Single-point-of-failure search device, single-point-of-failure search method, and program

Abstract

A single-point-of-failure search device as in one embodiment of this invention searches for a single point of failure in a network. The single-point-of-failure search device is equipped with a resource management unit, an optimal resource selection unit, and a single point of failure (SPOF) resource calculation unit. The resource management unit manages, for each of the layers of the network, construction properties including at least resource coordination overhead, a design period, a construction period, a lead time, a life cycle cost, and a cost value. In accordance with a search object given by designating a start point and an end point, the optimal resource selection unit optimizes a route between the start point and the end point on the basis of at least the cost value, and outputs an optimal route in the network. The SPOF resource calculation unit gives, to the optimal resource selection unit, a search object with which the cost value of paths constituting the optimal route has been set to a maximum value to search for a bypass route for the optimal route, and calculates a non-bypassable resource as being a single point of failure.

Description

Single point of failure detection device, single point of failure detection method, and program

[0001] One aspect of the present invention relates to a single point of failure detection device, a single point of failure detection method, and a program.

[0002] Services that utilize communication networks are expanding daily. In modern society, which depends on networks, it is important not to cause failures or disruptions. It is equally important to minimize the impact of any disruptions that do occur and to restore the system quickly. Redundancy is an effective technique for improving the robustness of networks.

[0003] A single point of failure (SPOF: Single Point Of Failure) is an important consideration when making a network redundant. An SPOF can be said to be the Achilles' heel of a network, and if a failure occurs here, the entire system may stop. Therefore, when designing a network, it is necessary to accurately identify SPOFs and take countermeasures. Designing a network requires a consistent determination based on a combination of various types of information.

[0004] Japanese Patent No. 6655524, Japanese Patent No. 7107158, Japanese Patent No. 6837022

[0005] Networks are hierarchical from the physical layer to the top logical layer, making it difficult to maintain consistency in design policies. Also, in existing technologies, the person in charge of design is different for each layer, making it difficult to grasp the entire network. Especially in recent highly complex network environments, it is very difficult to determine SPOFs.

[0006] Whether combining information between layers for unified management or determining SPOFs, it is still difficult to determine across the entire network because it is necessary to design all layers in a chain. In the conventional method of sequentially assuming and searching for resources where endpoints are not connected, a huge amount of calculation time is required as the number of related resources increases.

[0007] Therefore, the object of the present invention is to provide a technology that enables the speedup of single-point fault detection.

[0008] According to the embodiment, the single point of failure (SPOF) search device searches for a single point of failure in a network. The SPOF search device comprises a resource management unit, an optimal resource selection unit, and a SPOF resource calculation unit. The resource management unit manages construction characteristics, including at least resource adjustment overhead, design period, construction period, lead time, lifecycle cost, and cost value, for each network layer. The optimal resource selection unit optimizes the path between the specified starting point and ending point based on at least the cost value, according to the search target provided by specifying the starting point and ending point, and outputs the optimal path in the network. The SPOF resource calculation unit provides the optimal resource selection unit with a search target whose cost value for the path constituting the optimal path is set to a maximum value, searches for alternative paths to the optimal path, and finally calculates resources that cannot be bypassed as single points of failure.

[0009] According to one aspect of this invention, a single fault point can be searched for at high speed.

[0010] Figure 1 is a functional block diagram showing an example of a single point of failure (SPOF) device according to an embodiment. Figure 2 is a diagram showing an example of the data flow in the single point of failure (SPOF) device shown in Figure 1. Figure 3 is a diagram illustrating the procedure for searching for a SPOF using a railway route map as an example. Figure 4 is a diagram illustrating the procedure for searching for a SPOF using a railway route map as an example. Figure 5 is a diagram illustrating the procedure for searching for a SPOF using a railway route map as an example. Figure 6 is a diagram illustrating the procedure for searching for a SPOF using a railway route map as an example.

[0011] The embodiments will be described below with reference to the drawings. First, an overview will be provided. <Overview> In order to improve robustness at the service level, it is desirable to be able to determine and evaluate SPOF during network design. However, real-world networks consist of numerous devices, and the reliability and cost of these devices and resources vary. Even in such situations, there is a need for technology that can determine and evaluate SPOF across the entire network.

[0012] In SPOF (Single Point of Failure) searches, when a Start (starting point) and Goal (ending point) are specified, it is necessary to determine which resources cannot be bypassed in the event of a failure. Existing technologies assume that the resources related to the Start and Goal fail in stages, create a network configuration in which the connections from Start to Goal are sequentially disconnected, and then sequentially search for conditions under which the connection from Start to Goal is impossible. This is because it is necessary to search all possible paths in a state where the path taken is uncertain due to the impact of failures, meaning that the failure state of all related resources had to be calculated step by step. Consequently, the computation time became enormous as the number of related resources increased.

[0013] Next, we disclose a technology that can solve the above-mentioned difficulties. In this embodiment, we utilize general-purpose network configuration management (NOIM) information to search for the optimal path from information on the connection relationships between resources. Furthermore, by applying this technology, we extract resource information that cannot be bypassed as SPOFs through detour route searching.

[0014] <Configuration> Figure 1 is a functional block diagram showing an example of a single point of failure (SFO) device according to the embodiment. The SFO 1 designs a network as an object that can be abstracted into a management structure consisting of multiple layers. The network can be abstracted into multiple layers, including a physical layer and a logical layer, and each layer contains multiple resources.

[0015] In this embodiment, the construction characteristics for each layer are managed as linking information between layers. Here, the construction characteristics include at least the resource adjustment overhead, design period, construction period, lead time, lifecycle cost, and cost value for each layer of the hierarchical network.

[0016] Even with newly constructed networks, the fact that they ultimately converge on the physical layer (physical equipment, geography, etc.) allows for the integration of layers and the automation of network design. Resources can be broken down into a hierarchical management structure and abstracted into a general-purpose data model. This approach is applicable to any industry where relationships between layers can be considered, integrated, and evaluated; it is not limited to networks. For example, resources other than communication networks, such as transportation networks and water pipelines, can also be managed using this general-purpose data model.

[0017] The single-point-of-fault (SFL) device 1 shown in Figure 1 is a computer comprising a memory for storing a program and a processor such as a CPU (Central Processing Unit) or MPU (Micro Processing Unit). The SFL in this embodiment is realized by the processor executing the program. That is, the program includes instructions that cause the computer to function as the SFL in this embodiment.

[0018] In Figure 1, the single point of failure (SPOF) detection device 1 includes a display unit 10 that provides an operating environment via a GUI (Graphical User Interface), a SPOF resource calculation unit 20, an optimal resource selection unit 30, and a resource management unit 40.

[0019] The resource management unit 40 includes a data format conversion unit 41 that acquires network design information 70 and converts the data format, and a network configuration holding unit 42. The network configuration holding unit 42 centrally manages the construction characteristics of each network layer and resource information using a database.

[0020] The optimal resource selection unit 30 comprises an optimal resource selection processing unit 31 and a calculation result storage unit 32. The optimal resource selection processing unit 31 outputs the optimal route in the modeled network by optimizing the route between the starting point and the ending point based on the cost value of the construction characteristics, according to the search target given by specifying the starting point and the ending point. The route may be a route involving multiple layers or a route involving only a single layer.

[0021] The optimal resource selection processing unit 31 determines and selects the optimal resource configuration from resources that have been combined with construction characteristics as coefficients, based on a search algorithm such as Dijkstra's algorithm. Information representing the designed optimal path is stored in the calculation result storage unit 32.

[0022] The SPOF resource calculation unit 20 provides the optimal resource selection unit 30 with a search target for the optimal route calculated by the optimal resource selection unit 30, setting the cost value of the paths constituting the optimal route to its maximum value, and searches for alternative routes to the optimal route. Finally, it calculates resources that cannot be bypassed as single failure points.

[0023] Here, the SPOF resource calculation unit 20 includes a data receiving unit 21, an element extraction / model conversion unit 22, an SPOF resource extraction unit 23, a detour route database (DB) 24, a calculation result storage unit 25, and a data output unit 26.

[0024] The data receiving unit 21 receives input of a SPOF search target 50 with specified start and goal positions. The element extraction and model conversion unit 22 extracts the requirements and elements of the search target from the given SPOF search target 50 and converts them into a general-purpose data model.

[0025] The SPOF resource extraction unit 23 provides the optimal resource selection processing unit 31 of the optimal resource selection unit 30 with a reset condition that sets the cost value of the path in the data model to a maximum value, and extracts resources that cannot be bypassed. The bypass routes extracted during the calculation process are stored in the bypass route database 24. The resources that cannot be bypassed extracted by the SPOF resource extraction unit 23 are stored in the calculation result storage unit 25 and passed to the display unit 10 and the data output unit 26 to be output as the SPOF resource list 60.

[0026] Figure 2 shows an example of the data flow in the single point of failure (SPOF) search device shown in Figure 1. In Figure 2, assume that a SPOF search target is given, including the Start position and Goal position (e.g., [('Nishi-Oshima', 'Oshima')]). The optimal route is calculated by the optimal resource selection unit 30 through optimal route design, while referring to network configuration information. In this process, alternative routes are also searched for and sequentially accumulated in the alternative route database.

[0027] The SPOF resource calculation unit 20 edits the configuration data for detour route search. Since the target with the optimal cost value is selected during optimal design, here the data is edited to a cost value that would not be selected even when conditions for maximum detour are taken. In other words, it extracts routes other than the optimal resource and changes the cost values ​​of all resources extracted as the optimal resource. For example, it sets the cost of all paths on the optimal route to the maximum value and finds the optimal route under those conditions.

[0028] The SPOF resource calculation unit 20 searches for alternative routes using network configuration information with edited cost values ​​(maximizing the cost values ​​only for resources extracted as optimal resources), and performs optimal design on the alternative route search database with edited cost values. The optimal route that is ultimately selected under these conditions is extracted as an SPOF resource. The SPOF resource calculation unit 20 then extracts the overlapping areas between the optimal resources and alternative resources, calculates an SPOF resource list, and outputs it.

[0029] Here, network configuration information stores the relationships between resources and their cost values. For example, network configuration information in a station route map can be stored in data in the format (resource, related resource, cost value). This configuration information is, for example, [['Nishimagome', 'Magome', 1188.1511925761429], ['Magome', 'Nakanobu', 2217.644094697656], ['Nakanobu', 'Togoshi', 2068.08476888505], ['Togoshi', 'Gotanda', 2552.917193830189], ['Gotanda','Takanawadai', 2269.5258145339167], ['Takanawadai', 'Sengakuji', 1926.5248229166373], ['Sengakuji','Mita', 2480.4129925407033], ['Mita', It can be represented as JSON format data such as 'Daimon', 2407.7948390839483], ['Daimon', 'Shinbashi',2161.5972973008174],・・・・・, ['Higashi-Nihonbashi', 'Bakuro-Yokoyama', 0]].

[0030] Furthermore, the obtained SPOF resource list is output as JSON data, for example, [('Nishi-Ojima', 'Sumiyoshi'), ('Higashi-Nakano', 'Nakai'), ('Nishi-Shinjuku 5-chome', 'Nakano-Sakaue'), ('Funabori', 'Higashi-Ojima'), ('Tochomae', 'Nishi-Shinjuku 5-chome'), ('Sumiyoshi', 'Kikukawa'), ('Nakano-Sakaue', 'Higashi-Nakano'), ('Kikukawa', 'Morishita'), ('Ojima', 'Nishi-Ojima'), ('Shinjuku', 'Tochomae'), ('Higashi-Ojima', 'Ojima')].

[0031] <Operation> Referring to Figures 3 to 6, the procedure for searching for SPOFs using a railway route map as an example will be explained. In Figure 3, all resources (paths) between stations have predetermined cost values ​​set in advance. In this route map, we consider searching for SPOFs on a route starting at Funabori Station and ending at Nakai Station (Start: Funabori, Goal: Nakai).

[0032] As shown by the hatched arrows in Figure 4, let's assume that one optimal path has been found. Then, the SPOF resource calculation unit 20 extracts all the resources (optimal resources) that this optimal path passes through. Then, as shown in Figure 5, the SPOF resource calculation unit 20 resets the cost values ​​of all these optimal resources to their maximum values ​​(for example, 99999).

[0033] In general terms, the SPOF resource calculation unit 20 changes the cost value of the resources extracted at the time of optimal resource search to a cost value that bypasses the target resource. This cost value is greater than the sum of the costs of all paths. After changing the cost value, the shortest path is searched again. In this way, the optimal path is searched with all cost values ​​at their maximum, and the resource selected that is the same as the resource of the shortest path can be said to be the SPOF.

[0034] Figure 6 shows that in addition to the optimal route (shaded hatching) and the detour route (upward diagonal hatching), SPOF (downward diagonal hatching) was also searched. In Figure 6, if there is a possible detour on SPOF, other routes can be calculated as part of the search results.

[0035] In this embodiment, the search target is narrowed down by focusing on the SPOF conditions. Specifically, the technique of designing optimal resources for the network using NOIM is applied, and by applying the following reinterpretation conditions (A, B) to this technique to search for detour routes, the search target is narrowed down and processing speed is increased. At that time, the cost of all paths on the optimal route is set to the maximum value to find the optimal route, and in the process, any path that is selected as the optimal route even if its cost value is the maximum value is output as an SPOF.

[0036] [Reinterpretation Condition A] "A single point of failure (SPOF) is an object that cannot be bypassed when a failure occurs." This can be reinterpreted as "an SPOF is a resource that must be selected under any circumstances." Furthermore, "a resource that is selected under any circumstances" can be reinterpreted as "a resource that is selected even when its cost value is extremely low." Therefore, an SPOF is synonymous with "a resource that is selected even when its cost value exceeds the sum of the cost values ​​of all resources under normal circumstances (i.e., a maximum cost value)."

[0037] [Reinterpretation Condition B] By utilizing the fact that "the optimal resource is selected under normal circumstances," SPOF becomes synonymous with "the resource that must be selected even if the cost values ​​of all resources on the path selected as the optimal resource reach a maximum value." Therefore, it is possible to perform calculations that focus only on the resources extracted as the optimal resources. In other words, the target resources can be narrowed down to just one path.

[0038] In this embodiment, resources are extracted by focusing on the extraction of SPOFs. If only SPOFs are being searched, there is no need to search all detour routes. Therefore, the process is sped up by reducing the number of times the detour route database is recreated by changing the cost value in stages, and by reducing the number of comparisons. In other words, the system compares whether the resources extracted with the optimal resource under normal cost conditions match the resources extracted when the cost value of "only the resources extracted with the optimal resource under normal conditions" is changed to the maximum value. As a result, there is no need to process in stages, and processing speed is increased. That is, the cost values ​​of resources in a single route can be changed all at once instead of changing each one in stages, so the results can be obtained in a single search.

[0039] <Effects> As explained above, in this embodiment, the optimal path from Start to Goal is determined, and the optimal path from Start to Goal is calculated again under the condition that the cost of each path on the obtained optimal path is reset to its maximum value. Then, the paths common to the previously determined optimal path and the later determined optimal path are calculated as SPOF.

[0040] In other words, when determining SPOFs from the network configuration, the system manages the relationships between the configurations to be investigated (the relationships between resources whose impact we want to understand), searches for the optimal path from Start to Goal using this configuration information, and then dynamically changes the resource cost values ​​to extract points where bypassing is impossible. Finally, a list of targets that cannot be bypassed is extracted as SPOFs.

[0041] This process makes it possible to identify a single point of failure (SPOF) in a short amount of time. Consequently, it becomes possible to quickly determine whether or not to address the network's redundant configuration. Therefore, according to this embodiment, it is possible to speed up the search for a single point of failure.

[0042] It should be noted that the present invention is not limited to the embodiments described above, and can be modified in various ways without departing from its essence during implementation. Furthermore, each embodiment may be combined as appropriate, and in that case, the combined effects can be obtained. Moreover, the above embodiments include various inventions, and various inventions can be extracted by selecting combinations from the multiple constituent elements disclosed. For example, if the problem can be solved and effects obtained even if some constituent elements are deleted from all the constituent elements shown in the embodiment, then the configuration with these deleted constituent elements can be extracted as an invention.

[0043] 1...Single Point of Failure (SPOF) Search Device 10...Display Unit 20...SPOF Resource Calculation Unit 21...Data Reception Unit 22...Element Extraction / Model Conversion Unit 23...SPOF Resource Extraction Unit 24...Detour Route Database 25...Calculation Result Storage Unit 26...Data Output Unit 30...Optimal Resource Selection Unit 31...Optimal Resource Selection Processing Unit 32...Calculation Result Storage Unit 40...Resource Management Unit 41...Data Format Conversion Unit 42...Network Configuration Holding Unit 50...SPOF Search Target 60...SPOF Resource List 70...Network Design Information.

Claims

1. A single point of failure (SPOF) search device for searching for a single point of failure in a network, comprising: a resource management unit that manages construction characteristics for each layer of the network, including at least resource adjustment overhead, design period, construction period, lead time, life cycle cost, and cost value; an optimal resource selection unit that outputs an optimal route in the network by optimizing the path between the starting point and the ending point based at least on the cost value, according to a search target given by specifying the starting point and the ending point; and a SPOF resource calculation unit that searches for a bypass route of the optimal route by providing the optimal resource selection unit with a search target for which the cost value of the paths constituting the optimal route is set to a maximum value, and calculates resources that cannot be bypassed as a single point of failure.

2. The single point of failure (SPOF) search device according to claim 1, wherein the SPOF resource calculation unit comprises an element extraction / model conversion unit that extracts the requirements and elements of the given search target and converts them into a general-purpose data model, and an SPOF resource extraction unit that provides the optimal resource selection unit with a reset condition that sets the cost value of the path in the data model to a maximum value and extracts the unavoidable resources.

3. A single point of failure (SFO) method for searching for a single point of failure in a network using a computer, comprising: a process in which the computer manages construction characteristics for each layer of the network, including at least resource adjustment overhead, design period, construction period, lead time, life cycle cost, and cost value; a process in which the computer optimizes the path between the starting point and the ending point based at least on the cost value according to a given search target with a specified starting point and ending point, and outputs an optimal path in the network; and a process in which the computer sets the cost value of the paths constituting the optimal path to a maximum value, searches for alternative paths to the optimal path, and calculates resources that cannot be bypassed as single points of failure.

4. A program comprising instructions that cause a computer to function as a single-point-of-fault device according to claim 1 or claim 2.

Images