tracing.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN116232850B_ABST
    Figure CN116232850B_ABST
Patent Text Reader

Abstract

The embodiment of the present invention provides a network fault tracing method, including: summarizing the full-link logs in the network to be detected; constructing a complex network based on the full-link logs; when there are faulty nodes in the complex network, determining the network connectivity of the complex network; when the network connectivity is lower than a preset threshold, determining the fault period, and determining the root fault node within the fault period. This technical solution applies complex network technology to network fault tracing, simplifies the network by using the relationship between the in-and-out degrees of complex network nodes, and uses the connected path to trace the fault link, so that the entire network is judged according to a reliable physical connection, effectively ensuring the accuracy of the fault tracing results.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of network technology, and in particular to a network fault tracing method and system. Background Art

[0002] With the development of Internet technology, the composition of Internet websites is becoming more and more complex. The structure of the website's internal structure layer, physical layer, transmission layer, data layer, and application layer has become very complex, and there is a strong coupling relationship between the various layers in the website. When a single point in the system fails, it will affect other nodes, and then the fault propagation, cascading failure, and even the paralysis of the entire system will occur. Traditional fault analysis methods mainly include fault location based on fixed rules, fault location based on fault alarm frequency, manual fault location, reliable fault analysis, and fault root cause location.

[0003] In the process of implementing the present invention, the inventors found that there are at least the following problems in the prior art:

[0004] The current fault location method is based on fixed rules and thresholds to determine the root cause of the fault, which has limitations. The network environment is complex, and a single fault will hit multiple rules. At the same time, the existing method requires manual rule setting, and the judgment result is highly consistent with the manual setting, which is limited to the setting personnel's understanding of the overall architecture. In addition, the existing technical means only target single fault sources and simple scenarios, and cannot quickly detect faults in complex networks. All these factors have led to the inability of existing technologies to efficiently and accurately determine the root cause of the fault. Therefore, how to improve the accuracy of network fault analysis and location is a problem that needs to be solved. Summary of the invention

[0005] The embodiment of the present invention provides a network fault tracing method and system, which are used to solve the problem of low accuracy of network fault analysis and location in the prior art.

[0006] To achieve the above-mentioned purpose, on the one hand, an embodiment of the present invention provides a network fault tracing method, including: summarizing the full-link log in the network to be detected, the full-link log including a set of log records of all source nodes accessing the destination node; constructing a complex network based on the full-link log; when there is a faulty node in the complex network, determining the network connectivity of the complex network; when the network connectivity is lower than a preset threshold, determining the fault period, and determining the root fault node within the fault period.

[0007] On the other hand, an embodiment of the present invention provides a network fault tracing system, including: a log aggregation module, used to aggregate the full-link logs in the network to be detected, the full-link logs including a set of log records of all source nodes accessing the destination nodes; a complex network construction module, used to construct a complex network based on the full-link logs; a connectivity calculation module, used to calculate the network connectivity of the complex network when there are faulty nodes in the complex network; a fault tracing module, used to determine whether the network connectivity is lower than a preset threshold; if so, determine the fault period and determine the root fault node within the fault period.

[0008] At the same time, an embodiment of the present invention further provides a computer-readable storage medium on which a computer program is stored. When the program is executed by a processor, the aforementioned network fault tracing method is implemented.

[0009] In addition, an embodiment of the present invention also provides a computer device, which includes: one or more processors; a storage device for storing one or more programs; when the one or more programs are executed by the one or more processors, the one or more processors implement the aforementioned network fault tracing method.

[0010] The above technical solution has the following beneficial effects:

[0011] This technical solution applies complex network technology to network fault tracing and determines whether the network is faulty through network connectivity: the network is simplified by using the relationship between the in-and-out degrees of complex network nodes, so that the entire network is judged according to reliable physical connections, and the shortest path between nodes is used for fault tracing, meeting the needs of fault propagation analysis and cascading fault root cause analysis in complex network environments, and completely relying on complex network methods without the need for human intervention. The results are more objective and effectively ensure the accuracy of the fault tracing results.

[0012] At the same time, this technical solution also has the following characteristics:

[0013] 1. The analysis methods in the prior art rely on historical data and cannot cope with new problems arising from architecture changes. However, in the technical solution of the present application, real-time data is used as the basis for creating a network, and there is no obvious dependence on historical data, which lowers the threshold for use and improves applicability.

[0014] 2. The existing methods make judgments based on certain rules and are unable to provide the entire fault development process and the scope of the fault's impact on the architecture, and are relatively lacking in visualization of fault events. The method of this application can draw a graph of the faulty nodes and the affected links, making it easier to intuitively view the fault development and impact scope. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0016] Figure 1 is a flow chart of a network fault tracing method according to an embodiment of the present invention;

[0017] Figure 2 is a structural diagram of a network fault tracing system according to an embodiment of the present invention;

[0018] Figure 3 is a flow chart of a specific embodiment of the present invention;

[0019] Figure 4 is a schematic diagram of constructing a complex network in a specific embodiment of the present invention;

[0020] Figure 5 is a schematic diagram of determining a fault period in a specific embodiment of the present invention;

[0021] Figure 6 It is a schematic diagram of determining a root fault node in a specific embodiment of the present invention. DETAILED DESCRIPTION

[0022] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0023] First, the technical terms involved in the embodiments of the present invention are defined as follows:

[0024] Complex network: A network that has some or all of the properties of self-organization, self-similarity, attractor, small world, and scale-free is called a complex network.

[0025] Global network efficiency g1 (global_efficiency): Global efficiency refers to the reciprocal of the shortest average path, which is used to measure the global transmission capacity of the network. It can be calculated using the complex network library nx.global_efficiency(). In this article, g1 is the global network efficiency of the normal network, and g2 is the global network efficiency under fault conditions.

[0026] Network connectivity threshold k: used to determine whether a fault affects network operation. If it is lower than this threshold, the network operation efficiency is low and data transmission within the network is blocked. It is used to measure whether to make a fault determination.

[0027] Data warehouse: A data store that stores all types of data and supports data analysis and decision making.

[0028] Full-link log: A log used to ensure the quality of service access, including the source node and the destination node, as well as the log record of the entire link from the source call to the destination node, mostly used for performance analysis.

[0029] Connectivity: A constant of the network connectivity status calculated based on the global network efficiency of complex network analysis. It is used to express the overall connectivity of the current network and whether there are any faults, and serves as a basis for determining the fault time.

[0030] Fault line: A link consisting of faulty nodes and the call relationships between them, used to analyze clues to the root cause of the fault.

[0031] Error reporting node and destination node: The node where an error or exception occurs in the database record. The database records the full-link log. The source node of the error access and the destination node of the access are the error reporting node and the destination node.

[0032] like Figure 1 As shown, an embodiment of the present invention provides a network fault tracing method, including:

[0033] S101, summarizing the full-link log in the network to be detected, wherein the full-link log includes a set of log records of all source nodes accessing destination nodes;

[0034] S102, constructing a complex network according to the full-link log;

[0035] S103, when there is a faulty node in the complex network, determining the network connectivity of the complex network;

[0036] S104: When the network connectivity is lower than a preset threshold, determine a fault period, and determine a root fault node within the fault period.

[0037] Furthermore, the step S104 specifically includes:

[0038] S1041, taking the continuous period corresponding to when the network connectivity is lower than the preset threshold as the fault period;

[0039] S1042, taking all nodes that report errors during the fault period as fault nodes to be screened;

[0040] S1043, for each of the faulty nodes to be screened, determining in sequence the shortest links between the faulty node to be screened and other faulty nodes to be screened;

[0041] S1044. Gather all the shortest links into a shortest link set;

[0042] S1045, determining the number of times each of the to-be-screened fault nodes appears as a destination node in the shortest link set;

[0043] S1046: Select the to-be-screened fault node with the largest number of repetitions as the root fault node.

[0044] In the faulty link, the more times a certain faulty node to be screened appears as a destination node, the greater the impact of the point on the faulty link. Therefore, the faulty node to be screened with the most appearances can be selected as the root cause of the fault.

[0045] Furthermore, the step S102 specifically includes:

[0046] S1021. Use the source node and the destination node in the full-link log as nodes of the complex network;

[0047] S1022. Using the call relationship between the source node and the destination node in the full-link log as a directed edge of the complex network;

[0048] S1023. Gather all the nodes and the directed edges to form the complex network.

[0049] Furthermore, the step S103 specifically includes:

[0050] S1031, determining a second global network efficiency of the complex network when there is a faulty node;

[0051] S1032, retrieving a predetermined first global network efficiency of the complex network when there is no faulty node;

[0052] S1033: Determine network connectivity of the complex network according to the first global network efficiency and the second global network efficiency.

[0053] Furthermore, step S1033 is performed using the following formula:

[0054]

[0055] Among them, g2 is the second global network efficiency, g1 is the first global network efficiency, and t is the network connectivity.

[0056] Furthermore, the network fault tracing method further includes:

[0057] S106, constructing a fault network through the fault nodes to be screened and the shortest link set;

[0058] S107: In the complex network, display the fault network in color.

[0059] like Figure 2 As shown, an embodiment of the present invention provides a network fault tracing system, including:

[0060] The log aggregation module 21 is used to aggregate the full-link logs in the network to be detected;

[0061] A complex network construction module 22, used to construct a complex network according to the full-link log;

[0062] A connectivity calculation module 23, used for calculating the network connectivity of the complex network when there is a faulty node in the complex network;

[0063] The fault tracing module 24 is used to determine whether the network connectivity is lower than a preset threshold; if so, determine the fault period and the root fault node within the fault period.

[0064] Furthermore, the fault tracing module 24 is specifically used to: take the continuous time period when the network connectivity is lower than the preset threshold as the fault period; take all nodes that report errors within the fault period as fault nodes to be screened; for each of the fault nodes to be screened, determine the shortest links between it and other fault nodes to be screened in turn; gather all the shortest links into a shortest link set; determine the number of repetitions of each fault node to be screened appearing as a destination node in the shortest link set; select the fault node to be screened with the largest number of repetitions as the root fault node.

[0065] Furthermore, the complex network construction module 22 is specifically used to: use the source node and the destination node in the full-link log as the nodes of the complex network; use the calling relationship between the source node and the destination node in the full-link log as the directed edge of the complex network; and gather all the nodes and the directed edges to form the complex network.

[0066] Furthermore, the connectivity calculation module 23 is specifically used to: calculate the global network efficiency g2 of the complex network when there is a faulty node; retrieve the pre-calculated global network efficiency g1 of the complex network when there is no faulty node; and calculate the network connectivity t of the complex network according to the following formula:

[0067]

[0068] Furthermore, the fault tracing system based on complex network technology also includes a fault network construction module 25, which is used to construct a fault network through the fault nodes to be screened and the shortest link set.

[0069] The above technical solution of the embodiment of the present invention is described in detail below with reference to a specific application example:

[0070] The overall technical process of this specific embodiment is as follows Figure 3 As shown, execute in the order shown in the figure.

[0071] Overall technical description: Obtain full-link logs from the data warehouse in time series, process the full-link logs to obtain information such as source nodes, destination nodes, and access quality; build an overall network architecture in a complex network according to the directional relationship between source nodes and destination nodes; set the status of each node according to the access quality, set the faulty node to be disconnected, and determine whether the network connectivity is lower than the threshold; make a fault determination if it is lower than the threshold, retrieve the junction point where the connectivity is equal to the threshold as the beginning of the fault, and give the time period of the fault; search all faulty nodes and all links between faulty nodes according to the faulty time period; summarize the number of times each node appears as the destination node in the shortest link, and determine that the node with the largest value is the root cause of the fault; draw a fault network, mark the faulty network in red and display it in the overall network.

[0072] In this diagram, if the network connectivity is not lower than the threshold, there is no need for subsequent fault tracing, and the process returns to the revelation step to loop, retrieve the full-link log, rebuild the complex network, restart the connectivity calculation, etc. In actual operation, since the large changes in the computer node network are not particularly fast, there is no need to repeat the complex network construction frequently. It is sufficient to return to the connectivity calculation step to implement cyclic judgment. After a certain time, such as 1 hour, the process returns to the beginning, retrieves the full-link log, and rebuilds the complex network.

[0073] Among them, the data warehouse is used to store the full-link logs of each service node (logs include: access logs between programs, network layer transmission logs, DNS resolution logs, etc.), and the log fields include: access time, source node, destination node, access status code, access time, error information, etc., as the original data of the entire system. The access log is written to the data warehouse by the manager of each node. Therefore, when executing the fault tracing method of this application, it is necessary to first obtain the full-link log from the data warehouse.

[0074] The process of building a complex network is as follows Figure 4As shown in the figure, the source nodes and destination nodes in the data warehouse are regarded as the set of nodes V in the complex network, and the call relationships between the source and the destination are regarded as the set of directed edges E in the complex network. By directly constructing the network architecture G=(V, E) using the complex network library, the overall network architecture can be obtained. After the overall architecture of the complex network is completed, the global network efficiency g1 (global_efficiency) can be calculated using the complex network library.

[0075] Determination of the network connectivity of the complex network: Calculate the global network efficiency g2 using the complex network library to judge the network connectivity Whether it is lower than the preset network connectivity threshold k, and t < k triggers the traceability determination.

[0076] The process of judging the fault time period is as Figure 5 shown in the figure. Calculate the global network efficiency g at each time point, calculate the network connectivity t, obtain the time period when it is lower than the threshold (t < k), and then infer the time period when the fault occurs (the part between the start time and the end time in the figure).

[0077] It should be noted that if there is no time period during the entire judgment process when the network connectivity t is lower than the threshold k, it means that there are no fault nodes in the entire network during this process, or there are very few fault nodes, so it is easy to conduct intuitive fault analysis without the intervention of this method. When there are multiple fault time periods during the judgment process, subsequent fault line search and fault root cause judgment can be performed for each fault time period respectively.

[0078] The process of fault line search and fault root cause judgment is as Figure 6 shown in the figure. According to the fault time period, retrieve all fault nodes (including error reporting nodes and destination nodes) from the data warehouse (full-link log), and then use the complex network library to obtain the shortest paths Ei between all nodes (that is, for the set of all fault nodes, group each fault node with the rest of the fault nodes once, and find the shortest path between this group; until each fault node has completed a combination with other fault nodes except itself once), and obtain the destination nodes Vi of all the shortest paths (that is, Figure 6 all the nodes marked with double-ring circles in the upper right set in the figure).

[0079] After that, according to the fault route search result Ei, calculate the number of times the destination node appears in the fault link ∑V i , sort all the destination nodes (see Figure 6 the lower right set in the figure), and take the destination node V with the most occurrences. The fact that the node V appears the most times as the destination in the fault link indicates that this point has the greatest impact on the fault link, and it is judged that the node V is the root cause of the fault.

[0080] Finally, draw the fault network: use the complex network library to directly construct the network architecture Gi=(Vi,Ei) to obtain the fault network.

[0081] The disclosed embodiments are described above to enable any person skilled in the art to implement or use the present invention. Various modifications of these embodiments are obvious to those skilled in the art, and the general principles defined herein may also be applied to other embodiments without departing from the spirit and scope of the present disclosure. Therefore, the present disclosure is not limited to the embodiments given herein, but is consistent with the broadest scope of the principles and novel features disclosed in this application.

[0082] The specific implementation methods described above further illustrate the objectives, technical solutions and beneficial effects of the present invention in detail. It should be understood that the above description is only a specific implementation method of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A network fault tracing method, It is characterized in that include: Summarize the full-link log in the network to be detected, where the full-link log includes a set of log records of all source nodes accessing the destination node; Building a complex network based on the full-link log; When there is a faulty node in the complex network, determining the network connectivity of the complex network; When the network connectivity is lower than a preset threshold, determining a fault period, and determining a root fault node within the fault period; Determining the network connectivity of the complex network includes: determining a second global network efficiency of the complex network in the presence of a faulty node; Retrieving a predetermined first global network efficiency of the complex network when there is no faulty node; The network connectivity of the complex network is determined according to the first global network efficiency and the second global network efficiency.

2. The network fault tracing method according to claim 1, It is characterized in that The determining of the fault period and determining the root fault node within the fault period includes: The continuous period corresponding to when the network connectivity is lower than the preset threshold is used as the fault period; All nodes that report errors during the fault period are taken as fault nodes to be screened; For each of the faulty nodes to be screened, determining in sequence the shortest links between the faulty node to be screened and other faulty nodes to be screened; Gather all the shortest links into a shortest link set; Determine the number of times each of the to-be-screened fault nodes appears as a destination node in the shortest link set; The to-be-screened fault node with the largest number of repetitions is selected as the root fault node.

3. The network fault tracing method according to claim 1, It is characterized in that The constructing a complex network according to the full-link log includes: Using the source node and the destination node in the full-link log as nodes of the complex network; Using the call relationship between the source node and the destination node in the full-link log as a directed edge of the complex network; All the nodes and the directed edges are assembled to form the complex network.

4. The network fault tracing method according to claim 1, It is characterized in that The determining of the network connectivity of the complex network according to the first global network efficiency and the second global network efficiency , Use the following formula: ; in, is the second global network efficiency, is the first global network efficiency, and t is the network connectivity.

5. The network fault tracing method according to claim 2, It is characterized in that Also includes: Constructing a fault network through the fault nodes to be screened and the shortest link set; In the complex network, the fault network is displayed in color.

6. A network fault tracing system, It is characterized in that include: A log aggregation module is used to aggregate the full-link logs in the network to be detected, wherein the full-link logs include a set of log records of all source nodes accessing the destination nodes; A complex network construction module, used to construct a complex network according to the full-link log; A connectivity calculation module, used for determining the network connectivity of the complex network when there is a faulty node in the complex network; A fault tracing module, used to determine a fault period and a root fault node within the fault period when the network connectivity is lower than a preset threshold; The connectivity calculation module is specifically used to: determine the second global network efficiency of the complex network when there is a faulty node; retrieve the predetermined first global network efficiency of the complex network when there is no faulty node; The network connectivity of the complex network is determined according to the first global network efficiency and the second global network efficiency.

7. The network fault tracing system according to claim 6, It is characterized in that The fault tracing module is specifically used for: The continuous period corresponding to when the network connectivity is lower than the preset threshold is taken as the fault period; all nodes that report errors during the fault period are taken as fault nodes to be screened; for each of the fault nodes to be screened, the shortest links between the fault node to be screened and other fault nodes to be screened are determined in turn; Gather all the shortest links into a shortest link set; determine the number of times each of the faulty nodes to be screened appears as a destination node in the shortest link set; The to-be-screened fault node with the largest number of repetitions is selected as the root fault node.

8. A computer-readable storage medium having a computer program stored thereon, It is characterized in that When the program is executed by a processor, the network fault tracing method as described in any one of claims 1 to 5 is implemented.

9. A computer device, It is characterized in that It includes: one or more processors; A storage device for storing one or more programs; When the one or more programs are executed by the one or more processors, the one or more processors implement the network fault tracing method as described in any one of claims 1 to 5.

Citation Information

Patent Citations

  • A fault propagation path simulation method based on a complex network model

    CN109597752A

  • Power transmission network structure form evaluation method

    CN110969347A

  • Using a programmable resource dependency mathematical model to perform root cause analysis

    CN114070724A

  • Fault equipment positioning method and device, electronic equipment, medium and program product

    CN114710400A

  • Using a programmable resource dependency mathematical model to perform root cause analysis

    US20220045898A1