Method and system for detecting failure-inducing clients using failure-response edge server grouping

Abstract

To relate to detection of clients that induce network failures, and in particular provide a client terminal connection control method, system, and storage medium for detecting clients that induce network failures through grouping of failure response edge servers.SOLUTION: A method includes the steps of assigning communication processing of a first client group consisting of a plurality of client terminals to an edge server, assigning communication processing of a second client group including at least one client terminal belonging to the first client group to a failure-handling edge server when a failure occurs in the operation of the edge server, and determining whether a failure occurs in the operation of the failure-handling edge server.SELECTED DRAWING: Figure 17

Description

[Technical field]

[0001] The present invention relates to detection of clients that induce network failures, and more particularly to a method and system for detecting clients that induce network failures using failure response edge server grouping, which detects clients that induce network failures through grouping of failure response edge servers. [Background technology]

[0002] Detecting clients that cause network failures using the client route control system, for example identifying the IP address of a DDoS attack initiator, means detecting and identifying DDoS attacks by controlling the client route by specifying the edge server to which the user connects using multiple edge servers, client route control servers, and DNS (Domain Name Server).

[0003] Generally, DDoS attacks can be divided into two types. The first type is when a server is overwhelmed with too much work to handle, paralyzing it. When this type of attack occurs, the server is overwhelmed with work and cannot process other tasks, causing it to become paralyzed.

[0004] The second type is to flood the network line with traffic, rendering the line itself unusable from the start. This type of attack does not affect the server, but the line connected to the server cannot be used, making it impossible for the server and clients to communicate. In other words, even if there is no problem with the server, problems occur in the network, making it difficult to maintain service. In response to this situation, many DDoS security technologies have been developed, but the reality is that they do not provide a fundamental solution. Conventional DDoS defense technologies can defend against DDoS attacks themselves, but they cannot search for and identify the IP address of the attacker that is the main source of the attack.

[0005] Furthermore, if the line or IP band where the attack occurred is blocked in order to prevent a DDoS attack, normal users of the line or band cannot use the service either, which causes a serious cost and time loss for financial institutions, public institutions, game service providers, etc., who must maintain services 24 hours a day.

[0006] To solve this problem, we invented a method to detect and detect clients that cause network failures by combining several routes and providing each client with only one unique route by using edge servers placed in multiple stages, and we have registered a patent for this (Korean Patent 10-1569857, 2015.11.11).

[0007] However, the fault-inducing client detection method utilizing a client route control system, which is a registered patent, is not significantly different from the present invention when there is only one client inducing a network fault, but has the problem that there is a possibility of false detection when multiple clients cause network faults simultaneously.

[0008] For example, according to the conventional patent, when an edge server IP allocation metric for finding out a client causing a failure is configured in a 3x2 matrix arrangement for an edge server group consisting of six edge servers, it can be configured as shown in FIG. 11. The edge server group has six edge servers, from edge server 1 to edge server 6. When an edge server is assigned to a client using a 3x2 matrix by using the six edge servers, eight clients can be identified, from clients A to H. That is, edge server 1 is assigned to clients A to D, and edge server 4 is assigned to clients E to H. Then, when edge server 1 fails, edge server 2 is assigned to clients A and B, and edge server 5 is assigned to clients C and D, according to the edge server IP allocation metric. When edge server 5 fails, edge server 3 is assigned to client C, and edge server 6 is assigned to client D, according to the edge server IP allocation metric. After that, when a failure occurs in client 6, client D is detected as a client inducing a failure.

[0009] However, when clients A and F cause failures at the same time as shown in Figure 12, edge server IP allocation metrics indicate that edge server 1 and edge server 4 will fail first, then edge server 2 will fail, and then edge server 3 and edge server 6 will fail. Ultimately, when clients A and F cause failures at the same time, edge servers 1, 2, 3, 4, and 6 will fail. In this case, client B, who is assigned edge servers 1, 2, and 6, and client E, who is assigned edge servers 4, 2, and 3, will also be detected as failure-inducing clients. In other words, although the clients that actually cause failures are clients A and F, a problem occurs in that clients B and E are also recognized as failure-inducing clients. Summary of the Invention [Problem to be solved by the invention]

[0010] The problem to be solved by the present invention has been devised to solve the above-mentioned problems of the prior art, and is to provide a method and system for detecting a fault-inducing client using a fault response edge server grouping, which detects a client that induces a fault by detecting a fault-inducing client that induces a network fault and preventing a service server from being further affected by a service fault due to a network fault, by replacing the failed edge server with a fault response edge server sub-group selected from a fault response edge server group consisting of a plurality of fault response edge servers when a fault occurs in the replaced fault response edge server, and by repeating this process to replace it with another fault response edge server sub-sub-group when a fault occurs in the replaced fault response edge server. [Means for solving the problem]

[0011] A method for detecting a failure-inducing client using a failure handling edge server group according to the present invention for solving the technical problem includes the steps of: checking for a failure of an edge server in a network in which a plurality of edge servers are present on a path by which a client approaches a service server; when a failure occurs in an edge server, providing a number of failure handling edge servers (failure handling edge server sub-groups) selected from a failure handling edge server group consisting of a plurality of failure handling edge servers, which is greater than the number of the edge server in which the failure occurred, to a client connected to the failure-inducing edge server in place of the edge server in which the failure occurred; when a failure occurrence check shows that a failure has occurred in the substituted failure handling edge server, providing a failure handling edge server sub-group selected from the failure handling edge server group to a client in place of the failure-inducing edge server in which the failure occurred; and whenever a failure occurs in a failure handling edge server, substituting the failure-inducing edge server sub-group selected from the failure handling edge server group for the failure-inducing edge server in which the failure occurred, and when only one client is assigned to the failure handling edge server in which the failure occurred, detecting the one client as a failure-inducing client.

[0012] The failure handling edge server subgroup is characterized in that, when a failure occurs in an edge server or a failure handling edge server, the failure handling edge server subgroup is composed of a plurality of failure handling edge servers based on the number of failure handling edge servers not assigned from the failure handling edge server group by analyzing the number of clients assigned to the edge server or failure handling edge server and connection records.The failure occurrence check of the edge server or failure handling edge server is characterized in that it is performed by monitoring the occurrence of a failure of the edge server or failure handling edge server when a DNS query is received from a client or regardless of reception of a DNS query from a client.

[0013] The method for detecting a failure-induced client using a failure response edge server group according to the present invention is characterized in that, when a failure occurs in an edge server or a failure response edge server, a client IP or user information using the failed edge server or the failure response edge server is searched for, and when a DNS query request is received, a failure response edge server selected from the failure response edge server group is provided to the searched client.

[0014] The client includes an agent that sends a DNS query to connect to a failure response edge server selected from a failure response edge server group when a failure occurs in an edge server or a failure response edge server, and the agent requests a DNS query including user information, and upon receiving the DNS query from the agent of the client, extracts the user information from the DNS query of the client, and provides the client with a failure response edge server IP corresponding to the extracted user information selected from the failure response edge server group.

[0015] A failure-inducing client detection system using a failure handling edge server group according to the present invention for solving the above technical problem includes an edge server group including a plurality of edge servers present on a path where a client accesses a service server, a failure handling edge server present on a path where a client accesses a service server and which substitutes for the edge server where a failure occurs when the edge server fails, and another failure handling edge server which substitutes for the failure handling edge server, and when a failure occurs in an edge server or the failure handling edge server, a failure handling edge server subgroup including a number of failure handling edge servers greater than the number of the edge server where the failure occurs or the failure handling edge servers. a DNS control unit that provides an edge server IP or a failure-response edge server IP according to a client IP or user information in response to a DNS query requested by a client; a failure occurrence check unit that checks the occurrence of a failure in an edge server and a failure-response edge server; and an edge server control unit that, when a failure occurs in an edge server or a failure-response edge server, provides the failure-response edge server subgroup selected from the failure-response edge server group to a client in place of the edge server or failure-response edge server in which the failure has occurred, and, when there is only one client assigned to the failure-response edge server, detects the one client as a failure-inducing client.

[0016] The fault occurrence check unit checks whether there is a fault in the edge server or fault response edge server provided to the DNS control unit by the edge server control unit in response to a DNS query, and at this time, if there is no fault in the provided edge server or fault response edge server, the DNS control unit provides the provided edge server or fault response edge server information to a client via a DNS server.

[0017] The fault occurrence check unit monitors faults in an edge server or a fault response edge server regardless of a DNS query, and when a fault occurs in the edge server or the fault response edge server, the edge server control unit searches for client IP or user information using the edge server or the fault response edge server in which the fault has occurred, and assigns a fault response edge server IP selected from the fault response edge server group to the searched client IP or user information, and when a client corresponding to the searched client IP or user information requests a DNS query, the DNS control unit provides the fault response edge server IP assigned by the edge server control unit to the searched client via a DNS server, or provides the fault response edge server IP assigned by the edge server control unit to the searched client via a DNS server regardless of a DNS query.

[0018] The client includes an agent that transmits a DNS query including user information to reconnect to a failure handling edge server selected from the failure handling edge server group when a failure occurs in the edge server or failure handling edge server, and the edge server control unit extracts user information from the DNS query of the client when the edge server control unit receives a DNS query from the agent of the client. The edge server or failure handling edge server connects the client and server, and includes at least one of a program, a server, and hardware equipment having a service relay function or a service function.

[0019] In addition, a client terminal connection control method performed by a client terminal connection control device according to the present invention includes a step of assigning communication processing of a first client group consisting of a plurality of client terminals to an edge server, and when a failure occurs in the operation of the edge server, a step of assigning communication processing of a second client group generated by dividing the first client group to a failure-response edge server, and a step of determining whether a failure occurs in the operation of the failure-response edge server.

[0020] When a failure occurs in the operation of the failure response edge server and multiple client terminals are assigned to the failure response edge server, the communication processing of a sub-client group generated by dividing the client group assigned to the failure response edge server can be assigned to the failure response edge server.

[0021] When a failure occurs in the operation of the failure handling edge server and there is one client terminal assigned to the failure handling edge server, the one client terminal may be determined as a failure-inducing client terminal.

[0022] In addition, a client terminal connection control device according to the present invention includes a communication unit that performs data communication with client terminals, an edge server, and a fault handling edge server, and a processor that assigns communication processing of a first client group consisting of a plurality of client terminals to the edge server, and, if a fault occurs in the operation of the edge server, assigns communication processing of a second client group generated by dividing the first client group to the fault handling edge server, and determines whether a fault will occur in the operation of the fault handling edge server.

[0023] When a failure occurs in the operation of the failure handling edge server and a plurality of client terminals are assigned to the failure handling edge server, the processor can assign, to the failure handling edge server, communication processing of a sub-client group generated by dividing the client group assigned to the failure handling edge server.When a failure occurs in the operation of the failure handling edge server and a single client terminal is assigned to the failure handling edge server, the processor can determine the single client terminal as a failure induction client terminal.

[0024] The present invention also provides a program executed by a processor to carry out the above-described invention, and a recording medium that can be read by the processor and has the program recorded thereon.

[0025] According to the method and system for detecting a fault-inducing client using a fault handling edge server grouping according to the present invention, a fault-inducing client that induces a network fault is detected, and the service server is prevented from being affected by a service fault caused by the network fault any more. Therefore, the continuity of the service provided to the client can be guaranteed, and the client can continuously maintain network communication through the fault handling edge server even if a fault occurs in a part of the path.

[0026] According to the present invention, the edge server where a failure has occurred and the failure response edge server are replaced by the failure response edge server subgroup, and the client inducing the failure is accurately detected, so that the IP address of the DDoS attack initiator can be easily identified and the DDoS attack can be blocked. That is, by dynamically allocating the edge server or the failure response edge server subgroup to be used for each client, when a failure occurs in the edge server or the failure response edge server, it is possible to immediately search only for the client that caused the failure. As a result, by blocking only the client that caused the failure, it does not affect users who are using the service normally. Also, the time required to analyze logs for DDoS detection is reduced, so time loss can be reduced.

[0027] According to the present invention, the route is specified based on the client information, so it is versatile. In other words, since the present invention specifies the route based on the client information, whatever method the client uses, it is always applied to the routing system and the appropriate route is used. This is much more versatile than the ACL or Null Routing technique that filters only by IP or port.

[0028] In addition, the present invention can defend against various types of DDoS attacks. By using an edge server or a fault response edge server, the present invention prevents clients from directly connecting to the server. This means that the DDoS attack initiator cannot directly interfere with the server, and therefore cannot directly launch a DDoS attack on the server. Even if the DDoS attack initiator launches an attack on a network line, the relevant line is connected only to the DDoS attack initiator, so it is possible to block the line and defend without affecting other lines. Since the DDoS attack initiator can be found with a small number of edge servers, the cost of maintaining security is reduced, and the manpower and time loss required for the equipment and management required for the same is significantly reduced.

[0029] In addition, the edge server and the failure response edge server according to the present invention may be embodied with substantially the same hardware / software configuration, and a system administrator may adaptively manage the configuration of the edge server and the failure response edge server according to the network environment. More specifically, some of the failure response edge servers may be used as edge servers to provide sufficient traffic to client terminals connected to the network. Similarly, when it is estimated that a large number of failure-inducing clients are included among the client terminals connected to the network, some of the edge servers may be converted and used as failure response edge servers for network stability, thereby quickly selecting the failure-inducing clients included in the client terminals.

[0030] In this way, the method for detecting a fault-inducing client according to the present invention can provide a fast and stable environment for detecting a fault-inducing client, since the time complexity required for searching for a fault-inducing client connected to one edge server can be calculated as O(logSn), where step S is the number of fault-handling edge servers and n is the number of clients connected to the edge server. [Brief description of the drawings]

[0031] [Figure 1] 1 is a block diagram showing an example of the configuration of an entire system to which the present invention is applied. [Diagram 2] 1 is a block diagram illustrating a configuration of a system for detecting a fault-inducing client using a fault handling edge server grouping according to an embodiment of the present invention. [Diagram 3] 1 is a flowchart illustrating a method for detecting a failure-inducing client using a failure handling edge server grouping according to an embodiment of the present invention. [Figure 4]This is a flowchart showing one embodiment of detecting a failure-inducing client by checking whether the edge server or failure-response edge server has failed when a DNS query is received from a client, and replacing the failed edge server or failure-response edge server with a failure-response edge server subgroup. [Diagram 5] This is a flowchart showing one embodiment of detecting a failure-inducing client by checking whether the edge server or failure-response edge server has failed regardless of receiving a DNS query from a client, and replacing the failed edge server or failure-response edge server with a failure-response edge server subgroup. [Figure 6] 1 illustrates an edge server allocation scheme according to one embodiment of the present invention. [Figure 7] This shows a case where client 1 becomes a client causing a problem due to the influence of malicious code or hacking. [Figure 8] This shows that clients 1, 4, and 7 are assigned to failure response edge server subgroup 1 so that clients connected to edge server 1 under attack can be served by a failure response edge server. [Figure 9] This shows that client 1 and client 4 are separated and reassigned to failure-response edge server subgroup 2. [Figure 10] This shows the process of recognizing a client that is connected to one fault-handling edge server and attacks that fault-handling edge server as a client causing the fault. [Figure 11] FIG. 1 is a diagram illustrating a method for detecting a client that induces a failure, which utilizes a client path control system that is a conventional registered patent. [Figure 12] FIG. 1 is a diagram illustrating a method for detecting a client that induces a failure, which utilizes a client path control system that is a conventional registered patent. [Figure 13] 1 is a diagram for explaining a failure-inducing client detection method using grouping of failure response edge servers according to the present invention; [Figure 14] 1 is a diagram for explaining a failure-inducing client detection method using grouping of failure response edge servers according to the present invention; [Figure 15] 1 is a diagram for explaining a failure-inducing client detection method using grouping of failure response edge servers according to the present invention; [Figure 16] 1 is a diagram for explaining a failure-inducing client detection method using grouping of failure response edge servers according to the present invention; [Figure 17] FIG. 1 is a diagram illustrating a method for detecting a failure-inducing client according to an embodiment. [Figure 18] FIG. 1 is a diagram illustrating a method for detecting a failure-inducing client according to an embodiment. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0032] Hereinafter, a preferred embodiment of the present invention will be described in detail with reference to the accompanying drawings. The embodiment described in this specification and the configurations shown in the drawings are merely preferred embodiments of the present invention and do not represent the entire technical idea of ​​the present invention, so it should be understood that there are various equivalents and modifications that can replace them at the time of filing this application.

[0033] 1 is a block diagram showing an example of the configuration of an entire system to which the present invention is applied. The system to which the present invention is applied includes a plurality of clients 100, a DNS server 110, a client route control server 120, a service relay network 130, and a service server 140.

[0034] In the present invention, the service relay network 130 may be composed of an edge server group 132 consisting of a plurality of edge servers having a service relay function or including a server or hardware equipment having a service function, and network equipment such as a gateway, a router, a switch, a hub, etc. In addition, the service relay network 130 includes a failure response edge server group 134 that detects a failure-inducing client while taking the place of the edge server where the failure occurs when a failure occurs in the edge server.

[0035] Generally, a network can include various servers such as a web server and DNS, and network equipment such as a gateway, router, switch, and hub. Among these, the server mainly performs the function of receiving a request from a client and resolving the request, and the network equipment performs the function of transmitting packets sent and received by the client. Such equipment transmits packets received from the client to the server via a route with a fast transmission time without loss, and the server quickly processes the client request and transmits it to the client. However, such equipment is vulnerable as it lacks the ability to deal with problems when problems occur in the equipment or the line.

[0036] The edge server used in the present invention may be realized by a server and hardware equipment having a service relay function or a service function, such as a proxy server, a cache server, etc., for use in implementing the present invention. Also, the edge server may be a program having a service relay function or a service function, and may be used as a service path to which a client connects and function as a service server to which a client connects.

[0037] When a client 100 connects to a server, most clients use DNS to connect to a service server 140. The present invention includes an edge server, a failure response edge server, and a client route control server (hereinafter referred to as CRCS) 120 that manages them.

[0038] When the client 100 connects to the service server 140 to receive a service, the client 100 is first provided with the IP address of the service server 140 via the DNS server 110. To this end, the client 100 transmits a DNS query to the DNS server 110.

[0039] When the DNS server 110 receives a DNS query from the client 102 or 104 and the IP address of the corresponding service server is not stored, the DNS server 110 transmits the DNS query to the client routing server 120 .

[0040] The client route control server 120 receives a DNS query from the DNS server 110, assigns and provides an IP of an edge server belonging to the service relay network 130 to each client, and transmits the assigned edge server IP to the client via the DNS server 110.

[0041] To explain the client route control server (CRCS) 120 in more detail, the client route control server (CRCS) 120 checks the status of edge servers and failure handling edge servers, and manages data of clients, edge servers, and failure handling edge servers. It also designates and transmits edge server IP or failure handling edge server IP for each client. The CRCS 120 can receive a DNS request, check the status of edge servers and failure handling edge servers, designate the domain name and client IP or edge server IP corresponding to the client IP, store the designated information, and send a DNS response.

[0042] The client 100 that has requested the DNS query receives the IP information of the edge server or the failure handling edge server IP, and connects to the edge server or the failure handling edge server to receive a service from the service server 140 .

[0043] More specifically, the client 100 requests a DNS query to the DNS server 110 to find the address of the server. If the DNS server 110 does not have the domain name requested by the client in its own server, it requests a higher-level DNS server to search for an IP address corresponding to the domain name.

[0044] To explain the DNS server 110 in more detail, the DNS server 110 may be a general DNS device that receives a domain name requested by a client and responds with an IP corresponding to the domain name, and may include the corresponding concept and technology. The DNS server 110 transmits an IP corresponding to the domain name to the client. Through the above-mentioned process, the DNS query comes to the CRCS 120 having the domain name and client IP requested by the client 100. At this time, the CRCS 120 prepares the IP of the edge server or the IP of the failure response edge server used for the specific route for each client IP according to the routing algorithm.

[0045] 2 is a block diagram showing the configuration of a system for detecting a failure-induced client using a grouping of failure-response edge servers according to an embodiment of the present invention. Referring to FIG. 2, the system for detecting a failure-induced client using a grouping of failure-response edge servers according to an embodiment of the present invention includes an edge server group 260, a group of failure-response edge servers 270, and a client route control server 20.

[0046] The edge server group 260 includes multiple edge servers 260-1, 260-n that exist on the path that the client 210 takes to access the service server 280, and is a collection of edge servers that monitor traffic between the client and the service server and deliver packets from the client to the service server. An edge server can accept connections from one or more clients, and when a client first connects to a service server, an edge server in the edge server group is assigned to the client.

[0047] The failure handling edge server group 270 is located on the path by which the client 210 approaches the service server 280, and consists of failure handling edge servers 270-1, 270-m, 270-nm, 270-n that substitute for the edge server where the failure occurred when a failure occurs in the edge server 260-1, 260-n, and other failure handling edge servers 270-1, 270-m, 270-nm, 270-n that substitute for the failure handling edge server when a failure occurs in an edge server or failure handling edge server.In the event that a failure occurs in an edge server or failure handling edge server, the failure handling edge server subgroup 275 consists of failure handling edge servers 270-1, 270-m that is greater than the number of the edge server where the failure occurred or the failure handling edge servers.

[0048] The failure-response edge server group 270 is a collection of failure-response edge servers that are assigned to search for failure-induced clients when a client that was assigned to and used an edge server that has failed is reassigned an edge server. A failure-response edge server can accept connections from one or more clients, and a failure-response edge server is assigned when a client that was assigned to an edge server that has failed by the failure occurrence check unit 240 of the client route control server 20 in a general situation requests an edge server assignment.

[0049] When a failure handling edge server actually performs failure handling, the entire failure handling edge server group 270 is not processed, but a failure handling edge server subgroup is logically generated through a failure handling edge server subgroup control mechanism in the edge server control unit 250 of the client route control server 20, the failure handling edge server is included in the failure handling edge server subgroup, and then the failure handling edge server in the generated failure handling edge server subgroup 275 is assigned.

[0050] The failure handling edge server subgroup 275 is a virtual failure handling edge server group logically generated within the failure handling edge server group 270. When an edge server actually fails and the failure handling edge server starts to handle the failure, the entire failure handling edge server group is not assigned to the target. Instead, a virtual failure handling edge server subgroup is created in the failure handling edge server subgroup control mechanism of the edge server control unit 250, the number of clients already assigned to the edge server where the failure occurred and the connection records are analyzed, and an appropriate number is selected and assigned to the virtual failure handling edge server subgroup taking into consideration the number of failure handling edge servers that were not assigned from the failure handling edge server group, and then the number of clients selected and assigned to the virtual failure handling edge server subgroup are assigned to the failure handling edge server subgroup 275 generated via the edge server grouping assignment mechanism.

[0051] The fault-handling edge server subgroup, for which the fault-handling process is completed, is deleted, and the fault-handling edge servers assigned to it are left unassigned in the fault-handling edge server group. The fault-handling edge server subgroup can be realized by a virtual group that is generated only when necessary during edge server fault handling and deleted when the process is completed.

[0052] The edge server and the failure handling edge server are servers that play a role in controlling traffic between the client and the service server, and transmit connection information between the client 210 and the service server 280, system information of the edge server itself (CPU, memory, network, storage usage, usage rate, etc.), failure occurrence related information, etc. to the client route control server 20. In a general situation, a client is connected to the service server 280 via an edge server, and in a failure occurrence state, the client is connected to the service server 280 via the failure handling edge server. The edge server and the failure handling edge server have the same configuration.

[0053] The client route control server 20 controls the communication path between the client and the service server so that the communication path can proceed through the edge server, assigns an edge server and a failure handling edge server in response to a client request via DNS, receives communication-related information (connection, system status, traffic-related information) from the edge server and the failure handling edge server, checks whether a failure has occurred, and detects a failure-inducing client. The client route control server 20 also manages a blacklist of failure-inducing clients, blocks access to the service server in response to a DNS request from a client registered in the blacklist via DNS, and includes a DNS control unit 230, a failure occurrence check unit 240, and an edge server control unit 250. As a result, the client route control server 20 can also function as a client terminal connection control device that controls the connection of the client terminal.

[0054] The DNS control unit 230 provides an edge server IP or a fault handling edge server IP according to a client IP or user information in response to a DNS query requested by the client 100. That is, the DNS control unit 230 receives a DNS request of a client from the DNS server 220, requests the client route control server 20 to allocate an edge server, and transmits the result of the edge server control unit 250 to the DNS server 220.

[0055] The failure occurrence check unit 240 checks the occurrence of a failure in an edge server and a failure handling edge server, and receives communication-related information such as connection-related information, system status information, and traffic-related information from the edge server and the failure handling edge server to check whether a failure has occurred in the corresponding edge server (including the failure handling edge server). When the failure occurrence check unit 240 determines that a failure has occurred in the corresponding edge server (including the failure handling edge server) according to the failure-related policy, it transmits the details to the edge server control unit 250.

[0056] In response to a DNS query, the fault occurrence check unit 240 can check whether the edge server or fault handling edge server provided to the DNS control unit 230 by the edge server control unit 250 has a fault. If the provided edge server or fault handling edge server has no fault, the DNS control unit 230 provides the provided edge server or fault handling edge server information to the client 210 via the DNS server 220.

[0057] In addition, the failure occurrence check unit 240 can monitor the failure of an edge server or a failure handling edge server regardless of a DNS query, and in this case, when a failure occurs in the edge server or the failure handling edge server, the edge server control unit 250 searches for client IP or user information using the edge server or the failure handling edge server in which the failure occurs, and assigns a failure handling edge server IP selected in the failure handling edge server group 270 to the searched client IP or user information. In this case, when a client corresponding to the searched client IP or user information requests a DNS query, the DNS control unit 230 can provide the failure handling edge server IP allocated by the edge server control unit 250 to the searched client 100 via the DNS server 220. In addition, the DNS control unit 230 can provide the failure handling edge server IP allocated by the edge server control unit 250 to the searched client 100 via the DNS server 220 regardless of the DNS query.

[0058] When a failure occurs in an edge server or a failure handling edge server, the edge server control unit 250 provides the failure handling edge server subgroup 275 selected from the failure handling edge server group 270 to a client instead of the edge server or failure handling edge server where the failure occurred, and when there is only one client that can be assigned to the failure handling edge server, detects the one client as a failure-inducing client.The edge server control unit 250 includes an edge server allocation unit 252, a failure handling edge server subgroup control unit 254, and a failure-inducing client detection unit 256.

[0059] The edge server allocation unit 252 allocates an edge server of an edge server group through an edge server grouping allocation mechanism when a DNS request is received from a client 210 through the DNS control unit 230. When a failure inducing client is confirmed through a failure inducing client detection mechanism, the edge server allocation unit 252 adds the client to a blacklist, and after the failure inducing client is found, reallocates the client allocated to the failure handling edge server to an available edge server in the edge server group, and does not allocate an edge server or a failure handling edge server when there is a DNS request from a client registered in the blacklist.

[0060] When a DNS request for a corresponding client is received after an edge server failure is confirmed by the failure occurrence check unit 240, the failure response edge server subgroup control unit 254 logically generates a failure response edge server subgroup 275 for searching for a failed client in the failure response edge server group 270 through a failure response edge server subgroup control mechanism, assigns a failure response edge server to the failure response edge server subgroup, and then assigns a failure response edge server from the logically generated failure response edge server subgroup 275 through an edge server grouping assignment mechanism.

[0061] The fault-inducing client detection unit 256 analyzes the allocation state of the fault handling edge server and the information of the corresponding client through a fault-inducing client detection mechanism to find the fault-inducing client.

[0062] The client unit 210 may include agents 212 and 214 that transmit a DNS query including user information in order to reconnect to a failure handling edge server selected from the failure handling edge server group when a failure occurs in an edge server or a failure handling edge server. In this case, when the edge server control unit 250 receives a DNS query from the agent 212 or 214 of the client, it extracts user information from the DNS query of the client.

[0063] The client unit 210 is composed of a client that connects to the service server 280 and may include a client in which an agent is installed, and the agent requests information on the service server 280 from the DNS including the client's user information, receives edge server or fault response edge server information, and transmits it to the client's Internet application.

[0064] According to one embodiment of the present invention, the edge servers 260-1, 260-n or the fault response edge servers 270-1, 270-m, 270-nm, 270-n connect the client 100 and the service server 280, and may be embodied by programs, servers, and hardware equipment having a service relay function or a service function, such as a proxy server or cache server, and may be used as a service path to which the client connects and may also act as a service server to which the client connects.

[0065] In addition, the edge server group 260 and the fault response edge server group 270 can constitute a service relay network, and the service relay network can be composed of edge servers and fault response edge servers, which have a service relay function or include the concept of a server or equipment having a service function, as well as network equipment such as a gateway, a router, a switch, and a hub.

[0066] 3 is a flowchart illustrating a method for detecting a fault-inducing client using a fault handling edge server grouping according to an embodiment of the present invention. A method for detecting a fault-inducing client using a fault handling edge server grouping according to an embodiment of the present invention will be described with reference to FIGS. 1 to 3.

[0067] First, in the service relay network 130 in which there are multiple edge servers on the route by which the client 210 approaches the service server 280, the failure occurrence check unit 240 checks for a failure in the edge server (step S310). If the failure occurrence check shows that a failure has occurred in the edge server, the edge server control unit 250 provides a failure response edge server subgroup 275 selected from the failure response edge server group 270, which is made up of a greater number of failure response edge servers than the number of edge servers in which the failure has occurred, to the client connected to the failure response edge server in place of the edge server in which the failure has occurred (step S320).

[0068] Then, the failure occurrence check unit 240 checks for the occurrence of a failure in the failure handling edge server provided to the client (step S330). If the failure occurrence check result indicates that a failure has occurred in the substituted failure handling edge server, the edge server control unit 250 provides the client with another failure handling edge server sub-group in the failure handling edge server group 270 that is not provided to the client, instead of the failure handling edge server where the failure has occurred (step S340). Here, the failure handling edge server sub-group 275 may be composed of a plurality of failure handling edge servers based on the number of failure handling edge servers that have not been assigned from the failure handling edge server group by analyzing the number of clients and connection records assigned to the edge server where the failure has occurred or the failure handling edge server when a failure occurs in an edge server or a failure handling edge server.

[0069] The failure occurrence check unit 240 checks for a failure in the failure response edge server provided to the client (step S350), and whenever a failure occurs in the failure response edge server, the edge server control unit 250 generates a failure response edge server subgroup that is not provided to the client in the failure response edge server group 270 to replace the failed failure response edge server, and if there is only one client assigned to the failed failure response edge server, detects the one client as a failure-inducing client (step S360).

[0070] The check for a fault occurrence of the edge server or the fault-handling edge server can be performed by monitoring the occurrence of a fault of the edge server or the fault-handling edge server when a DNS query is received from a client, or regardless of receiving a DNS query from a client.

[0071] 4 is a flow chart showing an embodiment of detecting a failure-inducing client by checking whether the edge server or the failure handling edge server has failed when a DNS query is received from a client and replacing the failed edge server or the failure handling edge server with a failure handling edge server subgroup. Referring to FIG. 2 and FIG. 4, the DNS control unit 230 receives a DNS query from the DNS server 220 (step S410) and checks whether the received DNS query contains user information (step S415). If the user information is present, the user information is extracted (step S420).

[0072] The edge server allocation unit 252 in the edge server control unit 250 allocates an edge server IP corresponding to the extracted user information (step S430).

[0073] On the other hand, if the user information is not found in the DNS query in step S415, the edge server allocation unit 252 provides the edge server IP allocated to the client that requested the DNS query (step S425).

[0074] When the edge server IP is assigned, the failure occurrence check unit 240 checks whether a failure has occurred in the assigned edge server (step S435).

[0075] If the failure occurrence check unit 240 checks that no failure has occurred (S440), the edge server IP or failure response edge server IP allocated by the edge server allocation unit 252 is provided (step S445).

[0076] In the unlikely event that a failure occurs, a check is made to see whether only one client is assigned to the edge server or failure handling edge server (step S450). If only one client is assigned to the edge server or failure handling edge server, the client is detected as a failure-inducing client (step S460). If multiple clients are assigned to the edge server or failure handling edge server, the edge server or failure handling edge server where the failure occurred is replaced by a failure handling edge server group (step S455).

[0077] 5 is a flowchart showing an embodiment of detecting a failure-inducing client by checking whether the edge server or failure handling edge server has failed regardless of receiving a DNS query from a client, and substituting the failed edge server or failure handling edge server with a failure handling edge server subgroup. After substituting the failed edge server or failure handling edge server with the failure handling edge server subgroup, if a failure occurs in the edge server or failure handling edge server, a client IP or user information using the failed edge server or failure handling edge server is searched, and when a DNS query request is received, a failure handling edge server selected from the failure handling edge server group can be provided to the searched client.

[0078] 2 and 5, the failure occurrence check unit 240 monitors whether a failure occurs in an edge server or a failure handling edge server, regardless of whether a DNS query is received (step S510).

[0079] If the monitoring result shows that a failure occurs in an edge server or a failure response edge server (step S520), the edge server control unit 250 searches for clients that are using the failed edge server or the failure response edge server as a route to access the service server 280 and stores the client information (step S530).

[0080] After that, it is checked whether only one client is assigned to the edge server or failure handling edge server (step S540). If only one client is assigned to the edge server or failure handling edge server, the client is detected as a failure-inducing client (step S550). If multiple clients are assigned to the edge server or failure handling edge server, the edge server or failure handling edge server where the failure occurred is replaced by a failure handling edge server group (step S560).

[0081] Then, when a DNS query is received from the searched client (S570), the DNS control unit 230 provides the assigned edge server IP or failure handling edge server IP to the client that requested the DNS query (step S580).

[0082] Meanwhile, the client includes agents 212 and 214 that transmit a DNS query to connect to a failure handling edge server selected from a failure handling edge server group when a failure occurs in an edge server or a failure handling edge server, and the agents 212 and 214 can request a DNS query including user information. The user information that can be included in the DNS query includes information that can identify a user or information that can identify a device, and can include at least one of a login ID, a device ID (MAC, CPU ID, HDD Serial, etc.), a phone number, and an IP address.

[0083] When the edge server control unit 250 receives a DNS query from the client's agent, it can extract user information from the client's DNS query and provide the client with a fault response edge server IP corresponding to the extracted user information selected from the fault response edge server group.

[0084] Meanwhile, the method and system for detecting a fault-inducing client using a fault-handling edge server grouping according to an embodiment of the present invention will be described in more detail. First, in an embodiment of the present invention, edge server allocation can be performed as follows. FIG. 6 shows an edge server allocation method according to an embodiment of the present invention. Referring to FIG. 6, first, when a client 610 requests information on a service server 650 through a DNS server 620, the DNS server 620 transmits the DNS request of the client 610 to a client route control server 630. An edge server allocation unit (not shown) of the client route control server 620 allocates clients that receive requests to edge servers in the edge server group 640 it manages in order. The first client 611 is allocated to edge server 1 641, the client 2 612 to edge server 2 642, and the client 3 613 to edge server 3 643 in order. If there are more client requests than the number of edge servers in the edge server group 640, the edge servers are reallocated in the edge server group from the beginning. If there are three edge servers in the edge server group, clients 1 to 3 have already been assigned all of edge servers 1 to 3, so client 4 614 is reassigned to edge server 1 641, and client 5 615 is assigned to edge server 2 642. Clients already assigned can continue to use the assigned edge server without any changes. An edge server can accept one or more clients, and the maximum capacity is affected by the system performance of the edge server.

[0085] If an edge server is assigned and a fault occurs in the edge server, the fault-inducing client can be detected as follows. FIG. 7 shows a case where the client 1 611 becomes a fault-causing client due to the influence of malicious code or hacking. Referring to FIG. 7, when the edge server 1 641 to which the clients 1, 4, and 7 are connected is attacked by traffic from the client 1 611, the edge server 1 611 becomes unstable, and the edge server 1 641 transmits communication-related information such as system information, traffic information, and client connection information to a fault occurrence check unit (not shown) of the client route control server 630, and the fault occurrence check unit (not shown) of the client route control server 630 transmits the corresponding information to an edge server control unit (not shown) of the client route control server 630.

[0086] 8 shows that clients 1, 4, and 7 are assigned to failure handling edge server subgroup 1 810 in order to handle clients connected to edge server 1 641 under attack with a failure handling edge server. Referring to FIG. 8, after confirming the number of clients (3) connected to edge server 1 641 where a failure has occurred through a failure handling edge server subgroup control mechanism of an edge server control unit (not shown), failure handling edge server subgroup 1 810 is logically generated within a failure handling edge server group (not shown), and failure handling edge server 1 812 and failure handling edge server 2 814 are included in the corresponding failure handling edge server subgroup 1 810 and then assigned to the failure handling edge server subgroup 1 810.

[0087] When client 1 611, client 4 614, and client 7 617 assigned to edge server 1 641 make a DNS request to be assigned an edge server again, an edge server assignment unit (not shown) of the client route control server 630 assigns them to a failure handling edge server in failure handling edge server subgroup 1 810 in a failure handling edge server group that is not an edge server group. Here, it is assumed that there are two failure handling edge servers 812 and 814 in failure handling edge server subgroup 1 810 (assuming that the failure handling edge server subgroup control mechanism includes two failure handling edge servers).

[0088] Since there are three clients (clients 1, 4, and 7) assigned to the edge server 1 641, the client 1 611 and the client 4 614 are assigned to the failure handling edge server 1 812, and the client 7 617 is assigned to the failure handling edge server 1 812. The system state of the failure handling edge server 1 812 to which the clients 1 and 4 are connected becomes unstable due to a traffic attack from the client 1 611. The failure handling edge server 1 812 transmits communication related information to a failure occurrence check unit (not shown) of the client route control server 630. The failure occurrence check unit (not shown) of the client route control server 630 transmits the corresponding information to an edge server control unit (not shown) of the client route control server 630. Here, the failure induced client detection mechanism in the edge server control unit of the client route control server reserves judgment since it is not yet the time when the failure handling edge servers and clients are assigned in a one-to-one ratio.

[0089] Then, after isolating the clients connected to failure handling edge server 1 812 of failure handling edge server group 1 810 under attack, a new failure handling edge server subgroup 2 910 is logically generated via the failure handling edge server subgroup control mechanism of the edge server control unit (not shown), and failure handling edge server 3 912 and failure handling edge server 4 914 are included in failure handling edge server subgroup 2 910 and then assigned to failure handling edge server subgroup 2 910.

[0090] 9 shows that client 1 611 and client 4 614 are separated and reallocated to failure handling edge server subgroup 2 910. Referring to FIG. 9, when client 1 611 and client 4 614, which are allocated to failure handling edge server 1 812 of failure handling edge server subgroup 1 810, make a DNS request to reallocate an edge server, an edge server allocation unit (not shown) of client route control server 630 allocates them to a failure handling edge server in failure handling edge server subgroup 2 910, which is not failure handling edge server subgroup 1 810. Here, it is assumed that there are two failure handling edge servers 912 and 914 in failure handling edge server subgroup 2 910 (assuming that two failure handling edge servers are included in failure handling edge server subgroup 2 by the failure handling edge server subgroup control mechanism).

[0091] Also, client 7 617, which was connected to failure handling edge server 2 814 of failure handling edge server subgroup 1 810, is not a target for failure induction client, so it is assigned to edge server 4 644, which is a new edge server in the edge server group, in an edge server assignment unit (not shown) of the client route control server 630. It is possible to assign it to an available edge server, or to an edge server that has room in the system among the already assigned edge servers.

[0092] Since there are two clients (clients 1 and 4) assigned to the failure handling edge server 1 812 of the failure handling edge server subgroup 1 810 and there are also two failure handling edge servers belonging to the failure handling edge server subgroup 2 910, the client 1 611 is assigned to the failure handling edge server 3 912 and the client 4 614 is assigned to the failure handling edge server 4 914. The edge server assignment unit (not shown) of the client route control server 630 transmits the corresponding assignment information to the failure induced client detection unit (not shown) of the client route control server 630. The system state of the failure handling edge server 3 912 connected to the client 1 611 becomes unstable due to a traffic attack from the client 1 611. The failure response edge server 3 912 transmits communication-related information to a failure occurrence checking unit (not shown) of the client route control server 630, and the failure occurrence checking unit (not shown) of the client route control server 630 transmits the corresponding information to an edge server control unit (not shown) of the client route control server 630.

[0093] 10 shows a process of recognizing a client that connects to one failure handling edge server and attacks the failure handling edge server as a failure causing client. Referring to FIG. 10, a failure inducing client detection unit (not shown) of the client route control server 630 recognizes the client 1 611 connected to the failure handling edge server 1 812 as a failure causing client because the allocation ratio between the failure handling edge server and the client is one to one and the failure handling edge server 1 812 has caused a failure. The failure inducing client detection unit (not shown) of the client route control server 630 records the client 1 611 in a blacklist and transmits the information to the edge server allocation unit (not shown) of the client route control server 630. And, the edge server allocation unit (not shown) of the client route control server 630 can allocate the client 4 614 to the edge server 4 644 that has a spare edge server in terms of the system among the edge servers in the edge server group because the client 4 614 is not a failure inducing client.

[0094] Even if client 1 611, which is a client causing a problem and registered on the blacklist, again makes a DNS request to request information from the service server 650, the edge server allocation unit (not shown) of the client route control server 630 responds by processing it as null without allocating an edge server, so that the client cannot receive an edge server allocation and cannot connect to the service server 650.

[0095] Meanwhile, the difference between the fault-inducing client detection method using a client route control system, which is a conventional registered patent, and the fault-inducing client detection method using grouping of fault response edge servers according to the present invention will be described in more detail.

[0096] There is not much difference between the conventional patent method and the method of the present invention when a failure occurs in only one client, but when failures occur simultaneously in multiple clients, there are differences in detection method and accuracy, and the present invention can accurately detect the client that induces the failure without false positives.

[0097] 11 and 12 are diagrams for explaining a method for detecting a client that induces a fault using a client route control system, which is a conventional registered patent. Figs. 13 to 16 are diagrams for explaining a method for detecting a client that induces a fault using grouping of fault handling edge servers according to the present invention.

[0098] As described above in the background art of the present invention, referring to Fig. 11, when an edge server group 11 consisting of six edge servers is configured with a 3x2 matrix arrangement for searching for a faulty client, the matrix is ​​configured as follows: When clients A to H are assigned edge servers according to the above matrix, if clients A and F are faulty clients, edge servers 1, 2, 3, 4, and 6 in the edge server group will have faults due to clients A and F.

[0099] However, as can be seen from the matrix in Fig. 12, not only clients A and F but also other clients assigned to edge servers 1, 2, 3, 4, and 6 in the edge server group where the failure occurred are also designated as failure-inducing clients, so that clients B and E are also recognized as clients causing the failure. In other words, clients B and E may be erroneously detected as failure-inducing clients.

[0100] On the other hand, in comparing the fault-induced client detection method using grouping of fault-response edge servers according to the present invention with the conventional method, it is assumed that there are the same clients A to H as in the metrics method, and that there are two edge servers 1 and 2 in edge server group 13 for grouping, and nine fault-response edge servers 11 to 19 in fault-response edge server group 14, as shown in Figure 13.

[0101] Referring to Figure 14, since there are only two edge servers in the edge server group, clients A to H are assigned an edge server and provided with services as shown in Figure 14. If clients A and F are failure-inducing clients, both edge servers 1 and 2 in edge server group 13 will experience failures.

[0102] First, failure handling edge server subgroups 1 and 2 are generated in the failure handling edge server group by a failure handling edge server subgroup control mechanism and an edge server grouping allocation mechanism of the edge server control unit (not shown), and failure handling edge server group 1 includes failure handling edge servers 11 and 12, while failure handling edge server group 2 includes failure handling edge servers 13 and 14. Then, clients A to H assigned to edge servers 1 and 2 where a failure has occurred are assigned to failure handling edge server subgroups 1 and 2 as shown in FIG.

[0103] Referring to FIG. 15, since the failure cause clients, client A and client F, are assigned to the failure handling edge server 11 of the failure handling edge server subgroup 1 13-1 and the failure handling edge server 13 of the failure handling edge server subgroup 2 13-2, the corresponding failure handling edge servers will experience a failure.

[0104] Failure response edge server subgroup 3 13-3 and failure response edge server subgroup 4 13-4 are generated by the failure response edge server subgroup control mechanism and edge server grouping allocation mechanism of the edge server control unit (not shown), and of the failure response edge servers that were not assigned, as shown in Figure 16, failure response edge servers 15 and 16 are included in failure response edge server subgroup 3 13-3, and failure response edge servers 17 and 18 are included in failure response edge server subgroup 4 13-4, and clients A, B, E, and F connected to the failure response edge servers 11 and 13 that have failed are assigned.

[0105] 16, failure handling edge server 15 in failure handling edge server subgroup 3 13-3 and failure handling edge server 18 in failure handling edge server subgroup 4 13-4 have failed, and since there is a one-to-one mapping between clients and failure handling edge servers, client A assigned to failure handling edge server 15 and client F assigned to failure handling edge server 18 are determined to be the clients that caused the failure. As described above, the method of the present invention detects only A and F as failure-inducing clients and does not detect clients B and E, which were erroneously detected by the conventional method, as failure-inducing clients.

[0106] Hereinafter, a method for controlling a client terminal connection by a client terminal connection control device according to an embodiment using the above-mentioned failure-inducing client detection method will be described with reference to Fig. 17. The client terminal connection control device may include a communication unit that performs data communication with at least one of a client terminal, an edge server, and a failure response edge server, and a processor for performing the following processes.

[0107] First, the client terminal connection control device can assign communication processing of a first client group consisting of a plurality of client terminals to an edge server (S1710). Next, when a failure occurs in the operation of the edge server, the client terminal connection control device can assign communication processing of a second client group including at least one client terminal belonging to the first client group to a failure handling edge server (S1720). Here, the second client group can be configured to include only at least one client terminal belonging to the first client group. Alternatively, the second client group can be configured to further include other client terminals in addition to the at least one client terminal belonging to the first client group.

[0108] Next, the client terminal connection control device can determine whether a failure occurs in the operation of the failure handling edge server (S1730). The client terminal connection control device can control the connection of client terminals depending on whether a failure occurs. For example, when a failure occurs in the operation of the failure handling edge server and the number of client terminals assigned to the failure handling edge server is equal to or greater than a predetermined number, the client terminal connection control device can assign communication processing of a sub-client group including at least one client terminal belonging to the client group assigned to the failure handling edge server to the failure handling edge server. For example, the client terminal connection control device can assign communication processing of the sub-client group to a failure handling edge server in which no failure occurs. Here, the predetermined number can be 1 or 2, or any other set value.

[0109] In more detail, the client terminal connection control device can generate a sub-client group including at least one client terminal belonging to the client group assigned to the failure handling edge server, and assign communication processing of the generated sub-client group to the failure handling edge server. As described above, the sub-client group can be configured to include only the at least one client terminal belonging to the client group assigned to the failure handling edge server where the failure occurred. Alternatively, the sub-client group can be configured to further include other client terminals in addition to the at least one client terminal belonging to the client group assigned to the failure handling edge server where the failure occurred.

[0110] Meanwhile, when a failure occurs in the operation of the failure handling edge server and the number of client terminals allocated to the failure handling edge server is a predetermined number, the client terminal connection control device can determine the predetermined number of client terminals as failure-inducing client terminals.

[0111] Hereinafter, a method for controlling a client terminal connection using the above-described fault-inducing client detection method in a client terminal connection control system according to an embodiment will be described with reference to Fig. 18. The client terminal connection control system may include an edge server, a fault handling edge server, and a connection control device.

[0112] First, the edge server can perform communication processing for a first client group consisting of multiple client terminals (S1810). Next, when a failure occurs in the operation of the edge server, the failure handling edge server can perform communication processing for a second client group including at least one client terminal belonging to the first client group (S1820).

[0113] Next, the connection control device can determine whether a failure occurs in the operation of the failure handling edge server (S1830). Thereafter, the connection control device can control the connection of client terminals depending on whether a failure occurs. For example, when a failure occurs in the operation of the failure handling edge server and a predetermined number of client terminals are assigned to the failure handling edge server, the connection control device can assign communication processing of a sub-client group including at least one client terminal belonging to a client group assigned to the failure handling edge server to the failure handling edge server. For example, the connection control device can assign communication processing of the sub-client group to a failure handling edge server in which no failure occurs. Here, the predetermined number can be 1 or 2, or any other set value.

[0114] On the other hand, when a failure occurs in the operation of the failure handling edge server and the number of client terminals allocated to the failure handling edge server is a predetermined number, the connection control device can determine the predetermined number of client terminals as failure induction client terminals.

[0115] The method, device and system according to the above-mentioned embodiment can be embodied as a program code readable by a computer (including all devices having information processing capabilities) on a computer-readable recording medium. The computer-readable recording medium includes all kinds of recording devices in which data that can be read by a computer system is stored. Examples of computer-readable recording devices include ROM, RAM, CD-ROM, magnetic tape, floppy disk, optical data storage device, etc.

[0116] The present invention has been described with reference to the embodiments shown in the drawings, but these are merely examples, and a person skilled in the art would understand that various modifications and equivalent embodiments are possible. Therefore, the true technical scope of the present invention should be determined by the technical ideas of the appended claims. [Industrial Applicability]

[0117] The present disclosure can be used to detect clients that induce network impairments. [Explanation of symbols]

[0118] 100 Client 102, 104 Client 14 Agent 110 DNS Server 120 Client Routing Server 130 Service Relay Network 132 Edge Server Group 134 Edge Server Group for Disaster Response 140 Service Server 20 Client Routing Server 210 Client 210-1 Client 1 212 Agent 201-n Client n 124 Agent 220 DNS Server 230 DNS Control Section 240 Error Occurrence Check Section 250 Edge Server Control Unit 252 Edge Server Allocation Unit 254 Fault-Response Edge Server Subgroup Control Unit 256 Failure Inducing Client Detection Unit 260 Edge Server Group 260-1 Edge Server 1 260-n Edge server n 270 Edge Server Group for Disaster Response 275 Edge Server Subgroup for Disaster Recovery 270-1 Edge server for fault handling 1 270-m Fault response edge server 270-nm Fault-Response Edge Server nm 270-n Fault response edge server n 610 Client 611 Client 1 612 Client 2 613 Client 3 614 Client 4 615 Client 5 620 DNS Server 630 Client Routing Server 640 Edge Server Group 641 Edge Server 1 642 Edge Server 2 643 Edge Server 3 644 Edge Server 4 650 Service Server 810 Disaster Response Edge Server Subgroup 1 812 Edge Server 1 for fault handling 814 Disaster response edge server 2 910 Disaster Response Edge Server Subgroup 2 912 Edge Server 3 for Disaster Recovery 914 Disaster Response Edge Server 4 11 Edge Server Group 12 Edge Server Allocation Matrix 13 Edge Server Group 14. Edge Server Group for Disaster Response 13-1 Disaster Response Edge Server Subgroup 1 13-2 Disaster Response Edge Server Subgroup 2 13-3 Disaster Response Edge Server Subgroup 3 13-4 Disaster Response Edge Server Subgroup 4

Claims

1. A client terminal access control method executed by a client terminal access control device, comprising: allocating communication processing of a first client group, the first client group including a plurality of client terminals, to a server; When a failure occurs in the operation of the server, allocating a communication process of a second client group including at least one client terminal belonging to the first client group to a first failure handling server; determining whether operation of the first failure handling server is impaired; When a failure occurs in the operation of the first failure handling server, determining whether to proceed with a first process or a second process based on the number of client terminals assigned to the first failure handling server; Equipped with The first process is to determine the client terminal assigned to the first failure support server as a failed client terminal, the second process is a process of allocating, to a second failure support server, a communication process of a sub-client group including at least one client terminal allocated to the first failure support server; A method according to claim 1, wherein the number of terminals assigned to the second failure handling server is smaller than the number of terminals assigned to the first failure handling server.

2. The method according to claim 1 , wherein it is determined to proceed with the second process based on the number of the client terminals assigned to the first failure handling server being greater than a predetermined number.

3. The method according to claim 1 , wherein it is determined to proceed with the first process based on the number of the client terminals assigned to the first failure handling server being equal to or less than a predetermined number.

4. The method of claim 2 , wherein the predetermined number is one.

5. A client terminal access control device, a communication unit configured to perform data communication with the server, the first failure handling server, and the second failure handling server; allocating communication processing of a first client group, the first client group including a plurality of client terminals, to a server; When a failure occurs in the operation of the server, allocating a communication process of a second client group including at least one client terminal belonging to the first client group to a first failure handling server; determining whether the operation of the first failure handling server is impaired; When a failure occurs in the operation of the first failure handling server, determining whether to proceed with a first process or a second process based on the number of client terminals assigned to the first failure handling server; A processor configured to: Equipped with The first process is to determine the client terminal assigned to the first failure support server as a failed client terminal, the second process is a process of allocating, to the second failure support server, a communication process of a sub-client group including at least one client terminal allocated to the first failure support server; The number of terminals to be assigned to the second failure support server is smaller than the number of terminals to be assigned to the first failure support server.

6. A client terminal access control method executed by a client terminal access control system having a server, a first failure handling server, a second failure handling server, and an access control device, comprising: executing, by the server, a communication process for a first client group including a plurality of client terminals; When a failure occurs in the operation of the server, executing a communication process of a second client group including at least one client terminal belonging to the first client group by the first failure handling server; determining by the access control device whether an operation of the first failure handling server is faulty; When a failure occurs in the operation of the first failure handling server, determining, by the access control device, whether to proceed with a first process or a second process based on the number of client terminals assigned to the first failure handling server; Equipped with The first process is to determine the client terminal assigned to the first failure support server as a failed client terminal, the second process is a process of allocating, to the second failure support server, a communication process of a sub-client group including at least one client terminal allocated to the first failure support server; A method according to claim 1, wherein the number of terminals assigned to the second failure handling server is smaller than the number of terminals assigned to the first failure handling server.

7. A client terminal access control system, comprising: A server; a first failure response server; a second failure response server; An access control device; Equipped with The server, Executing a communication process for a first client group having a plurality of client terminals; When a failure occurs in the operation of the server, the first failure handling server executes a communication process for a second client group including at least one client terminal belonging to the first client group; The access control device determines whether the operation of the first failure handling server is faulty; When a failure occurs in the operation of the first failure handling server, the access control device determines whether to proceed with the first process or the second process based on the number of client terminals assigned to the first failure handling server; The first process is to determine the client terminal assigned to the first failure support server as a failed client terminal, the second process is a process of allocating, to the second failure support server, a communication process of a sub-client group including at least one client terminal allocated to the first failure support server; A system in which the number of terminals assigned to the second failure support server is smaller than the number of terminals assigned to the first failure support server.

8. A computer program stored on a computer readable recording medium for carrying out the method according to any one of claims 1 to 4 and 6.

9. A computer-readable recording medium storing a computer program for causing a computer to execute the method according to any one of claims 1 to 4 and 6.

Images