CN112231523B - Network fault positioning and troubleshooting method and system based on directed acyclic graph - Google Patents

Abstract

The invention relates to a method and system for locating and troubleshooting network faults based on directed acyclic graphs. The method includes: obtaining a computer network system and its component equipment; drawing a directed acyclic graph based on the computer network system and its component equipment; The acyclic graph obtains the faulty loop where the network fault occurs; obtains the troubleshooting method of the faulty loop based on the difficulty of implementation or contribution; uses the troubleshooting method to locate the faulty loop; determines whether the fault is the equipment under inspection; if so, displays the preview of the equipment under inspection. If not, use troubleshooting methods to locate faults in the upstream direction Internet loop and downstream direction equipment terminal loops; until the fault nodes in the upstream direction Internet loop and downstream direction equipment terminal loops are obtained and the fault node presets are displayed Troubleshooting. The invention can quickly realize network fault location and reduce time and labor costs.

Description

A network fault location and troubleshooting method and system based on directed acyclic graph

Technical field

The invention relates to the field of computer network industry operation and maintenance, and in particular to a network fault location and troubleshooting method and system based on directed acyclic graphs.

Background technique

The importance of the network to modern enterprise production is increasing day by day. The operation and management of production equipment, internal communication, and external display all require smooth network connections. However, failures caused by physical (equipment aging, replacement, etc.) and logic (configuration errors, IP changes, etc.) are difficult to avoid. Moreover, as the network structure becomes more and more complex, the number of paths that may fail increases non-linearly with the increase in the number of devices, making it increasingly difficult to locate faults.

In the existing technology, a processing method based on monitoring tools and expert experience is generally adopted: monitoring tools (such as zabbix) can alert operation and maintenance personnel as soon as a fault occurs. However, the failure of one device often causes data anomalies in its upstream and downstream devices, resulting in an "alarm storm" phenomenon (a large number of devices alarming at the same time); which does not help solve the problem. For non-operation and maintenance department employees, due to the lack of monitoring tools and operation and maintenance experience, they can only explore on their own when they cannot connect to the Internet, and then report if they are unsuccessful, which is inefficient.

Experienced experts will first clarify the upstream and downstream relationships between equipment, use the most effective testing methods to quickly locate the problem equipment based on the current fault symptoms, and then determine the most likely failure based on the history of previously handled cases. direction (improper configuration, aging equipment, loose ports), and then handle the failure with the lowest-cost solution.

Ideally, that is, 1) changes in network topology (addition or deletion of equipment, connection methods between devices, etc.) can be synchronized to each expert in real time; and 2) when a fault occurs, the experts can arrive at the scene as soon as possible and with a stable spirit and physical status handling failures. In this way, losses caused by network failures can be minimized. However, there is no doubt that the human cost of achieving this state will be unbearable for enterprises.

Contents of the invention

The purpose of the present invention is to provide a network fault location and troubleshooting method and system based on directed acyclic graphs, which can quickly realize fault location, provide solutions, and reduce network maintenance time and labor costs.

In order to achieve the above objects, the present invention provides the following solutions:

A network fault location and troubleshooting method based on directed acyclic graph, including:

Acquire computer network systems and their component equipment;

Draw a directed acyclic graph according to the acquisition of the computer network system and its component equipment;

Obtain the fault loop where the network fault occurs according to the directed acyclic graph;

Obtain the troubleshooting method of the fault loop based on the difficulty of implementation or degree of contribution. The troubleshooting method is divided into bypassing the inspected equipment, upstream direction Internet loop, and downstream direction equipment terminal loop according to different detection directions;

Use the troubleshooting method to locate the fault of the fault circuit;

Determine whether the fault is the equipment under inspection;

If so, display the preset fault solution for the device under inspection;

If not, use the troubleshooting method to locate the fault in the upstream direction Internet loop and the downstream direction equipment terminal loop;

Until the faulty node in the upstream direction Internet loop and the downstream direction equipment terminal loop is obtained and the preset fault solution for the faulty node is displayed.

Optionally, drawing a directed acyclic graph based on the acquisition of the computer network system and its component devices includes:

The component devices are depicted as nodes, and the connection relationships between the component devices are depicted as edges.

Optionally, the troubleshooting methods include:

The upstream direction Internet loop detection method, the downstream direction device terminal loop detection method, and the detection method for bypassing the inspected equipment.

Optionally, the upstream direction Internet loop detection method includes:

Obtain the node where the inspected device is located and its sibling nodes in the directed acyclic graph;

Send ping packets from the device under inspection and its sibling node devices to the upstream direction to the Internet;

If the device under inspection cannot connect to the Internet, but its sibling nodes can, it can be determined that the fault lies with the device;

If all sibling nodes cannot connect, the fault occurs in the upstream direction Internet loop.

Optionally, the downstream direction equipment terminal loop detection method includes:

Send a ping packet from the downstream direction device terminal to the device under inspection;

If the downstream device terminal can communicate with the device being inspected, the fault occurs in the upstream Internet loop;

If the downstream equipment terminal cannot communicate with the equipment being inspected, the fault occurs in the downstream equipment terminal loop.

Optionally, the method for bypassing the detected device detection includes:

Send ping packets from the downstream direction device terminal connection to the upstream direction Internet;

If the downstream device terminal connection and the upstream Internet can be connected, the fault occurs in the device being inspected;

If the downstream device terminal connection and the upstream Internet cannot be connected, the problem occurs in the upstream Internet loop or the downstream device terminal loop.

Optionally, the greater the implementation difficulty, the greater the corresponding preset value.

Optionally, the greater the contribution, the greater the preset value.

Optionally, the troubleshooting method for obtaining the fault loop based on implementation difficulty or contribution includes:

Determine whether the preset values of implementation difficulty of the troubleshooting method are equal;

If not, select the fault loop troubleshooting method with a smaller preset difficulty level;

If so, select the fault loop troubleshooting method with a larger preset contribution value.

A network fault location and troubleshooting system based on directed acyclic graph, including:

The device acquisition module is used to acquire the computer network system and its component devices;

An image drawing module, used to draw a directed acyclic graph according to the acquisition computer network system and its component equipment;

A fault loop determination module, configured to obtain a fault loop in which a network fault occurs based on the directed acyclic graph;

A troubleshooting method selection module is used to obtain the troubleshooting method of the fault loop based on implementation difficulty or contribution. The troubleshooting method is divided into bypassing the inspected equipment, upstream direction Internet loop, and downstream direction equipment terminal loop according to different detection directions;

A first fault location module, configured to use the troubleshooting method to locate faults on the fault circuit;

A judgment module, used to judge whether the fault is the equipment under inspection;

The first display module is used to display the preset fault solution of the inspected equipment when the inspected equipment fails;

The second fault location module is used to use the troubleshooting method to locate faults on the upstream direction Internet loop and the downstream direction equipment terminal loop;

The second display module is used to display the preset fault solutions for faulty nodes in the upstream direction Internet loop and the downstream direction equipment terminal loop.

According to the specific embodiments provided by the present invention, the present invention discloses the following technical effects:

The present invention designs a fault handling method and system based on directed acyclic graphs and expert knowledge. Causal relationship reasoning based on directed acyclic graphs is used to troubleshoot and locate problems; expert knowledge is used to provide equipment inspection methods at each step of reasoning. , and provide troubleshooting methods after locating. Using the present invention, as if a network expert is assisting, the user can handle the fault with high quality as soon as the fault is discovered and minimize the losses caused by the fault.

Description of the drawings

In order to explain the embodiments of the present invention or the technical solutions in the prior art more clearly, the drawings needed to be used in the embodiments will be briefly introduced below. Obviously, the drawings in the following description are only some of the drawings of the present invention. Embodiments, for those of ordinary skill in the art, other drawings can also be obtained based on these drawings without exerting creative efforts.

Figure 1 is a flow chart of the network fault location and troubleshooting method based on directed acyclic graph according to the present invention;

Figure 2 is a module diagram of the network fault location and troubleshooting system based on directed acyclic graph according to the present invention;

Figure 3 is a schematic structural diagram of a directed acyclic graph of the present invention.

Detailed ways

The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only some of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts fall within the scope of protection of the present invention.

The purpose of the present invention is to provide a network fault location and troubleshooting method and system based on directed acyclic graphs, which can quickly realize fault location and provide solutions.

In order to make the above objects, features and advantages of the present invention more obvious and understandable, the present invention will be described in further detail below in conjunction with the accompanying drawings and specific embodiments.

Figure 1 is a flow chart of the network fault location and troubleshooting method based on directed acyclic graphs of the present invention. As shown in Figure 1, a network fault location and troubleshooting method based on directed acyclic graphs includes:

Step 101: Obtain the computer network system and its component equipment;

Step 102: Draw a directed acyclic graph based on the acquisition of the computer network system and its component devices;

Step 103: Obtain the fault loop where the network fault occurs according to the directed acyclic graph;

Step 104: Obtain the troubleshooting method of the fault loop based on the difficulty of implementation or degree of contribution. The troubleshooting method is divided into bypassing the inspected equipment, upstream direction Internet loop, and downstream direction equipment terminal loop according to different detection directions;

Step 105: Use the troubleshooting method to locate the fault of the fault circuit;

Step 106: Determine whether the fault is the equipment under inspection;

Step 107: If yes, display the preset fault solution for the device under inspection;

Step 108: If not, use the troubleshooting method to locate faults in the upstream direction Internet loop and the downstream direction equipment terminal loop;

Step 109: Obtain the faulty node in the upstream direction Internet loop and the downstream direction equipment terminal loop and display the preset fault solution for the faulty node.

The above solution of the present invention will be further elaborated below:

The basic data structure of the present invention is a directed acyclic graph.

The structure of a directed graph reflects the causal connections between nodes, as shown in Figure 3. Problem links are always marked in dark gray in the graph. Each node represents a real device, and the corresponding inspection items, repair methods and "contribution degree" are stored as attributes of the node; the causal link is determined by the edge and its direction.

The subdivision structure inside the node is also stored by the directed acyclic graph graph. Expert knowledge is therefore fully expressed by a directed graph of multiple "levels". In the process of users using the system, lower-level knowledge can be obtained through continuous drilling actions.

The specific process includes:

1. Obtain the name and type of the faulty device

2. Eliminate redundant loops of faulty equipment

A device can connect to the Internet through multiple paths ("pc_1" in Figure 3, which can connect to the Internet through wired and wireless methods). In a directed graph (also called a directed acyclic graph), a node has multiple parent nodes. In the worst case, the number of possible paths increases exponentially with the number of such nodes (on the connection path from the problem terminal to the Internet), and the time complexity of fault detection therefore increases exponentially.

But the reality is that the failure path is usually certain and known. For example, in Figure 3, although "pc_1" can access the switch through either a network cable or a wireless access point, the problem of which connection method is used when a fault occurs is known to the user. With just one interaction, the system can eliminate redundant loops at a node and identify faulty loops.

Step 3, specific fault detection

To detect whether a device has a fault, there are three reasoning logics depending on the detection direction:

Upstream direction ("upstream"; relative to the device being inspected, the Internet is the upstream direction, and the terminal device is the downstream direction, the same below): For example, sending ping packets from a DNS server to other servers on the Internet. At this time, by judging the sibling nodes in the directed graph (referring to having the same parent node; "pc_1" and "mac_imac" in Figure 3 are sibling nodes because they have the same parent node "switch_office_1"; click here Definition, "pc_1" is also its own sibling node) to determine whether there is a fault. Specifically, if the device cannot connect to the Internet but its sibling nodes can, the fault can be determined to be with that device. If all sibling nodes cannot connect, the fault occurs upstream (parent node and above).

Downstream direction ("downstream"): Send a connection request from the problem terminal to the device under inspection. For example, send a ping packet from the problem terminal to a DNS server. If the connection is possible, the fault occurs upstream of the DNS server; if not, the fault occurs downstream of the DNS server.

Bypass ("bypass"): Bypassing the inspected device from the problematic terminal to connect to the Internet. For example, bypassing a DNS server through an IP connection from the problematic endpoint, or bypassing a wireless access point through a network cable. If it can be connected, the fault occurs in the device. If not, the problem occurs in the upstream or downstream of the device being inspected. The above inference logic is summarized in Table 1.

Step 4, drill into the internal structure

Nodes may have internal structures. Ignoring the internal structure will lead to inaccurate problem location, and building the entire graph directly from the bottom layer will make the graph too complex. The present invention solves this dilemma through modularization. After the system initially locates the fault point or fault segment, the user can continue to drill down to analyze the lower-level structure layer by layer. That is, for each layer, it can be understood as jumping to "Get the name and type of the faulty device" , loop steps 2-3 until the specific fault node is located.

Step 5, statistical information storage, display and update of diagnosis/repair methods

Information on detection and repair methods includes quantitative descriptions, including two values: "contribution degree" and "implementation difficulty." These values are initialized by experts. Among them, the implementation difficulty remains unchanged after initialization; the contribution value will be automatically updated after each diagnosis.

An example is:

The current check proceeds to "Huawei switch 1". The equipment node in the knowledge base has previously entered 3 detection methods and their corresponding "implementation difficulty" and "contribution degree", which are:

1) "Bypass the switch and directly connect to the upstream router", implementation difficulty 30, contribution 2;

2) "Change an interface", implementation difficulty is 10, contribution is 50;

3) "Replace the network cable connected to the switch", the implementation difficulty is 30 and the contribution is 10.

The order of inspection suggestions given by the system will be 2) 3) 1); that is, the simplest inspection method is ranked first. If the difficulty is the same (such as 1) and 3)), the one with greater contribution will be ranked first.

During the event processing, if the question option corresponding to a detection method is selected (and therefore contributes to problem location), the "contribution degree" value of the detection method is +1; the final contribution value of each detection method is according to the formula ( 1) The rules shown (the maximum contribution value is normalized to 100, and the remaining contribution values are scaled and rounded in the same proportion) are updated simultaneously:

Formula (1) is:

Contribution of this method = int (contribution of this method / maximum value of contribution of each detection method * 100)

Similarly, if a repair method fixes the fault, the "contribution degree" value of the method is +1; in the end, the possibility of each repair method is updated simultaneously according to formula (2)

Formula (2) is: (The maximum contribution value is normalized to 100, and the remaining contribution values are scaled in the same proportion and rounded)

Contribution of this method = int (contribution of this method / maximum value of contribution of each repair method * 100)

The specific "implementation difficulty" and "contribution degree" need to be considered in the process of fault detection and fault repair: according to the input name and type of faulty equipment, the database will match the corresponding detection methods and solutions. These two Methods will be ranked according to their degree of difficulty based on "difficulty of implementation" and "contribution" before deciding which method to use.

The following is a description of a specific diagnostic process:

1. Determine the problem domain and related equipment. You can click on it, or you can enter the question using natural language and use the system's matching function to determine it. For example, the current problem device is pc_1 (the name of the device in the system).

2. Remove redundant loops through interaction.

3. Enter the diagnosis process. After the system sorts the detection methods according to the "Implementation Difficulty" value, it lists all detection suggestions corresponding to the current step in the "Troubleshooting" column.

4. The user conducts detection according to the recommended detection method and feedbacks the answer by clicking on the question, and the system makes inferences based on the answer;

5. During the diagnosis process, the system will give corresponding conclusions based on different feedback from users, or raise further questions and continuously narrow down the scope of investigation.

Repeat steps 3-5 until the problem is located. The system gives suggestions after comprehensive ranking based on the contribution and implementation difficulty of the repair methods. If the located device has internal structures, the user can continue to drill.

According to the event processing results - the fault is solved according to the method recommended by the system, the fault is solved according to the user's own method, and the fault is not solved - the system automatically generates a report. Users can edit and submit as needed, update the system's statistical data, and retain it as historical data.

In view of the above method, the present invention also discloses a network fault location and troubleshooting system based on directed acyclic graph, as shown in Figure 2, including:

The device acquisition module 201 is used to acquire the computer network system and its component devices;

The image drawing module 202 is used to draw a directed acyclic graph according to the acquisition computer network system and its component equipment;

The fault loop determination module 203 is used to obtain the fault loop in which a network fault occurs according to the directed acyclic graph;

The troubleshooting method selection module 204 is used to obtain the troubleshooting method of the fault loop based on the difficulty of implementation or degree of contribution. The troubleshooting method is divided into bypassing the inspected device, upstream direction Internet loop, and downstream direction equipment terminal loop according to different detection directions;

The first fault location module 205 is used to locate faults on the fault circuit using the troubleshooting method;

Determination module 206, used to determine whether the fault is the device being inspected;

The first display module 207 is used to display the preset fault solution of the inspected equipment when the inspected equipment fails;

The second fault location module 208 is used to use the troubleshooting method to locate faults on the upstream direction Internet loop and the downstream direction equipment terminal loop;

The second display module 209 is used to display the preset fault solutions for faulty nodes in the upstream direction Internet loop and the downstream direction equipment terminal loop.

The beneficial effects brought by the technical solution of the present invention are:

Based on the characteristics of computer networks, the present invention designs a data structure that is convenient for storing expert knowledge, classifies detection methods, and builds reasoning logic on this basis. This can provide expert advice on locating and repairing faults stably and efficiently, reducing losses caused by network connection interruptions; and recording expert knowledge in the system in an easy-to-understand, comprehensive, and quantifiable manner, and sharing it with users along with the system; In addition, system operation is less dependent on a single expert, reducing single-point risks; it can also automatically complete event recording and statistics, making it easier to quantitatively assess the sources of system instability and then make targeted improvements.

Each embodiment in this specification is described in a progressive manner. Each embodiment focuses on its differences from other embodiments. The same and similar parts between the various embodiments can be referred to each other. As for the system disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple. For relevant details, please refer to the description in the method section.

This article uses specific examples to illustrate the principles and implementation methods of the present invention. The description of the above embodiments is only used to help understand the method and the core idea of the present invention; at the same time, for those of ordinary skill in the art, according to the present invention There will be changes in the specific implementation methods and application scope of the ideas. In summary, the contents of this description should not be construed as limitations of the present invention.

Claims (10)

1. A network fault locating and checking method based on a directed acyclic graph is characterized by comprising the following steps:

acquiring a computer network system and constituent equipment thereof;

drawing a directed acyclic graph according to the acquired computer network system and the constituent equipment thereof;

acquiring a fault loop with network faults according to the directed acyclic graph;

the method comprises the steps of obtaining the troubleshooting method of the fault loop according to implementation difficulty or contribution, wherein the troubleshooting method is divided into a bypass device to be checked, an upstream direction internet loop and a downstream direction device terminal loop according to different detection directions;

adopting the troubleshooting method to locate the fault of the fault loop;

judging whether the fault is the detected equipment or not;

if yes, displaying a fault solution preset by the detected equipment;

if not, adopting the troubleshooting method to locate faults of the upstream direction Internet loop and the downstream direction equipment terminal loop;

and until fault nodes in the upstream direction Internet loop and the downstream direction equipment terminal loop are obtained, and a fault solution preset by the fault nodes is displayed.

2. The directed acyclic graph-based network fault localization and investigation method according to claim 1, wherein the drawing the directed acyclic graph according to the acquired computer network system and its constituent devices comprises:

the constituent devices are depicted as nodes, and the connection relationship between the constituent devices is depicted as edges.

3. The directed acyclic graph-based network fault location troubleshooting method of claim 1, wherein the troubleshooting method comprises:

an upstream direction internet loop detection method, a downstream direction equipment terminal loop detection method and a bypass detected equipment detection method.

4. A directed acyclic graph-based network fault localization and investigation method according to claim 3, wherein the upstream direction internet loop detection method comprises:

acquiring a node where detected equipment is located and a brother node of the node in the directed acyclic graph;

sending a ping packet to the Internet from the detected device and the devices of the brothers of the detected device to the upstream direction;

if the detected equipment cannot be connected to the Internet and the brother node of the detected equipment can be connected, the fault can be determined to be in the equipment;

if none of the sibling nodes is able to connect, then the failure occurs in the Internet loop in the upstream direction.

5. The directed acyclic graph-based network fault location and investigation method according to claim 3, wherein the downstream direction equipment terminal loop detection method comprises:

sending a ping packet from the downstream direction equipment terminal to the equipment to be detected;

if the downstream direction equipment terminal can be communicated with the equipment to be detected, a fault occurs in an upstream direction Internet loop;

if the downstream direction equipment terminal cannot be communicated with the equipment to be detected, a fault occurs in a downstream direction equipment terminal loop.

6. A directed acyclic graph-based network fault localization method according to claim 3, wherein said bypassing inspected device detection method comprises:

sending a ping packet to the Internet from the downstream direction equipment terminal connection to the upstream direction;

if the downstream equipment terminal connection and the upstream Internet can be communicated, a fault occurs in the detected equipment;

if the downstream direction device termination connection is not connected to the upstream direction internet, a problem arises with the upstream direction internet loop or the downstream direction device termination loop.

7. The directed acyclic graph-based network fault location and investigation method according to claim 1, wherein the greater the implementation difficulty, the greater the corresponding preset value.

8. The directed acyclic graph-based network fault location and investigation method according to claim 7, wherein the larger the contribution is, the larger the corresponding preset value is.

9. The directed acyclic graph-based network fault location and investigation method according to claim 8, wherein the investigation method for obtaining the fault loop according to implementation difficulty or contribution degree comprises the following steps:

judging whether implementation difficulty preset values of the checking method are equal or not;

if not, selecting a fault loop checking method with small implementation difficulty preset value;

if yes, a fault loop checking method with a large contribution degree preset value is selected.

10. A directed acyclic graph-based network fault location and investigation system, comprising:

the device acquisition module is used for acquiring a computer network system and the devices formed by the computer network system;

the image drawing module is used for drawing a directed acyclic graph according to the acquired computer network system and the component equipment thereof;

the fault loop determining module is used for acquiring a fault loop with network faults according to the directed acyclic graph;

the checking method selection module is used for acquiring checking methods of the fault loop according to implementation difficulty or contribution, wherein the checking methods are divided into a bypass checked device, an upstream direction internet loop and a downstream direction device terminal loop according to different detection directions;

the first fault locating module is used for locating faults of the fault loop by adopting the troubleshooting method;

the judging module is used for judging whether the fault is the detected equipment or not;

the first display module is used for displaying a fault solution preset by the detected equipment when the detected equipment fails;

the second fault locating module is used for locating faults of the upstream direction Internet loop and the downstream direction equipment terminal loop by adopting the troubleshooting method;

and the second display module is used for displaying fault solutions preset by fault nodes in the upstream-direction Internet loop and the downstream-direction equipment terminal loop.