KR102408172B1 - Method and device of providing a solution for network interruption by remote automatic network analysis under network interruption - Google Patents

Abstract

Disclosed is a method for supporting network failure resolution through remote automatic network analysis when a network failure occurs, that is, (a) a network manager, when a failure resolution request is obtained from a user of the network, refers to the failure resolution request, and the network selecting a specific network device on the (b) causing, by the network manager, the specific network device to perform at least one specific analysis operation selected according to the failure resolution request among at least one preset analysis operation; and (c) providing, by the network manager, at least a part of a specific analysis result of the specific analysis operation to the administrator of the network, thereby supporting the administrator to solve the failure. do.

Description

A method and apparatus for supporting network failure resolution through remote automatic network analysis in the event of a network failure

The present invention relates to a method and apparatus for supporting network failure resolution through remote automatic network analysis when a network failure occurs.

When a network failure is reported by a user of the network, the administrator analyzes the cause and proceeds with a procedure to solve it. For this purpose, the cause analysis procedure performed by the administrator is largely (i) checking the user's problem device, (ii) pinging the problem device in various directions to check whether it is normal, (iii) when the ping problem is confirmed, related network equipment and (iv) repair the faulty network equipment. The current situation is that such a procedure is performed by a manager directly visiting the site.

However, when several networks are managed together, it is somewhat difficult for an administrator to visit the site one by one to perform the above-described process. Therefore, there is a need for a method to solve network failures while minimizing field visits by managers.

An object of the present invention is to solve the above problems.

Another object of the present invention is to provide a method for supporting network failure resolution through remote automatic network analysis in the event of a network failure, thereby minimizing an administrator's on-site visit and solving the network failure.

Another object of the present invention is to provide a method for selecting an optimal specific network device for performing a network check in order to solve a network connection failure.

Another object of the present invention is to enable an administrator to remotely analyze a network by performing a ping check and a route check in a selected specific network device and transmitting the information to an administrator.

Another object of the present invention is to provide a method for receiving a problem solving request from a user of a network through a chatbot, so that the user's convenience in using a service can be improved.

The characteristic configuration of the present invention for achieving the object of the present invention as described above and for realizing the characteristic effects of the present invention to be described later is as follows.

According to one aspect of the present invention, in a method for supporting network failure resolution through remote automatic network analysis when a network failure occurs, (a) a network manager, when a failure resolution request is obtained from a user of the network, receives the failure resolution request with reference, selecting a specific network device on the network; (b) causing, by the network manager, the specific network device to perform at least one specific analysis operation selected according to the failure resolution request among at least one preset analysis operation; and (c) providing, by the network manager, at least a part of a specific analysis result of the specific analysis operation to the administrator of the network, thereby supporting the administrator to solve the failure. do.

As an example, in step (b), the network manager, when the failure resolution request corresponds to a network connection failure, causes a specific switch selected as the specific network device to perform (i) internal ping check and external ping check a process of acquiring at least one of internal and external connection information by performing at least one of A method is disclosed, comprising performing said specific analysis operation by causing at least a portion of the process to be performed.

As an example, in step (b), when the failure resolution request corresponds to an internal network connection failure, the network manager performs the internal ping check and the internal route check, so that the internal connection information and the internal Acquire connection path information, and when the failure resolution request corresponds to an external network connection failure, perform the external ping check and the external route check to obtain the external connection status information and the internal connection path information A method is disclosed.

As an example, in step (a), the network manager refers to the failure resolution request, and a first specific switch physically connected to a specific terminal having a failure and a second specific switch having the highest packet congestion among switches on the network A method is disclosed, characterized in that at least one of the switches is selected as the specific network equipment.

As an example, in step (a), when the network manager includes information on the specific faulty terminal in the fault resolution request, and there is a valid connection between the specific terminal and the first specific switch, the first A specific switch is selected as the specific network device, and when the specific terminal information is not included in the failure resolution request or there is no valid connection, the second specific switch having the highest packet congestion degree is selected as the specific network device Disclosed is a method characterized in that

As an example, in step (a), the network manager refers to the ARP table, and a method is disclosed, characterized in that the first specific switch physically connected to the specific terminal is found.

As an example, in the step (a), a method is disclosed, wherein the network manager obtains the failure resolution request from the user through a chatbot.

As an example, in step (a), after (i) authentication is performed by the user through the chatbot, (ii) the first degree of association between the specific terminal and the specific terminal is determined according to an instruction of the chatbot. After self-diagnosis is performed using at least some of the neighboring terminals having a threshold of 1 or higher, (iii) (iii-1) the first self-diagnosis result through the specific terminal and (iii-2) the second self-diagnosis through the neighboring terminal Disclosed is a method in which the network manager acquires the failure resolution request by inputting the failure resolution request in response to a second correlation between diagnosis results being equal to or greater than a second threshold.

As an example, in the step (c), the network manager provides at least a part of the specific analysis result to the manager of the network in the form of a detailed report, and sends an acceptance notification indicating that the failure resolution request has been received to the chatbot A method is disclosed for providing to the user of the network via

As an example, in step (a), the network manager (1) inputs at least one piece of data for learning to the natural language processing network in a state in which learning is completed, and then causes the natural language processing network to include at least one data for the training. A process of generating an intermediate processing vector for learning by applying an encoding operation, and (2) causing the data processing network to apply at least one neural network operation to the intermediate processing vector for learning to generate a final information vector for learning, and then the final information for learning When the process of generating a loss by referring to the information vector and the correct answer final information vector and learning at least some of the parameters of the data processing network by backpropagating it is completed, the natural language processing network included in the chatbot , by applying the encoding operation to the user's input to generate an intermediate processing vector for testing, then causing the data processing network included in the chatbot to apply the neural network operation to the intermediate processing vector for testing. Disclosed is a method characterized in that the failure resolution request is obtained by extracting each piece of information as a component of the failure resolution request from the user's input.

According to another aspect of the present invention, there is provided an apparatus for supporting network failure resolution through remote automatic network analysis when a network failure occurs, comprising: one or more memories for storing instructions; and one or more processors configured to perform the instructions, wherein the processor is configured to: (I) select a specific network device on the network with reference to the failure resolution request when a failure resolution request is obtained from a user of the network process; (II) a process of causing the specific network equipment to perform at least one specific analysis operation selected according to the failure resolution request among at least one preset analysis operation; and (III) providing at least a part of a specific analysis result of the specific analysis operation to an administrator of the network to perform a process for supporting the administrator to solve the failure.

As an example, the (II) process may be performed by the processor, when the failure resolution request corresponds to a network connection failure, causes the specific switch selected as the specific network device to perform (i) an internal ping check and an external ping check a process of acquiring at least one of internal and external connection information by performing at least one; and (ii) a process of acquiring at least one of internal and external connection path information by performing at least one of an internal route check and an external route check Disclosed is an apparatus characterized in that the specific analysis operation is performed by causing at least a portion of the analysis operation to be performed.

As an example, in the (II) process, when the failure resolution request corresponds to an internal network connection failure, the processor performs the internal ping check and the internal route check to determine whether the internal connection information and the internal connection Obtaining route information, and performing the external ping check and the external route check when the failure resolution request corresponds to an external network connection failure to obtain the external connection status information and the internal connection path information An apparatus is disclosed.

As an example, in the process (I), the processor refers to the failure resolution request, and a first specific switch physically connected to a specific terminal having a failure and a second specific switch having the highest packet congestion among switches on the network Disclosed is an apparatus, characterized in that selecting at least one of the specific network equipment as the specific network device.

As an example, in the (I) process, when the processor includes information on the specific terminal having the failure in the failure resolution request, and there is a valid connection between the specific terminal and the first specific switch, the first specific Selecting a switch as the specific network device, and selecting the second specific switch with the highest packet congestion degree as the specific network device when the specific terminal information is not included in the failure resolution request or there is no valid connection Disclosed is an apparatus characterized in that.

As an example, in the process (I), the processor searches for the first specific switch physically connected to the specific terminal by referring to the ARP table.

As an example, in the process (I), an apparatus is disclosed, wherein the processor obtains the failure resolution request from the user through a chatbot.

As an example, after the user performs (i) authentication through the chatbot, (ii) according to an instruction of the chatbot, at least one of the specific terminal and adjacent terminals having a first degree of association with the specific terminal equal to or greater than a first threshold After self-diagnosis is performed using a part, (iii) a second correlation between (iii-1) a first self-diagnosis result through the specific terminal and (iii-2) a second self-diagnosis result through the adjacent terminal An apparatus is disclosed, wherein the processor obtains the failure resolution request by inputting the failure resolution request in response to is greater than or equal to a second threshold.

As an example, in the process (III), the processor provides at least a part of the specific analysis result in the form of a detailed report to the manager of the network, and sends an acceptance notification indicating that the failure resolution request has been received to the chatbot and providing to the user of the network via

As an example, in the (I) process, the processor (1) inputs at least one piece of data for training to the natural language processing network in a state where learning is completed, and then causes the natural language processing network to encode at least one of the data for training. A process of generating an intermediate processing vector for learning by applying an operation, and (2) causing the data processing network to generate a final information vector for learning by applying at least one neural network operation to the intermediate processing vector for learning, and then the final information for learning When the process of generating a loss by referring to the vector and the correct answer final information vector and backpropagating it is completed, the process of learning at least some of the parameters of the data processing network is completed, causing the natural language processing network included in the chatbot to: After applying the encoding operation to the user's input to generate an intermediate processing vector for testing, the data processing network included in the chatbot applies the neural network operation to the intermediate processing vector for testing, so that the user Disclosed is an apparatus characterized in that the failure resolution request is obtained by extracting each piece of information as a component of the failure resolution request from the input of .

The present invention provides a method for supporting network failure resolution through remote automatic network analysis when a network failure occurs, thereby minimizing an administrator's on-site visit and enabling the network failure to be resolved.

In addition, the present invention is effective in providing a method for selecting an optimal specific network device for performing a network check in order to solve a network connection failure.

In addition, the present invention has the effect of allowing the administrator to remotely analyze the network by performing a ping check and a route check in a selected specific network device and delivering the information to the administrator.

In addition, the present invention provides a method for receiving a problem solving request from a user of a network through a chatbot, thereby increasing the user's convenience in using the service.

1 is a diagram illustrating a configuration of a network manager performing a network failure resolution support method through remote automatic network analysis when a network failure occurs according to an embodiment of the present invention, and the interworking relationship with the network and chatbot.
2 is an overall flowchart illustrating a method for supporting network failure resolution through remote automatic network analysis when a network failure occurs according to an embodiment of the present invention.
3 is a flowchart illustrating a process of receiving a failure resolution request from a user through a chatbot while performing a network failure resolution support method through remote automatic network analysis when a network failure occurs according to an embodiment of the present invention.
4 is a detailed flowchart illustrating a method for supporting network failure resolution through remote automatic network analysis when a network failure occurs according to an embodiment of the present invention.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS [0010] DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS [0010] DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS [0013] Reference is made to the accompanying drawings, which show by way of illustration specific embodiments in which the present invention may be practiced. These embodiments are described in sufficient detail to enable those skilled in the art to practice the present invention. It should be understood that the various embodiments of the present invention are different but need not be mutually exclusive. For example, certain shapes, structures, and characteristics described herein may be implemented in other embodiments with respect to one embodiment without departing from the spirit and scope of the invention. In addition, it should be understood that the location or arrangement of individual components within each disclosed embodiment may be changed without departing from the spirit and scope of the present invention. Accordingly, the following detailed description is not intended to be taken in a limiting sense, and the scope of the present invention, if properly described, is limited only by the appended claims, along with all scope equivalents to those claimed. Like reference numerals in the drawings refer to the same or similar functions throughout the various aspects.

Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the accompanying drawings in order to enable those of ordinary skill in the art to easily practice the present invention.

1 is a diagram illustrating a configuration of a network manager performing a network failure resolution support method through remote automatic network analysis when a network failure occurs according to an embodiment of the present invention, and the interworking relationship with the network and chatbot.

Referring to FIG. 1 , the network manager 100 may include a communication unit 110 , a memory 115 , and a processor 120 . In this case, the input/output and operation processes of the network manager 100 may be performed by the communication unit 110 and the processor 120, respectively. However, in FIG. 1 , a detailed connection relationship between the communication unit 110 and the processor 120 is omitted. In addition, the memory 115 may be in a state in which various instructions to be described later are stored, and the processor 120 is configured to execute the instructions stored in the memory, thereby performing processes to be described later to carry out the present invention. Even if the network manager 100 is depicted as described above, the case where the network manager 100 includes an integrated processor in which a medium, a processor, and a memory are integrated for implementing the present invention is not excluded.

Such a network manager 100 may operate in conjunction with the network 200 and the chatbot 300 . Here, the network 200 may include a plurality of terminals and a plurality of network equipment, and the network manager 100 manages at least some of them, checks their status, and gives commands to them. It may be in a set state. Here, the first specific switch 210 may be a switch directly connected to the specific terminal 211 and the adjacent terminal 212 having a first degree of association therewith equal to or greater than the first threshold. Also, the second specific switch 220 may be a switch having the highest packet congestion level due to the highest traffic in the network 200 . The chatbot 300 may have a virtual conversation with the user, receive a request for resolving a problem from the user, and transmit a request for a subsequent reception to the user. Unlike the drawing, the chatbot 300 may be implemented in the network manager 100 .

As the configuration of the network manager 100 according to an embodiment of the present invention has been described above, a method for supporting network failure resolution through remote automatic network analysis when a network failure occurs according to an embodiment of the present invention will be specifically described. . For this, refer to FIG. 2 .

2 is an overall flowchart illustrating a method for supporting network failure resolution through remote automatic network analysis when a network failure occurs according to an embodiment of the present invention.

Referring to FIG. 2 , when a failure resolution request is obtained from a user of the network 200 , the network manager 100 may select a specific network device on the network 200 with reference to the failure resolution request ( S01 ). Thereafter, the network manager 100 may cause the specific network device to perform at least one specific analysis operation selected according to a failure resolution request among at least one preset analysis operation ( S02 ). In addition, the network manager 100 may provide at least a part of the specific analysis result of the specific analysis operation to the manager of the network 200 to support the manager to solve the failure ( S03 ). Each step will be described in more detail.

First, the network manager 100 may obtain a failure resolution request from the user through the chatbot 300 . The chatbot 300 basically transmits a question message to the user in a predetermined order, and processes the user's answer accordingly to derive problem resolution request information. For example, the chatbot 300 may initially transmit a message requesting the user to perform authentication, and then transmit a message requesting to perform self-diagnosis and self-action. Thereafter, the chatbot 300 may receive such state information from the user if the network failure is not resolved even after the user has performed self-diagnosis and self-action, and further may obtain information about a specific terminal in which the failure is occurring. . Here, the self-diagnosis process is a process performed using at least some of the specific terminal and the adjacent terminals having a first degree of association with the specific terminal equal to or greater than a first threshold. For example, the adjacent terminal is connected to the same switch as the specific terminal. It may be a connected terminal included in the same subnet. For example, in the case of FIG. 1 , the terminal 212 included in the same subnet as the specific terminal 211 may be included in the adjacent terminal, but the other terminals 221 and 222 are connected to the adjacent terminal to the specific terminal 211 . Considered to be excluded, they may not be used in the self-diagnosis process. Here, the first degree of association may be calculated as an inverse number of the number of switches located between each terminal and a specific terminal. Since a terminal included in the same subnet has one switch in between, it can be calculated as 1, and a terminal included in the immediately adjacent subnet can be calculated as 0.5 because two switches are sandwiched therebetween. In this case, the first threshold may be set to about 0.8. At this time, the user can determine whether the corresponding neighboring terminal has a network failure problem like a specific terminal, and determine whether it is a problem with the entire subnet or only with a specific terminal. That is, when the second correlation between the first self-diagnosis result through the specific terminal 211 and the second self-diagnosis result through the adjacent terminal 212 is equal to or greater than the second threshold, the specific terminal 211 and the adjacent terminal 212 If all shows a network failure, it is determined that the problem is highly likely to be a problem with the entire subnet of the specific switch 210 , and a failure resolution request is received by the user and a subsequent process may be performed. The degree of correlation between the first self-diagnosis result and the second self-diagnosis result is determined by calculating the degree of similarity of the failures occurring in the two terminals (that is, whether they correspond to the same category of failure, whether the overlapping degree of the types of failure is greater than or equal to a predetermined threshold). can be derived. In order to describe such a series of processes, reference is made to FIG. 3 .

3 is a flowchart illustrating a process of receiving a failure resolution request from a user through a chatbot while performing a network failure resolution support method through remote automatic network analysis when a network failure occurs according to an embodiment of the present invention.

Referring to FIG. 3 , the chatbot 300 may conduct a conversation for user authentication. That is, a name, an authentication number, and a network name may be input (S01-1). For example, in the case of SchoolNet, the user is a teacher, the authentication number may be the last digit of a mobile phone number, and the network name may be the name of the corresponding school. Although not shown in the drawing, information on a specific terminal in which a failure has occurred may be obtained in this step. Thereafter, the chatbot 300 performs user authentication using the input information, and then the chatbot 300 may transmit a message so that the user can self-diagnose ( S01-2 ). As in the above example, this may be a message to check whether the adjacent terminal also has a problem. Although omitted in the drawing, if there is no problem in the adjacent terminal, it is highly likely that the problem is in the setting of a specific terminal, and thus an instruction message for this may be transmitted to the user. As described above, if there is a problem in the adjacent terminal, it is highly likely that the problem is in the entire subnet. Therefore, a failure resolution request is received and the remote automatic network analysis process proceeds. Step S01-3 of the drawing will be described later.

When a failure resolution request is received by the chatbot 300 , a specific analysis operation is performed by the network manager 100 . In order to describe this, reference is made to FIG. 4 .

4 is a detailed flowchart illustrating a method for supporting network failure resolution through remote automatic network analysis when a network failure occurs according to an embodiment of the present invention.

Referring to FIG. 4 , when a failure resolution request is received from the chatbot 300 , the network manager 100 may first check the status of each switch in the network 200 ( S02-1 ). Each of the switches may be in a state of being connected in hardware through a web socket so that they can report their respective statuses by periodically communicating with the network manager 100 and receive instructions from the network manager 100 . In this case, the network manager 100 may differently select a specific network device for performing automatic remote analysis according to whether information on a specific terminal having a failure is included in the failure resolution request information (S02-2).

That is, as shown in the figure, when information on a specific terminal having a failure exists in the failure resolution request information, a switch physically connected thereto can be searched for. In this case, when a physically connected switch exists and the switch is effectively connected to a specific terminal because the power is not turned off, the network manager 100 may select the first specific switch as a specific network device. At this time, the first specific switch may be searched with reference to the ARP table. This is because, when physically connected, information on a specific terminal will be included in the ARP table of the corresponding switch. Conversely, if information on a specific terminal does not exist, or if it is not effectively connected to a specific terminal due to a problem such as a switch physically connected thereto is turned off, another second specific switch may be selected as a specific network device. can For example, a second specific switch with the highest packet congestion level may be selected among the switches on the corresponding network 200 due to the highest traffic, which means that the traffic of the faulty specific terminal is the switch with a lot of traffic in the corresponding network 200 . because it is most likely to have passed. In the case of the drawing, it is assumed that the first specific switch 210 is selected.

When a specific network device is selected as described above, the network manager 100 may cause the specific network device to perform a specific analysis operation corresponding to the corresponding category among at least one analysis operation preset for each category of the failure resolution request. have. For example, when the failure resolution request corresponds to a network connection failure category, the network manager 100 may cause a specific network device to perform a ping check and a route check ( S02-3 and S02-4 ). As shown in the figure, this may be an order in which a ping check is performed first and then a root check is performed, but this is only an example, and it is ok to perform either one first or the two processes in parallel. First, in the case of a ping check, the network manager 100 may cause a specific network device to transmit a ping to an internal or external IP. When three pings are transmitted at a time, and no response to them is received, or a response to any one of the three pings is not normal, the ping can be retransmitted up to three times. Thereafter, the result may be stored in the memory 115 . Next, in the case of a route check, the network manager 100 may cause a specific network device to perform a traceroute command to an internal or external IP, and store the information in the memory 115 . At this time, when the traceroute command is normally completed, only information about it is saved, and when the timeout occurs, hop information up to this point may be saved. Unlike the above example, if the category of the failure resolution request is not a network failure, for example, a speed decrease problem, an operation different from the above may be selected and performed.

In the above description, when performing the ping check and the route check, it has been explained that it can be performed on the internal or external IP. to do it That is, the network manager 100 causes a specific network device to perform (i) a process of performing at least one of an internal ping check and an external ping check to obtain at least one of internal and external connectivity information; and (ii) At least one of an internal route check and an external route check may be performed to perform at least a part of a process of acquiring at least one of internal and external connection path information. Here, the internal IP may be used for an internal ping check and an internal route check, and the external IP may be used for an external ping check and an external route check. The network connection failure can be divided into an internal network connection failure and an external network connection failure. The internal network connection failure may be due to a problem inside the network 200, such as file sharing between terminals, and an external network connection failure. , may be due to a problem in communication equipment existing outside the network 200 or at a boundary line such as a gateway, such as not being able to connect to a site outside the network 200 . In case of internal network connection failure, internal ping check and internal route check can be performed to derive internal connection information and internal connection path information. In case of external network connection failure, external ping check and external route check by performing external connection information and external connection path information can be derived. By dividing the cases in this way and performing network analysis, more efficiency can be obtained.

As such, when a specific analysis operation is completed by the network manager 100 and a specific analysis result is derived, the network manager 100, through the chatbot 300, provides information on at least some of them, as shown in FIG. 3 . may be provided to the manager of the network 200 in the form of a detailed report, and a notification of receipt notifying that the failure resolution request has been received may be provided to the user (S01-3). Administrators need detailed information, but users do not need detailed information, so different notifications are sent like this.

Here, an additional configuration of the chatbot 300 will be described. The process of the user inputting information into the chatbot 300 may be a process in which the chatbot 300 simply gives a view and the user selects a specific view from among them, but it is a process in which a natural language processing algorithm is adopted for the convenience of the user. It may be possible, so I will explain it. For this, two neural networks are needed, one is a natural language processing network and the other is a data processing network. Natural language processing networks can use algorithms that are widely known as de facto standards in the field of natural language processing, such as BERT, and according to popular usage of BERT, BERT has already been trained and can extract information by decomposing sentences by itself. can In this situation, the programmer can add a DNN to the output stage so that the desired operation is performed, which is the data processing network in the present invention. That is, the data processing network may be a network that reprocesses the basically extracted information and divides it into vectorized information according to the template of the failure resolution request, that is, category information of the failure resolution request and specific terminal information in which the problem occurred. As the structure of the data processing network, any structure of a well-known DNN may be used.

Before the above-described process of the present invention is performed, the learning device (not shown) inputs at least one piece of data for learning into the natural language processing network in a state in which learning is completed, and then causes the natural language processing network to include at least one of the data for learning. A process of applying an encoding operation to generate an intermediate processing vector for learning, and (2) causing the data processing network to apply at least one neural network operation to the intermediate processing vector for learning to generate a final information vector for learning, then the final information vector for learning and a process of learning at least some of the parameters of the data processing network by generating a loss with reference to the correct answer final information vector and backpropagating it. Here, the correct answer final information vector may include information as each component of the above-described failure resolution request, for example, category information of the failure resolution request and information on the terminal in which the failure occurs as each component of the vector. Accordingly, when learning is completed, the network manager 100 causes the natural language processing network included in the chatbot 300 to generate an intermediate processing vector for testing by applying an encoding operation to the user's input, and then the chatbot 300 Troubleshooting by extracting each information as a component of the failure resolution request from the user's input, that is, category information and information on the terminal in which the failure occurred, by allowing the data processing network included in the to apply a neural network operation to the intermediate processing vector for testing request can be obtained. In this way, the user can more conveniently input a failure resolution request through the chatbot 300 .

The embodiments according to the present invention described above may be implemented in the form of program instructions that can be executed through various computer components and recorded in a computer-readable recording medium. The computer-readable recording medium may include program instructions, data files, data structures, etc. alone or in combination. The program instructions recorded on the computer-readable recording medium may be specially designed and configured for the present invention, or may be known and available to those skilled in the computer software field. Examples of the computer-readable recording medium include hard disks, magnetic media such as floppy disks and magnetic tapes, optical recording media such as CD-ROMs and DVDs, and magneto-optical media such as floppy disks. media), and hardware devices specially configured to store and execute program instructions, such as ROM, RAM, flash memory, and the like. Examples of program instructions include not only machine language codes such as those generated by a compiler, but also high-level language codes that can be executed by a computer using an interpreter or the like. The hardware device may be configured to operate as one or more software modules for carrying out the processing according to the present invention, and vice versa.

In the above, the present invention has been described with specific matters such as specific components and limited embodiments and drawings, but these are provided to help a more general understanding of the present invention, and the present invention is not limited to the above embodiments. , those of ordinary skill in the art to which the present invention pertains can devise various modifications and variations from such a machine.

Therefore, the spirit of the present invention should not be limited to the above-described embodiments, and not only the claims described below, but also all modifications equivalently or equivalently to the claims described below belong to the scope of the spirit of the present invention. will do it

Claims (20)

A method for supporting network failure resolution through remote automatic network analysis when a network failure occurs, the method comprising:
(a) selecting, by a network manager, a specific network device on the network by referring to the failure resolution request when a failure resolution request is obtained from a user of the network;
(b) causing, by the network manager, the specific network device to perform at least one specific analysis operation selected according to the failure resolution request among at least one preset analysis operation; and
(c) providing, by the network manager, at least a part of a specific analysis result of the specific analysis operation to a manager of the network, helping the manager to solve the failure
including,
The step (a) is,
The network manager, referring to the failure resolution request, selects at least one of a first specific switch physically connected to a specific terminal having a failure and a second specific switch having the highest packet congestion among switches on the network as the specific network device. choose,
The step (a) is,
The network manager selects the first specific switch as the specific network device when the fault resolution request includes information on the specific terminal having the fault and there is a valid connection between the specific terminal and the first specific switch; , when the failure resolution request does not include information on the specific terminal or there is no valid connection, selecting the second specific switch having the highest packet congestion as the specific network device,
The step (a) is,
the network manager obtains the failure resolution request from the user through the chatbot;
The step (a) is,
After the user performs (i) authentication through the chatbot, (ii) uses at least some of the specific terminal and the neighboring terminals having a first degree of association with the specific terminal equal to or greater than a first threshold according to an instruction of the chatbot to perform self-diagnosis, (iii) (iii-1) the first self-diagnosis result through the specific terminal and (iii-2) the second correlation between the second self-diagnosis result through the adjacent terminal The method according to claim 1, wherein the network manager obtains the failure resolution request by inputting the failure resolution request according to a threshold value or more. According to claim 1,
The step (b) is,
When the failure resolution request corresponds to a network connection failure, the network manager causes the specific switch selected as the specific network device to perform at least one of (i) an internal ping check and an external ping check to connect internal and external connections by performing at least part of a process of obtaining at least one of information on whether or not A method comprising performing an analysis operation. 3. The method of claim 2,
The step (b) is,
When the failure resolution request corresponds to an internal network connection failure, the network manager performs the internal ping check and the internal route check to obtain internal connection status information and internal connection path information, and the failure resolution request is In response to an external network connection failure, the external ping check and the external route check are performed to obtain the external connection status information and the internal connection path information. delete delete According to claim 1,
The step (a) is,
The method, characterized in that the network manager, by referring to the ARP table, finds the first specific switch physically connected to the specific terminal. delete delete According to claim 1,
Step (c) is,
providing, by the network manager, at least a part of the specific analysis result to the manager of the network in the form of a detailed report, and providing an acceptance notification indicating that the failure resolution request has been received to the user of the network through the chatbot; How to characterize. According to claim 1,
The step (a) is,
After the network manager (1) inputs at least one piece of data for learning into the natural language processing network in a state where learning is completed, the natural language processing network applies at least one encoding operation to the data for learning to generate an intermediate processing vector for learning and (2) causing the data processing network to generate a final information vector for learning by applying at least one neural network operation to the intermediate processing vector for learning, and then referencing the final information vector for learning and the final information vector for correct answer When the process of learning at least some of the parameters of the data processing network by generating a loss and backpropagating it is completed, the natural language processing network included in the chatbot performs the encoding operation on the user's input. After adding and generating the intermediate processing vector for the test, the data processing network included in the chatbot applies the neural network operation to the intermediate processing vector for the test, from the user's input to the failure resolution request. A method, characterized in that the failure resolution request is obtained by extracting each piece of information as a component. In an apparatus for supporting network failure resolution through remote automatic network analysis when a network failure occurs,
one or more memories storing instructions; and
one or more processors configured to perform the instructions, wherein the processor is configured to: (I) when a failure resolution request is obtained from a user of the network, referring to the failure resolution request, a process of selecting a specific network device on the network ; (II) a process for causing the specific network device to perform at least one specific analysis operation selected according to the failure resolution request among at least one preset analysis operation; and (III) providing at least a part of a specific analysis result of the specific analysis operation to an administrator of the network to perform a process for supporting the administrator to solve the failure,
The (I) process is
The processor selects, as the specific network device, at least one of a first specific switch physically connected to a specific terminal having a failure and a second specific switch having the highest packet congestion among switches on the network with reference to the failure resolution request do,
The (I) process is
the processor selects the first specific switch as the specific network device when the fault resolution request includes information on the specific terminal in which the fault has occurred and there is a valid connection between the specific terminal and the first specific switch; When the failure resolution request does not include the specific terminal information or there is no valid connection, the second specific switch having the highest packet congestion degree is selected as the specific network device,
The (I) process is
the processor obtains the failure resolution request from the user through the chatbot;
After the user performs (i) authentication through the chatbot, (ii) uses at least some of the specific terminal and the neighboring terminals having a first degree of association with the specific terminal equal to or greater than a first threshold according to an instruction of the chatbot to perform the self-diagnosis, and then (iii) (iii-1) the first self-diagnosis result through the specific terminal and (iii-2) the second correlation between the second self-diagnosis result through the adjacent terminal The apparatus according to claim 1, wherein the processor obtains the failure resolution request by inputting the failure resolution request according to a threshold value or more. 12. The method of claim 11,
The (II) process is
When the failure resolution request corresponds to a network connection failure, the processor causes the specific switch selected as the specific network device to (i) perform at least one of an internal ping check and an external ping check to determine whether internal and external connections are made and (ii) performing at least one of an internal route check and an external route check to obtain at least one of the internal and external connection path information by performing at least part of a process of acquiring at least one of information A device for performing an action. 13. The method of claim 12,
The (II) process is
When the failure resolution request corresponds to an internal network connection failure, the processor performs the internal ping check and the internal route check to obtain the internal connection status information and the internal connection path information, and the failure resolution request In response to the external network connection failure, the external ping check and the external route check are performed to obtain the external connection status information and the internal connection path information. delete delete 12. The method of claim 11,
The (I) process is
The processor refers to the ARP table, the device characterized in that it finds the first specific switch physically connected to the specific terminal. delete delete 12. The method of claim 11,
The (III) process is
providing, by the processor, at least a part of the specific analysis result in the form of a detailed report to the manager of the network, and providing an acceptance notification indicating that the failure resolution request has been received to the user of the network through the chatbot device characterized. 12. The method of claim 11,
The (I) process is
After the processor (1) inputs at least one training data to the natural language processing network in a state where learning is completed, the natural language processing network applies at least one encoding operation to the training data to generate an intermediate processing vector for learning and (2) causing the data processing network to generate the final information vector for learning by applying at least one neural network operation to the intermediate processing vector for learning, and then referencing the final information vector for learning and the final correct information vector for loss In a state in which the process of learning at least some of the parameters of the data processing network by generating and backpropagating a After generating the intermediate processing vector for testing, the data processing network included in the chatbot applies the neural network operation to the intermediate processing vector for testing, so as to configure the failure resolution request from the user's input An apparatus, characterized in that the failure resolution request is obtained by extracting each piece of information as an element.

Images