JP2007122639A - Reliability evaluation system, reliability evaluation method and reliability evaluation program of information system - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To make a system operating ratio favorable and to reduce a burden of a designer who designs an information system. <P>SOLUTION: This reliability evaluation system of the information system has: an input part 11 in which failure information, system configuration information and business processing information to a failure mode of components indicating each of software and hardware are inputted; an analysis part 31 which analysis the failure information to obtain failure analysis information; a creation part 32 which creates a diagram based on the system configuration information; an analysis part 34 which analyzes the business processing information to obtain business processing analysis information; a calculation part 35 which calculates the system operating ratio based on the failure analysis information and the business processing analysis information; a determination part 36 which determines whether or not the system operating ratio reaches a reference value; an extraction part 37 which extracts a basic event in relation to a rise of the system operating ratio when it is determined that the system operating ratio does not reach the reference value and reconfiguration parts 39, 40 which reconfigure new information based on whether or not the deterioration of the non-operating ratio of the extracted basic event is likely to occur. <P>COPYRIGHT: (C)2007,JPO&INPIT

Description

  The present invention relates to an information system reliability evaluation system, a reliability evaluation method, and a reliability evaluation program.

  The reliability evaluation of the information system is performed at the operation stage of the information system constituted by a computer network such as an online transaction system or the design stage of the information system. In this reliability evaluation, the operation rate of the system is evaluated. In the evaluation of the operating rate of this system, fault tree analysis is often used (see, for example, Non-Patent Document 1, Patent Document 1, and Patent Document 2).

  For example, in the evaluation of the operation rate of a system such as a nuclear power plant, the following is performed for the purpose of ensuring the safety of the plant. The evaluator who evaluates the system assumes a trouble event and calculates the probability of reaching this. Then, the evaluator quantitatively analyzes that the occurrence probability of the corresponding trouble hardly occurs. The main analysis method used at that time is a method called fault tree analysis. This fault tree analysis is used in reliability engineering and related fields, and is also used in information systems (for example, see Non-Patent Document 1).

  The following methods are used to evaluate the availability with fault tree analysis. The evaluator first selects the top event assumed in the system. The evaluator then explores the primary factors that lead to the top event and derives a logical relationship (AND, OR) between the top event and the primary factor. The evaluator expresses the logical relationship in a tree structure.

  In this tree structure, a top event is described, a logic symbol is described below the top event, and a primary factor is described below the logic symbol. Similarly, a logical symbol is described under each primary factor, a secondary factor is described below the logical symbol, and the same description is repeated for the tertiary factor and the quaternary factor. As described above, the tree structure is subdivided to a level at which the probability of occurrence of trouble can be estimated from experiments and the like.

  The top event is shown to be generated by a combination of events (basic events) at the lowest layer of the tree structure by using Boolean algebra. This combination (that is, a tree structure indicating a logical relationship) makes it possible to derive a non-operation rate (= 1-operation rate) corresponding to the top event, so that an operation rate corresponding to the top event is derived. Is possible.

  For example, in the reliability evaluation of a normal information system, a fault tree is generated from the equipment configuration of the factory system, and the degree of reliability of the factory system (system There is a reliability analysis that calculates the operating rate.

According to this reliability analysis, a fault tree is generated when elements that can be changed are changed as parameters based on production volume, equipment configuration, etc., and the degree of reliability of the factory system (system availability) ) Is calculated. Then, the calculated degrees of reliability are compared, and the equipment configuration with the highest degree of reliability (high system operation rate) is extracted, and the factory system is operated based on the extracted equipment configuration. Or a repair plan for the factory system is selected (see, for example, Patent Document 1 and Patent Document 2). As a result, the factory system can be designed so as to sufficiently satisfy the standard value of the system operation rate, and repairs can be performed while operating the factory system with a high system operation rate.
Japanese Patent Laid-Open No. 9-234652 Japanese Patent Laid-Open No. 9-237102 Kenji Kitagawa “Latest Design Examination Technology” Techno System December 4, 1987 (2nd edition)

  However, the above-described normal reliability evaluation has the following problems. In other words, in the information system, the number of business processes is not constant on average over time, the number of business processes fluctuates over time, and the business concentration time period when the number of business processes concentrates and the business quiet time period when the number of business processes is quiet Exists. For example, in the stock online trading system, as shown in FIG. 19, the business concentration time zone (around 9: 00 to 9:30, around 12:30 to 12:40, around 14:50 to 15:00) There is a large difference in the number of business processes in the off-hours (before 9:00, around 11:00 to 12:30, after 15:00).

  In such a case, a load is applied to the system in the business concentration time period, so the failure occurrence probability is high. On the contrary, in the idle business time period, the system is not loaded, so the failure occurrence probability is low. For this reason, since multiple components (hardware or software related to the operation of hardware) are likely to fail multiple times during the work concentration time zone, even if the system is a redundant system, Failure of the entire system is likely to occur.

  In addition, if the repair time of a component becomes longer than the interval of the work concentration time zone and crosses over the work concentration time zone, the tendency of multiple components to easily fail one after another becomes stronger. Failures are more likely to occur. For example, if the repair time of a component is increased from zero, if the repair time of the component exceeds the interval of the work concentration time period, the failure of the entire system is likely to occur suddenly and the system operation rate decreases. To do.

  In order to be able to evaluate the rapid change in system availability when the repair time of such a component changes in the vicinity of the interval of the work concentration time zone, It is necessary to evaluate the reliability of the information system in consideration of the fluctuation of the failure occurrence probability. However, in the reliability evaluation of a normal information system, the number of business processes is considered to be constant over time, and the system availability is evaluated. It is not possible to evaluate the rapid change in system availability when changing in the vicinity.

  For this reason, for example, even if an information system is designed so as to have the highest system operation rate, the system operation rate when the information system is actually operated may fall below a reference value. Even if the information system is changed to increase the system operating rate based on the above-described reliability evaluation at the operation stage of the information system, the system operating rate may fall below the reference value. As a result, there is a problem in that the design of the information system needs to be redone, and the burden on the information system designer becomes very large.

  The present invention has been made in view of the above, and an object thereof is to provide a reliability evaluation system for an information system that makes it possible to reduce the burden on a designer who designs an information system while improving the system operating rate. To provide a reliability evaluation method and a reliability evaluation program.

  A first feature according to an embodiment of the present invention is that, in an information system reliability evaluation system, software and a device for executing a function of a device are used as information necessary for the reliability evaluation of the information system including the device. Configuration of information system using failure information related to failure corresponding to failure mode of component that is configured by each component, or component that is configured by each component, and device and component or component group An information input unit for receiving information including at least system configuration information indicating information on the system, a reference value of the system operating rate indicating the operating rate of the entire information system, and business processing information regarding business processing of the information system, and a failure mode Failure information analysis unit that analyzes failure information every time and acquires failure analysis information as an analysis result, and based on system configuration information, Analyzing and analyzing business process information and a diagram creation unit that creates a diagram showing a hierarchical logical relationship from the top event indicating the failure mode of the information system to the basic event indicating the failure mode of the component or group of components As a result, the business process information analysis unit that acquires the business process analysis information, calculates the operating rate of the basic event based on the failure analysis information and the business process analysis information, and based on the failure analysis information, the business process analysis information, and the diagram, An operating rate calculation unit that calculates the operating rate of the top event as a system operating rate, an operating rate determination unit that determines whether the calculated system operating rate has reached a reference value, and the calculated system operating rate as a reference If it is determined that the value has not been reached, the system performance is determined based on the relationship between the non-operation rate corresponding to the basic event and the system non-operation rate corresponding to the top event. The basic event extraction unit that extracts basic events related to the rate increase, the changeability determination unit that determines whether the downtime corresponding to the extracted basic event can be reduced, and the extracted If it is determined that the downtime corresponding to the basic event can be reduced, new failure analysis information corresponding to the extracted basic event is reset in the uptime calculation unit, and the extracted basic event is handled. And a first resetting unit that resets new failure information and new system configuration information in the information input unit when it is determined that the unavailability cannot be reduced.

  The second feature according to the embodiment of the present invention is that, in the reliability evaluation method for an information system, software and a device for executing a function of the device as information necessary for the reliability evaluation of the information system including the device are provided. Configuration of information system using failure information related to failure corresponding to failure mode of component that is configured by each component, or component that is configured by each component, and device and component or component group Enter the information that includes at least the system configuration information that indicates information about the system, the reference value of the system operating rate that indicates the operating rate of the entire information system, and the business processing information related to the business processing of the information system. Analyzing and analyzing business process information to obtain business process analysis information as an analysis result, analyzing failure information for each failure mode, Perform failure information analysis to obtain failure analysis information as a result of analysis, and based on the system configuration information, hierarchically change from the top event indicating the failure mode of the information system to the basic event indicating the failure mode of the component or component group Create a diagram that shows a logical relationship, calculate the availability of basic events based on failure analysis information and business process analysis information, and calculate the top event based on failure analysis information, business process analysis information, and diagrams The operating rate is calculated as the system operating rate, the operating rate is determined to determine whether the calculated system operating rate has reached the reference value, and the calculated system operating rate is set to the reference value. If it is determined that the system has not reached, the system unavailability corresponding to the basic event and the system unavailability corresponding to the top event Perform basic event extraction to extract basic events related to the increase in system availability, perform change possibility determination to determine whether the downtime corresponding to the extracted basic events can be reduced, and extract If it is determined that it is possible to reduce the unavailability corresponding to the extracted basic event, new failure analysis information corresponding to the extracted basic event is reset in the availability calculation unit, and the extracted basic event If it is determined that the unavailability corresponding to the above cannot be reduced, the first resetting is performed to reset new failure information and new system configuration information in the information input unit, and the extracted basic event is handled. When new failure analysis information is reset, processing after the operation rate calculation is performed, and when new failure information and new system configuration information are reset, processing after failure information analysis is performed.

  According to a third feature of the embodiment of the present invention, in the reliability evaluation program for an information system, software and a device for executing the function of the device are used as information necessary for the reliability evaluation of the information system including the device. Configuration of information system using failure information related to failure corresponding to failure mode of component that is configured by each component, or component that is configured by each component, and device and component or component group Enter the information that includes at least the system configuration information that indicates information about the system, the reference value of the system operating rate that indicates the operating rate of the entire information system, and the business processing information related to the business processing of the information system. Business process information analysis to obtain business process analysis information as an analysis result and analyze failure information for each failure mode. Failure information analysis to obtain failure analysis information as an analysis result, and from the top event indicating the failure mode of the information system to the basic event indicating the failure mode of the component or component group based on the system configuration information Create a diagram that shows a hierarchical logical relationship, calculate the availability of basic events based on failure analysis information and business process analysis information, and based on failure analysis information, business process analysis information and diagrams, Calculates the operating rate to calculate the operating rate of the top event as the system operating rate, determines the operating rate to determine whether the calculated system operating rate has reached the reference value, and the calculated system operating rate is the standard If it is determined that the value has not been reached, based on the relationship between the unavailability corresponding to the basic event and the system unavailability corresponding to the top event Perform basic event extraction to extract basic events related to system availability increase, perform change possibility determination to determine whether or not downtime corresponding to the extracted basic event can be reduced, and extract If it is determined that it is possible to reduce the unavailability corresponding to the extracted basic event, new failure analysis information corresponding to the extracted basic event is reset in the availability calculation unit, and the extracted basic event If it is determined that the unavailability corresponding to the above cannot be reduced, the first resetting is performed to reset new failure information and new system configuration information in the information input unit, and the extracted basic event is handled. When new failure analysis information is reset, the processing after the availability calculation is performed, and when new failure information and new system configuration information are reset, the function to perform the processing after failure information analysis is added to the computer. Executed It is to let you.

  According to the present invention, there are provided an information system reliability evaluation system, a reliability evaluation method, and a reliability evaluation program that can improve the system operation rate and reduce the burden on the designer who designs the information system. can do.

  An embodiment of the present invention will be described with reference to the drawings.

(Information system reliability evaluation system)
The information system according to the present embodiment includes a plurality of devices (for example, servers). Each device includes software for executing the function of the device and hardware constituting the device. The reliability evaluation system for an information system according to this embodiment is a system that performs an evaluation of a system operation rate and cost evaluation of such an information system.

  FIG. 1 is a diagram showing a block configuration of an information system reliability evaluation system (hereinafter referred to as a reliability evaluation system) according to the present embodiment.

  The reliability evaluation system according to the present embodiment includes a system evaluation unit 1 that exchanges information between the system management database 2 and the user terminal 3.

  The system evaluation unit 1 exchanges various types of information including information necessary for reliability evaluation of the information system with the system management database 2. The system management database 2 stores various types of information including information necessary for reliability evaluation of the information system.

  The system evaluation unit 1 extracts information necessary for reliability evaluation of the information system from the system management database 2 and inputs it, and performs predetermined processing / calculation based on the input information. An information processing / arithmetic unit 12 to perform, and an information output unit 13 that outputs the result of the predetermined processing / calculation to the user terminal 3 are provided.

  A user terminal 3 is connected to the system evaluation unit 1. The user terminal 3 outputs various information such as various instruction information (for example, information instructing to perform reliability evaluation of a predetermined information system) to the system evaluation unit 1, or displays various information. It is a device to do.

  The information extraction / input unit 11 includes, as information necessary for reliability evaluation of an information system including a device (for example, a server), software for executing the function of the device and hardware constituting the device. Failure information (for example, failure information included in reliability-related information 22 described later) relating to a failure corresponding to a failure mode of a component (a software component and a hardware component described later), the device, and the component System configuration information (for example, system configuration included in system function configuration information 21 and system reference information 24 described later) indicating system configuration information using information, and system operating rate indicating the operating rate of the entire information system A reference value (for example, a reference value of the system operation rate included in the system reference information 24 described later) and the information system Business process information about Tsutomu process (e.g., time variation data of the business processing number of information systems in the business process-related information 23 to be described later) is the information input unit and the information including at least is input.

  The information extraction / input unit 11 includes a facility cost indicating a cost necessary for the hardware component and the software component, a cost allowable value indicating an allowable range of the cost necessary for the information system, and information An operating loss amount (for example, a system stoppage loss amount) indicating an operating loss amount due to the system being stopped for a unit period and an operation period of the information system are input.

  FIG. 2 is a diagram illustrating information input to the information extraction / input unit 11.

  The input information includes system function configuration information 21, reliability related information 22, business process related information 23, system reference information 24, and cost related information 25. These pieces of information are associated with the information system.

  The system function configuration information 21 includes component information including hardware and software included in the information system. This component information includes hardware components (hereinafter referred to as hardware components) and software components (hereinafter referred to as software components). A hardware component is a component (for example, a component or a circuit) that configures a device (for example, a server) that configures an information system. A software component is a component for executing functions of devices constituting the information system (for example, a command or a procedure for executing each function described by a program code or the like).

  As an apparatus constituting the information system, for example, there are a WEB server 1, a WEB server 2,... Belonging to a WEB server group for providing content to a user using a WEB browser. In other words, the WEB server group indicates a set of WEB servers 1, 2,... Constituting the information system. In this case, as a software component corresponding to the WEB server 1, for example, there is WEB server 1 software for causing the function of the WEB server 1 to be executed. Examples of hardware components corresponding to the WEB server 1 include a hard disk, a CPU, and a memory that constitute the WEB server 1.

  Further, the system function configuration information 21 includes a function-specific classification of components including hardware and software included in the information system. The functional classification of components including hardware and software includes a functional classification of hardware components and a functional classification of software components. The hardware component classification by function indicates information in which each hardware component is classified according to the function of the device. The software component classification by function indicates information in which each software component is classified according to the function of the device.

  For example, in the function-based classification of software components, the functions of the WEB server 1 are classified so as to be associated with WEB program 1, WEB program 2,... , AP programs 2... Are associated with each other. An AP server is a server that belongs to an AP server group for receiving a request from a user and executing a business system process using a database or the like. In other words, the AP server group indicates a set of AP servers 1, 2,... Constituting the information system.

  Further, the system function configuration information 21 includes a connection of components including hardware and software included in the information system. The connection between components including hardware and software includes a connection between hardware components, a connection between software components, and a connection between hardware components and software components. The connection between hardware components is, for example, an electrical connection relationship between hardware components in each device. The connection between software components is, for example, a flowchart showing a relationship between functional blocks indicating functions of software components in each device. The connection between the hardware component and the software component is associated with, for example, the hardware component related to the function of the software component for each function of the software component in each device. It is information.

  In addition, the system function configuration information 21 includes the processing contents of the components including the hardware and software included in the information system (the processing contents of the hardware components and the processing contents of the software components).

  The reliability-related information 22 includes failure intervals of failure modes of components including hardware and software included in the information system (failure intervals of failure modes of hardware components and failure modes of failure modes of software components) In addition, failure information such as a repair time of a failure mode of a component including hardware and software (a repair time of a failure mode of a hardware component and a repair time of a failure mode of a software component) is included. For example, the failure interval of the hardware component failure mode includes a failure interval τx of the hard disk failure (failure mode) of the AP server 1, and the failure interval of the failure mode of the software component includes the AP server 1 failure interval. There is a failure interval τy of an AP failure (failure mode) of the AP software.

  The business process related information 23 includes time variation data of the number of business processes in the information system. The time variation data is data indicating the time variation of the number of business processes per time interval of the information system. For example, when the information system is a stock online trading system, the data is time variation data of the number of transactions processed per time interval.

  The system reference information 24 includes the system configuration of the information system. This system configuration indicates a connection between devices included in the information system. The system standard information 24 includes a system operating rate reference value (hereinafter referred to as a system operating rate reference value) indicating an operating rate of the entire information system, an operation period of the information system, and a cost spent for the information system. A cost tolerance that indicates the allowable range is included.

  The cost related information 25 includes equipment costs of components including hardware and software included in the information system (equipment costs indicating costs required for hardware components and costs required for software components). Equipment costs shown). In addition, the cost related information 25 includes an operating loss amount indicating an operating loss amount (which may be an actual measurement value or a predicted value) due to the information system being stopped for a unit period (for example, one day), for example, a system stopping loss amount. Is included. Equipment costs are costs for components, such as raw material costs for hardware components and software components, and costs required to incorporate the components into the information system.

  Information input by the information extraction / input unit 11 is sent to the information processing / calculation unit 12.

  FIG. 3 is a diagram showing a detailed configuration of the information processing / arithmetic unit 12.

  The information processing / arithmetic unit 12 includes a field data analysis unit 31, a fault tree creation unit 32, a fault tree-logic conversion unit 33, a business process number data analysis unit 34, an operation rate calculation unit 35, a system operation rate determination unit 36, an important A degree analysis unit 37, a changeability determination unit 38, a system configuration resetting unit 39, a failure information resetting unit 40, a cost calculation unit 41, a cost determination unit 42, and a cost information resetting unit 43.

<Field data analysis department>
The field data analysis unit 31 is a failure information analysis unit that acquires failure analysis information as an analysis result by analyzing failure information regarding failures corresponding to failure modes of hardware components and software components for each failure mode. .

  For example, the field data analysis unit 31 includes failure information related to a failure corresponding to the failure mode among the information input by the information extraction / input unit 11 for each device (for example, the AP server 1 and the AP server) included in the information system. 2) is classified for each device group (for example, AP server group, WEB server group, etc.) and a failure mode indicating a set of devices classified by processing function, and the above-described analysis is performed based on the classification result. The specific explanation is as follows.

  For example, as failure information (information related to failure) of each failure mode (for example, hard disk failure of the AP server 1), the use start date and time of the above component (for example, hard disk of the AP server 1), failure (for example, disk failure) ) Date of occurrence, failure interval, failure recovery date and time, and repair time. The field data analysis unit 31 includes failure information related to a failure corresponding to the failure mode (for example, a use start date and time, a failure occurrence date and time, a failure interval corresponding to a disk failure of the hard disk of the AP server 1) among various pieces of input information. The failure recovery date and time and the repair time) are classified by associating them with device groups and failure modes (for example, AP server group, disk failure, etc.). In this way, the field data analysis unit 31 generates a classification table in which failure information corresponding to failure modes of hardware components and software components is classified in association with device groups and failure modes.

  Here, FIG. 4 is a diagram illustrating a classification table (hereinafter referred to as a hardware classification table) in which failure information corresponding to a failure mode of a hardware component is classified in association with a device group and a failure mode. Failure information is associated with each device group, device, component, and failure mode (for example, disk failure). FIG. 5 is a diagram showing a classification table (hereinafter referred to as a software classification table) in which failure information corresponding to a failure mode of a software component is classified in association with a device group and a failure mode. Failure information is associated with each device group, device, component, and failure mode (for example, OS failure, application failure, middleware failure, user input error, etc.). The DB server is a server belonging to a DB server group that manages databases included in the information system. In other words, the DB server group indicates a set of DB servers 1, 2,... Constituting the information system.

  The field data analysis unit 31 uses the hardware classification table shown in FIG. 4 (or the software classification table shown in FIG. 5) to determine the failure mode for each failure mode of the hardware component (or software component). Calculate the failure probability. For example, the field data analysis unit 31 records the failure interval of the failure mode of the hardware component (or software component) input to the information extraction / input unit 11 (the above hardware classification table (or software classification table)). The failure probability of the failure mode of the hardware component (or software component) is calculated based on the failure interval of each failure mode) and the Weibull distribution indicating the distribution of failure probability with respect to the failure interval. The specific explanation is as follows.

For example, the field data analysis unit 31 sets the failure interval τ corresponding to the disk failure of the hard disk of the AP server 1 recorded in the hardware classification table (or the OS failure of the software of the AP server 1 recorded in the software classification table). By applying ₁ , τ ₂ , τ ₃ ,... To the Weibull distribution formula, the failure probability using the Weibull distribution formula is calculated. For example, the distribution function F (τ) of the failure probability with respect to the failure interval τ is expressed by the following equation (1).

  Here, β is a shape parameter, and θ is a scale parameter.

The field data analysis unit 31 can determine (estimate) β and θ by fitting the above-described failure intervals τ ₁ , τ ₂ , τ ₃ ,... Thereby, the field data analysis unit 31 can calculate the failure probability distribution function F (τ) of the failure mode of the hardware component and the software component.

  The failure pattern is one of an initial failure pattern in which the failure rate decreases with the passage of time, an accidental failure pattern in which the failure rate is constant regardless of the passage of time, and an aging failure pattern in which the failure rate increases with the passage of time. being classified. Here, when β <1, it corresponds to an initial failure pattern, when β = 1, it corresponds to an accidental failure pattern, and when β> 1, it corresponds to an aging failure pattern. For this reason, the failure probability calculated as described above corresponds to any failure pattern among all failure patterns.

  In addition, the field data analysis unit 31 uses the hardware classification table shown in FIG. 4 (or the software classification table shown in FIG. 5) to detect a failure for each hardware component (or software component) failure mode. An average repair time, which is an average value of the mode repair time, is calculated.

  The field data analysis unit 31 may calculate a failure interval from the use start date / time and the failure occurrence date / time, and calculate the failure probability using the calculated failure interval. The field data analysis unit 31 may calculate the repair time from the failure occurrence date and time and the failure recovery date and time, and may calculate the average repair time using the calculated repair time. The failure analysis information obtained as an analysis result by the field data analysis unit 31 is sent to the operation rate calculation unit 35.

<Fault tree creation part>
The fault tree creation unit 32 indicates the failure mode of the information system based on the system configuration information from the information extraction / input unit 11 (for example, the system configuration included in the system function configuration information 21 and the system reference information 24). It is a diagram creation unit that creates a fault tree indicating a hierarchical logical relationship from an event to a basic event indicating a failure mode of hardware components and software components.

  In this fault tree, the failure mode of the hardware component and the failure mode of the software component cannot be determined any more because the top event is a loss of function of the entire information system or an undesirable event in operation or operation. A hierarchical logical relationship from the top event to the basic event is shown. That is, in the fault tree, each event developed as a relationship between an event and a factor is connected by a logical symbol (for example, an OR symbol or an AND symbol).

  Specifically, the fault tree creation unit 32 creates a fault tree as follows. Component information including hardware and software input by the information extraction / input unit 11, classification according to function of components including hardware and software, connection of components including hardware and software, and hardware The processing contents of the components including software and the system configuration are input to the fault tree creation unit 32. The fault tree creation unit 32 creates a fault tree based on the input information.

  It is a well-known technique to create a fault tree based on classification of hardware components by function, connection between hardware components, processing contents of hardware components, and system configuration. In the case of the present embodiment as well, the fault tree is created in the same manner as this well-known technique, so a detailed description of the fault tree creation is omitted here.

  6 to 12 are diagrams illustrating an example of the fault tree created by the fault tree creating unit 32. FIG.

  In this example, it is assumed that the information system to be evaluated includes the above-described Web server group, AP server group, and DB server group. Each server group is composed of two servers. Each server is assumed to be composed of a plurality of hardware components (hard disk, CPU, memory, etc.). The software components for executing the functions of each server are Web server 1 software, Web server 2 software, AP server 1 software, and so on. Software component failure modes include OS failures, application failures, middleware failures, and user input errors.

  6 to 12, a failure event (for example, Web server 1) lower than a hardware failure event (for example, Web server 1 hardware loss, Web server 2 hardware loss, etc.) of each server. In some cases, failure events are logically connected by AND symbols or OR symbols below the hardware failure events of each server.

  In FIG. 6, a top event (an event indicating a failure of the entire information system, for example, loss of system function) and a first level event (an event hierarchically lower than the top event, for example, a WEB server group) Loss of function, etc.) are connected by logical symbols such as AND symbols and OR symbols. In FIG. 6, the event of the first layer and the event of the second layer (the event two layers lower than the top event, for example, the loss of the function of the WEB server 1) are connected by the above logical symbols. ing. Further, in FIG. 6, the event of the second layer and the event of the third layer (the event three layers lower than the top event, for example, the loss of the function of the WEB server 1 hardware, the loss of the function of the WEB server 1 software) Etc.) are connected by the above logic symbols.

  Here, among the events of the third hierarchy, the hardware failure event of each server (for example, loss of function of the WEB server 1 hardware) is a basic event related to hardware failure. In addition, among the events of the third layer, the failure event of the software of each server (for example, the loss of the function of the WEB server 1 software) further includes the event of the fourth layer (see FIG. 7 to FIG. 12). For example, an OS failure of the WEB server 1 software, an application failure, a middleware failure, a user input error, and the like) are connected by logical symbols. Here, the events of the fourth hierarchy shown in FIGS. 7 to 12 are basic events related to software failures.

  In such a fault tree, when a hierarchical logical relationship from the top event to the basic event is shown, each event (event of each layer, basic event) includes not only hardware failure events but also software events. There is a fault event related to. The created fault tree is sent to the fault tree-logic conversion unit 33.

<Fault tree-logic converter>
The fault tree-logic conversion unit 33 converts the data of the fault tree into a predetermined relational expression described later using a Boolean algebra. The specific explanation is as follows.

6 to 12, a predetermined event (for example, WEB server 1 software loss of function) and a plurality of lower events (WEB server 1 software OS failure, application failure, middleware failure, WEB server 1 software) Is connected with an OR symbol, when any of the plurality of lower events (for example, OS failure) occurs, the predetermined event also occurs. Therefore, when a predetermined event and a lower event of the predetermined event are connected by an OR symbol, a Boolean variable corresponding to the predetermined event is set to Y, and Y = 1 when a predetermined event occurs, When a predetermined event does not occur, Y = 0 is defined. A Boolean variable corresponding to a lower event of the predetermined event is defined as Yj. When a lower event of the predetermined event occurs, Yj = 1, When Yj = 0 is defined when no event occurs, a Boolean variable Y of a predetermined event is calculated by the following equation (2) using a Boolean algebra.

  Here, N is the total number of lower events.

In addition, in FIGS. 6 to 12, a predetermined event (for example, loss of function of the WEB server group) and a plurality of lower-level events (loss of function of the WEB server 1, loss of function of the WEB server 2) are connected by AND symbols. If all of the plurality of lower events occur (for example, loss of function of the WEB server 1, loss of function of the WEB server 2), the predetermined event occurs. Therefore, when a predetermined event and a plurality of lower-level events are connected by an AND symbol, a Boolean variable corresponding to the predetermined event is set to Y, and when a predetermined event occurs, Y = 1, a predetermined event When an event does not occur, Y = 0 is defined, and a Boolean variable corresponding to a lower event of the predetermined event is defined as Yj. When an event lower than the predetermined event occurs, Yj = 1, the predetermined event is When Yj = 0 is defined when it does not occur, a Boolean variable Y of a predetermined event is calculated by the following equation (3) using a Boolean algebra.

  Here, N is the total number of lower events.

  The fault tree-logic conversion unit 33 converts the fault tree data into a predetermined relational expression related to Y from the Boolean variable of the top event, using the fault trees, formulas 2 and 3 in FIGS. . In this predetermined relational expression, the Boolean variable Y corresponding to the top event is, for example, a predetermined expression using Yj (Yj is a Boolean variable of a basic event, j = 1 to N, N is the number of basic events). It is represented by The predetermined relational expression is sent to the operation rate calculation unit 35.

<Business Processing Number Data Analysis Department>
The business process number data analysis unit 34 analyzes the information related to the time variation data of the number of business processes per time interval of the information system as the business process related information 23, thereby acquiring the business process analysis information as the analysis result. Information analysis department.

  For example, the business process number data analysis unit 34 performs the above-described analysis based on the information related to the time variation data of the business process number per time interval of the information system among the information input by the information extraction / input unit 11. . The specific explanation is as follows.

  The time variation data of the number of business processes per time interval of the information system includes, for example, the time and the number of business processes per time interval at that time. Based on the data, the business processing number data analysis unit 34 generates a table (hereinafter referred to as a time variation table of the number of business processing cases) indicating the time variation of the number of business processing cases per time interval as shown in FIG.

  The business processing number data analysis unit 34 uses the time variation table of the number of business processing cases shown in FIG. 13 described above, and the failure occurrence probability per time interval of the failure mode of the hardware component (or software component). The weight factor to be applied to is calculated. Here, the weighting factor is defined by a numerical value obtained by dividing the number of business processes per time interval at each time by the average value of the number of business processes per time interval.

  For example, the business process count data analysis unit 34 sums the business process counts per unit time of each time recorded in the time variation table of the business process counts to obtain the average of the business process counts per time interval. The average value is calculated. Then, the business processing number data analysis unit 34 divides the number of business processings per time interval at each time by the average value, thereby obtaining the per time time interval of the failure mode of the hardware component (or software component). The weighting factor to be applied to the failure occurrence probability is calculated.

  When the time variation table of the number of business processes shown in FIG. 13 is used, the failure occurrence probability of each time interval in the failure mode of the hardware component (or software component) as shown in FIG. A table showing the time variation of the weighting factor (hereinafter referred to as the time variation table of the weighting factor applied to the failure occurrence probability) is obtained. The business process analysis information obtained as an analysis result by the business process number data analysis unit 34 is sent to the operation rate calculation unit 35.

<Occupancy rate calculation unit>
The operation rate calculation unit 35 analyzes the analysis result (calculated failure probability of each failure mode, average repair time) by the field data analysis unit 31 and the analysis result (calculated hardware component (calculated hardware component ( (Or a weighting factor applied to the probability of occurrence of failure per time interval of each failure mode of the software component) failure time to calculate the failure rate corresponding to the failure mode of the hardware component and software component A rate calculation function, an analysis result by the field data analysis unit 31 (a failure probability of each failure mode, an average repair time), and an analysis result by the business processing number data analysis unit 34 (a calculated hardware component (or software) Component) weighting factor for failure occurrence probability per time interval of failure mode) and Based on a Boolean variable Yj representing the state of the basic event obtained from the fault tree (for example, Boolean variable Yj representing the state of the hard disk due to a disk failure) and a predetermined relational expression regarding the Boolean variable Y representing the state of the top event A system operating rate calculation function for calculating an operating rate corresponding to the event as a system operating rate; The specific explanation is as follows.

  The operating rate calculation unit 35 first calculates an operating rate corresponding to each basic event included in the fault tree by evaluation using a Monte Carlo simulation. The operating rate corresponding to the basic event is calculated by 1- (non-operating rate corresponding to the basic event), for example. The field data analysis unit 31 outputs the calculated failure probability F (t) of each failure mode to the operation rate calculation unit 35. The failure probability F (t) indicates the probability that a failure will occur between time 0 and time t when each failure mode is repaired. The failure rate p (t) is a normal state from time 0 to time t when each failure mode is repaired, and the probability that a failure will occur per unit time at time t (p (t) = ( dF (t) / dt) / (1-F (t))).

  The business process number data analysis unit 34 outputs the weighting factor w (t) related to the failure occurrence probability per time interval between the time t and the time t + dt in the failure mode to the operation rate calculation unit 35. From these, it is assumed that the failure occurrence probability per time interval in which each failure mode is in the normal state from time 0 to time t when the failure mode is repaired and becomes in the failure state between time t and time t + dt is w (t) p ( t) Redefine as dt. Here, assuming that the weighting factor w (t) relating to the failure occurrence probability per time interval between the time t and the time t + dt in the failure mode is 1, the number of normal business processes is regarded as being constant on average over time. It becomes evaluation. The operating rate calculation unit 35 calculates the operating rate corresponding to each basic event as follows.

"Calculation of utilization rate corresponding to each basic event"
(1) operating rate calculating unit 35, a simulation evaluation start time (hereinafter, evaluation start time) as the t _{0 =} 0, the simulation evaluation ending time (hereinafter, evaluation ending time) to the case of a t _e, evaluation start time At t ₀ , a failure corresponding to the basic event does not occur, and the hardware component or software component corresponding to the basic event is recognized as being in a normal state. That is, the operation rate calculation unit 35 recognizes that the probability F (t ₀ ) that the failure corresponding to the basic event occurs and the failure rate p (t ₀ ) are both 0. The operating rate calculation unit 35 recognizes that the cumulative operating time corresponding to the basic event indicating the time during which the component corresponding to the basic event is operating between the evaluation start time and the evaluation end time is zero.

(2) The operating rate calculation unit 35 is configured to perform a failure per time interval in which a failure corresponding to a basic event occurs between a simulation evaluation time (hereinafter referred to as evaluation time) t ₀ and an evaluation time t ₀ + dt (= t ₁ ). The failure occurrence probability (probability that the hardware component or software component corresponding to the basic event will fail) is assumed to be w (t ₀ ) p (t ₀ + dt / 2) dt.

The operating rate calculation unit 35 compares the above-described probability w (t ₀ ) p (t ₀ + dt / 2) dt with the random number r ₁ generated in the range of 0 to 1, and w (t ₀ ) p (t ₀ + dt / 2) When dt ≧ r ₁ , it is recognized that a hardware component or software component (hereinafter simply referred to as a component) has transitioned to a failure state corresponding to a basic event.

  The operating rate calculation unit 35 adds the time until the failure corresponding to the basic event is repaired and the component returns to the normal state to the accumulated operating time corresponding to the basic event after recognizing that the transition has occurred. First, dt is added to the failure time 0 corresponding to the basic event. The failure time corresponding to the basic event indicates the time during which the component corresponding to the basic event has failed, and is cleared to 0 once the component returns to the normal state.

The operation rate calculation unit 35 compares the probability that a failure corresponding to the above-described basic event occurs with the random number r ₁ generated in the range of 0 to 1, and w (t ₀ ) p (t ₀ + dt / 2). In the case of dt <r ₁ , it is recognized that the component remains in a normal state. In this case, the operation rate calculation unit 35 adds the time dt to the accumulated operation time corresponding to the basic event.

(3) Next, the operation rate calculation unit 35 evaluates the state transition of the component corresponding to the basic event between the evaluation time t ₁ and the evaluation time t ₁ + dt (= t ₂ ). If the components in evaluation time t ₁ corresponding to the basic event is normal state, operating rate calculating unit 35 performs the same processing as above. The specific explanation is as follows.

In the same manner as described above, the operation rate calculation unit 35 calculates the failure occurrence probability per time interval at which the failure of the component corresponding to the basic event occurs between the evaluation time t ₁ and the evaluation time t ₁ + dt as w ( Let t ₁ ) p (t ₁ + dt / 2) dt.

The operation rate calculation unit 35 compares the probability w (t ₁ ) p (t ₁ + dt / 2) dt with the random number r ₂ generated in the range of 0 to 1, and calculates w (t ₁ ) p (t ₁ + dt / 2) When dt ≧ r ₂ , it is recognized that the component has transitioned to the failure state.

  The operating rate calculation unit 35 adds the time until the failure corresponding to the basic event is repaired and the component returns to the normal state to the accumulated operating time corresponding to the basic event after recognizing that the transition has occurred. First, dt is added to the failure time corresponding to the basic event.

The operating rate calculation unit 35 compares the above-described probability w (t ₁ ) p (t ₁ + dt / 2) dt with the random number r ₂ generated in the range of 0 to 1, and w (t ₁ ) p (t ₁ + dt / 2) in the case of dt _{<r 2} recognizes components to remain in the normal state. In this case, the operation rate calculation unit 35 adds the time dt to the accumulated operation time corresponding to the basic event.

Furthermore, the operation rate calculation unit 35 sets the evaluation time to t ₃ (= t ₂ + dt), t ₄ (= t ₃ + dt),..., T _n (= t _n−1 + dt),. The process of (3) described above is repeated while increasing by dt.

  In addition, in parallel with the above-described processing, the operation rate calculation unit 35 compares the failure time corresponding to the basic event with the average repair time of the failure mode corresponding to the basic event, and the failure time corresponding to the basic event. It is determined whether or not the average repair time has been reached.

When the failure time corresponding to the basic event reaches the average repair time, the operating rate calculation unit 35 recognizes that the component corresponding to the basic event has returned to the normal state. In this case, the operation rate calculation unit 35 resets the failure time corresponding to the basic event to 0, and also resets the failure rate and failure probability to 0. Then, when the time when the failure rate or the like is reset is t _m , the operating rate calculation unit 35 performs the process (4).

(4) The operation rate calculation unit 35 calculates a failure occurrence probability per time interval in which a failure corresponding to the basic event occurs between the evaluation time t _m and the evaluation time t _m + dt (= t _{m + 1} ) w (t _m) and _{p (t m -t m + dt} / 2) dt.

The operating rate calculation unit 35 compares the above-described probability w (t _m ) p (t _m −t _m + dt / 2) dt with the random number r ₃ generated in the range of 0 to 1, and w (t _m ) When p (t _m −t _m + dt / 2) dt ≧ r ₃ , it is recognized that the component has transitioned to the failure state corresponding to the basic event.

  The operating rate calculation unit 35 adds the time until the failure corresponding to the basic event is repaired and the component returns to the normal state to the accumulated operating time corresponding to the basic event after recognizing that the transition has occurred. First, dt is added to the failure time 0 corresponding to the basic event. The failure time corresponding to the basic event indicates the time during which the component corresponding to the basic event has failed, and is cleared to 0 once the component returns to the normal state.

The operation rate calculation unit 35 compares the probability that a failure corresponding to the basic event described above will occur with a random number r ₃ generated in the range of 0 to 1, and calculates w (t _m ) p (t _m −t _m + dt / 2) When dt <r ₃ , it is recognized that the component remains in a normal state. In this case, the operation rate calculation unit 35 adds the time dt to the accumulated operation time corresponding to the basic event.

Next, the operation rate calculation unit 35 evaluates the state transition of the component corresponding to the basic event between the evaluation time t _{m + 1} and the evaluation time t _{m + 1} + dt (= t _{m + 2} ). When the component corresponding to the basic event is in the normal state at the evaluation time t _{m + 1} , the operating rate calculation unit 35 performs the same process as described above. The specific explanation is as follows.

In the same manner as described above, the operation rate calculation unit 35 sets the failure occurrence probability per time interval at which the failure of the component corresponding to the basic event occurs between the evaluation time t _{m + 1} and the evaluation time t _{m + 1} + dt to w ( t _{m + 1} ) p (t _{m + 1} −t _m + dt / 2) dt.

The operating rate calculation unit 35 compares the probability w (t _{m + 1} ) p (t _{m + 1} −t _m + dt / 2) dt with the random number r ₄ generated in the range from 0 to 1, and determines w (t _{m + 1} ) p ( t _{m + 1} −t _m + dt / 2) When dt ≧ r ₄ , it is recognized that the component has transitioned to the failure state.

  The operating rate calculation unit 35 adds the time until the failure corresponding to the basic event is repaired and the component returns to the normal state to the accumulated operating time corresponding to the basic event after recognizing that the transition has occurred. First, dt is added to the failure time corresponding to the basic event.

The operation rate calculation unit 35 compares the above-described probability w (t _{m + 1} ) p (t _{m + 1} −t _m + dt / 2) dt with the random number r ₄ generated in the range of 0 to 1, and w (t _{m + 1} ) When p (t _{m + 1} −t _m + dt / 2) dt <r ₄ , it is recognized that the component remains in a normal state. In this case, the operation rate calculation unit 35 adds the time dt to the accumulated operation time corresponding to the basic event.

Furthermore, the operation rate calculation unit 35 sets the evaluation time to t _{m + 3} (= t _{m + 2} + dt), t _{m + 4} (= t _{m + 3} + dt),..., T _n (= t _n-1 + dt),. The above process is repeated by incrementing dt.

  In addition, in parallel with the above-described processing, the operation rate calculation unit 35 compares the failure time corresponding to the basic event with the average repair time of the failure mode corresponding to the basic event, and the failure time corresponding to the basic event. It is determined whether or not the average repair time has been reached.

When the failure time corresponding to the basic event reaches the average repair time, the operating rate calculation unit 35 recognizes that the component corresponding to the basic event has returned to the normal state. In this case, the operation rate calculation unit 35 resets the failure time corresponding to the basic event to 0, and also resets the failure rate and failure probability to 0. When the time to failure rate and the like are reset and the t _k, operating rate calculating unit 35 performs processing for the t _m in the above-described processing of (4) was replaced by t _k.

(5) operating rate calculating section 35, (2) above, the process of (3) and (4) are repeated until the evaluation time becomes _{t e.} Instead of the average repair time, the repair time for each failure mode preset in the operation rate calculation unit 35 may be used.

If the evaluation time has reached a t _e, operating rate calculating unit 35, operation rate q ₁ to the value obtained by dividing the accumulated operating time corresponding to the basic event evaluation time (t e _{-t 0),} corresponding to the basic event Calculate as Similarly, the operation rate calculation unit 35 performs the above-described simulation evaluation for all basic events included in the fault tree, and the operation rates q ₁ , q ₂ ,..., Q _n corresponding to the basic events. Is calculated.

  In the above description, the operation rate corresponding to the basic event is calculated by simulating the temporal behavior in which the component corresponding to the basic event transitions between the normal state or the failure state using the Monte Carlo method. The procedure is shown. The state transition behavior obtained from the above sequence of procedures (called a history in the field of Monte Carlo simulation) is one of all the histories stochastically expected. Therefore, the value of the operating rate calculated from this one history is one point in the statistical variation of the operating rate.

  For this reason, in order to evaluate the statistical average value of the operation rate, the operation rate calculation unit 35 repeatedly generates different random numbers many times in the above series of procedures, and obtains a large number of histories. Therefore, it is necessary to calculate the statistical average value of the operation rate. The statistical average value of the utilization rate converges to a specific value as the number of histories increases. In the simulation evaluation for calculating the operating rate described above, an evaluation error width of the operating rate to be evaluated is set in advance, and the repetition of the above procedure results in the convergence width (variation range) of the statistical average value of the operating rate. This is preferably performed until the evaluation error width is equal to or less.

  In calculating the operating rate corresponding to each basic event described above, assuming that the weight factor w (t) applied to the failure occurrence probability per time interval between the failure mode time t and time t + dt is 1, normal business It is possible to calculate the operation rate that the number of processing cases is regarded as constant over time.

  The operating rate calculation unit 35 calculates the system operating rate corresponding to the top event of the fault tree by evaluation using a Monte Carlo simulation. The operating rate calculation unit 35 calculates the system operating rate as follows.

"Calculating system availability"
(1) operating rate calculating unit 35, the evaluation start time and t _{0 =} 0, when the evaluation ending time and a t _e, the evaluation start time t _0, no occurrence of failures corresponding to the basic event, the basic The hardware component or software component corresponding to the event is recognized as being in a normal state. That is, the operation rate calculation unit 35 recognizes that the probability F (t ₀ ) and the failure rate p (t ₀ ) that a failure corresponding to the basic event occurs is 0. From the evaluation start time to the evaluation end time, the cumulative operating time corresponding to the top event indicating the time during which the system corresponding to the top event is operating is recognized as zero.

(2) The operation rate calculation unit 35 calculates a failure occurrence probability per time interval (a basic event is generated) between the evaluation time t ₀ and the evaluation time t ₀ + dt (= t ₁ ). The probability that the corresponding hardware component or software component will fail is assumed to be w (t ₀ ) p (t ₀ + dt / 2) dt.

The operating rate calculation unit 35 compares the above-described probability w (t ₀ ) p (t ₀ + dt / 2) dt with the random number r ₁ generated in the range of 0 to 1, and w (t ₀ ) p (t ₀ + dt / 2) When dt ≧ r ₁ , it is recognized that the component (hardware component or software component) has transitioned to the failure state corresponding to the basic event. In this case, the operation rate calculation unit 35 adds dt to the failure time 0 corresponding to the basic event. The failure time corresponding to the basic event indicates the time during which the component corresponding to the basic event has failed, and is cleared to 0 once the component returns to the normal state.

The operation rate calculation unit 35 compares the probability that a failure corresponding to the above-described basic event occurs with the random number r ₁ generated in the range of 0 to 1, and w (t ₀ ) p (t ₀ + dt / 2). In the case of dt <r ₁ , it is recognized that the component remains in a normal state.

The availability calculating unit 35 performs the above-described evaluation for all the basic events, and whether all the basic events are in a failure state between the evaluation time t ₀ and the evaluation time t ₀ + dt (= t ₁ ). , Recognize whether it is in a normal state. Predetermined logical relationship between the Boolean variable representing the state of the top event from the fault tree and the state of each basic event at the bottom of the tree structure indicating whether the Boolean variable representing the state of the top event is in the fault state or the normal state The following relational expression is obtained. The operating rate calculation unit 35 recognizes whether the top event is in a failure state or a normal state by substituting a Boolean variable representing the state of each basic event into this predetermined relational expression.

  The operation rate calculation unit 35 does not add the time from when the top event is recognized as having transitioned to the failure state to when the top event returns to the normal state, to the cumulative operation time corresponding to the top event. Moreover, when the operation rate calculation unit 35 recognizes that the top event remains in the normal state, the operation rate calculation unit 35 adds the time dt to the accumulated operation time corresponding to the top event.

(3) Next, the operation rate calculation unit 35 evaluates the state transition of the component corresponding to the basic event between the evaluation time t ₁ and the evaluation time t ₁ + dt (= t ₂ ). If the components in evaluation time t ₁ corresponding to the basic event is normal state, operating rate calculating unit 35 performs the same processing as above. The specific explanation is as follows.

In the same manner as described above, the operation rate calculation unit 35 calculates the failure occurrence probability per time interval at which the failure of the component corresponding to the basic event occurs between the evaluation time t ₁ and the evaluation time t ₁ + dt as w ( Let t ₁ ) p (t ₁ + dt / 2) dt.

The operation rate calculation unit 35 compares the probability w (t ₁ ) p (t ₁ + dt / 2) dt with the random number r ₂ generated in the range of 0 to 1, and calculates w (t ₁ ) p (t ₁ + dt / 2) When dt ≧ r ₂ , it is recognized that the component has transitioned to the failure state. In this case, the operation rate calculation unit 35 adds dt to the failure time corresponding to the basic event.

The operating rate calculation unit 35 compares the above-described probability w (t ₁ ) p (t ₁ + dt / 2) dt with the random number r ₂ generated in the range of 0 to 1, and w (t ₁ ) p (t ₁ + dt / 2) in the case of dt _{<r 2} recognizes components to remain in the normal state.

The operating rate calculation unit 35 performs the above-described evaluation for all the basic events, and whether all the basic events are in a failure state between the evaluation time t ₁ and the evaluation time t ₁ + dt (= t ₂ ). , Recognize whether it is in a normal state. Predetermined logical relationship between the Boolean variable representing the state of the top event from the fault tree and the state of each basic event at the bottom of the tree structure indicating whether the Boolean variable representing the state of the top event is in the fault state or the normal state The following relational expression is obtained. The operating rate calculation unit 35 recognizes whether the top event is in a failure state or a normal state by substituting a Boolean variable representing the state of each basic event into this predetermined relational expression.

  The occupancy rate calculation unit 35 does not add the time until the top event returns to the normal state after recognizing that the top event has transitioned to the failure state, and the top event is normal. If it is recognized that the state remains, the time dt is added to the cumulative operation time corresponding to the top event.

Furthermore, the operation rate calculation unit 35 sets the evaluation time to t ₃ (= t ₂ + dt), t ₄ (= t ₃ + dt),..., T _n (= t _n−1 + dt),. The process of (3) described above is repeated while increasing by dt.

  In addition, in parallel with the above-described processing, the operation rate calculation unit 35 compares the failure time corresponding to the basic event with the average repair time of the failure mode corresponding to the basic event, and the failure time corresponding to the basic event. It is determined whether or not the average repair time has been reached.

When the failure time corresponding to the basic event reaches the average repair time, the operating rate calculation unit 35 recognizes that the component corresponding to the basic event has returned to the normal state. In this case, the operation rate calculation unit 35 resets the failure time corresponding to the basic event to 0, and also resets the failure rate and failure probability to 0. Then, when the time when the failure rate or the like is reset is t _m , the operating rate calculation unit 35 performs the process (4).

(4) The operation rate calculation unit 35 calculates a failure occurrence probability per time interval in which a failure corresponding to the basic event occurs between the evaluation time t _m and the evaluation time t _m + dt (= t _{m + 1} ) w (t _m) and _{p (t m -t m + dt} / 2) dt.

The operating rate calculation unit 35 compares the above-described probability w (t _m ) p (t _m −t _m + dt / 2) dt with the random number r ₃ generated in the range of 0 to 1, and w (t _m ) When p (t _m −t _m + dt / 2) dt ≧ r ₃ , it is recognized that the component has transitioned to the failure state corresponding to the basic event. In this case, the operation rate calculation unit 35 adds dt to the failure time 0 corresponding to the basic event. The failure time corresponding to the basic event indicates the time during which the component corresponding to the basic event has failed, and is cleared to 0 once the component returns to the normal state.

The operation rate calculation unit 35 compares the probability that a failure corresponding to the basic event described above will occur with the random number r ₃ generated in the range of 0 to 1, and w (t _m ) p (t _m −t _m + dt / 2) When dt <r ₃ , it is recognized that the component remains in a normal state.

The occupancy rate calculating unit 35 performs the above-described evaluation on all the basic events, and whether all the basic events are in a failure state between the evaluation time t _m and the evaluation time t _m + dt (= t _{m + 1} ). , Recognize whether it is in a normal state. Predetermined logical relationship between the Boolean variable representing the state of the top event from the fault tree and the state of each basic event at the bottom of the tree structure indicating whether the Boolean variable representing the state of the top event is in the fault state or the normal state The following relational expression is obtained. The operating rate calculation unit 35 recognizes whether the top event is in a failure state or a normal state by substituting a Boolean variable representing the state of each basic event into this predetermined relational expression.

  The occupancy rate calculation unit 35 does not add the time until the top event returns to the normal state after recognizing that the top event has transitioned to the failure state, and the top event is normal. If it is recognized that the state remains, the time dt is added to the cumulative operation time corresponding to the top event.

Next, the operation rate calculation unit 35 evaluates the state transition of the component corresponding to the basic event between the evaluation time t _{m + 1} and the evaluation time t _{m + 1} + dt (= t _{m + 2} ). When the component corresponding to the basic event is in the normal state at the evaluation time t _{m + 1} , the operating rate calculation unit 35 performs the same process as described above. The specific explanation is as follows.

In the same manner as described above, the operation rate calculation unit 35 sets the failure occurrence probability per time interval at which the failure of the component corresponding to the basic event occurs between the evaluation time t _{m + 1} and the evaluation time t _{m + 1} + dt to w ( t _{m + 1} ) p (t _{m + 1} −t _m + dt / 2) dt.

The operating rate calculation unit 35 compares the probability w (t _{m + 1} ) p (t _{m + 1} −t _m + dt / 2) dt with the random number r ₄ generated in the range from 0 to 1, and determines w (t _{m + 1} ) p ( t _{m + 1} −t _m + dt / 2) When dt ≧ r ₄ , it is recognized that the component has transitioned to the failure state. In this case, the operation rate calculation unit 35 adds dt to the failure time corresponding to the basic event.

The operation rate calculation unit 35 compares the above-described probability w (t _{m + 1} ) p (t _{m + 1} −t _m + dt / 2) dt with the random number r ₄ generated in the range of 0 to 1, and w (t _{m + 1} ) When p (t _{m + 1} −t _m + dt / 2) dt <r ₄ , it is recognized that the component remains in a normal state.

The occupancy rate calculating unit 35 performs the above-described evaluation for all the basic events, and whether all the basic events are in a failure state between the evaluation time t _{m + 1} and the evaluation time t _{m + 1} + dt (= t _{m + 2} ). , Recognize whether it is in a normal state. Predetermined logical relationship between the Boolean variable representing the state of the top event from the fault tree and the state of each basic event at the bottom of the tree structure indicating whether the Boolean variable representing the state of the top event is in the fault state or the normal state The following relational expression is obtained. The operating rate calculation unit 35 recognizes whether the top event is in a failure state or a normal state by substituting a Boolean variable representing the state of each basic event into this predetermined relational expression.

  The occupancy rate calculation unit 35 does not add the time until the top event returns to the normal state after recognizing that the top event has transitioned to the failure state, and the top event is normal. If it is recognized that the state remains, the time dt is added to the cumulative operation time corresponding to the top event.

Furthermore, the operation rate calculation unit 35 sets the evaluation time to t _{m + 3} (= t _{m + 2} + dt), t _{m + 4} (= t _{m + 3} + dt),..., T _n (= t _n-1 + dt),. The above process is repeated by incrementing dt.

  In addition, in parallel with the above-described processing, the operation rate calculation unit 35 compares the failure time corresponding to the basic event with the average repair time of the failure mode corresponding to the basic event, and the failure time corresponding to the basic event. It is determined whether or not the average repair time has been reached.

When the failure time corresponding to the basic event reaches the average repair time, the operating rate calculation unit 35 recognizes that the component corresponding to the basic event has returned to the normal state. In this case, the operation rate calculation unit 35 resets the failure time corresponding to the basic event to 0, and also resets the failure rate and failure probability to 0. When the time to failure rate and the like are reset and the t _k, operating rate calculating unit 35 performs processing for the t _m in the above-described processing of (4) was replaced by t _k.

(5) operating rate calculating section 35, the above-mentioned (2), the processing of (3) and (4) are repeated until the evaluation time becomes _{t e.} Instead of the average repair time, the repair time for each failure mode preset in the operation rate calculation unit 35 may be used.

If the evaluation time has reached a t _e, operating rate calculating unit 35 is a utilization rate of a value obtained by dividing the accumulated operating time corresponding to the basic event evaluation time (t e _{-t 0),} corresponding to the top event Calculated as the system operating rate q. The calculated system operation rate q is sent to the system operation rate determination unit 36.

  The above explanation is based on the Monte Carlo method, and the system operation corresponding to the top event is simulated by simulating the temporal behavior of the system corresponding to the top event transitioning between the normal state or the failure state. The procedure for calculating the rate is shown. The state transition behavior obtained from the above sequence of procedures (called a history in the field of Monte Carlo simulation) is one of all the histories stochastically expected. Therefore, the value of the operating rate calculated from this one history is one point in the statistical variation of the operating rate.

  For this reason, in order to evaluate the statistical average value of the operation rate, the operation rate calculation unit 35 repeatedly generates different random numbers many times in the above series of procedures, and obtains a large number of histories. Therefore, it is necessary to calculate the statistical average value of the operation rate. The statistical average value of the utilization rate converges to a specific value as the number of histories increases. In the simulation evaluation for calculating the operating rate described above, an evaluation error width of the operating rate to be evaluated is set in advance, and the repetition of the above procedure results in the convergence width (variation range) of the statistical average value of the operating rate. This is preferably performed until the evaluation error width is equal to or less.

  In the above-described calculation of the system operation rate, if the weighting factor w (t) relating to the failure occurrence probability per time interval between the failure mode time t and time t + dt is 1, the number of normal business processes is temporal. It is possible to calculate the system operation rate that is considered to be constant on average.

<System availability check part>
The system operation rate determination unit 36 is an operation rate determination unit that determines whether the system operation rate calculated by the operation rate calculation unit 35 is greater than or equal to the system operation rate reference value from the information extraction / input unit 11. . When the calculated system operating rate is smaller than the system operating rate reference value, the system operating rate determining unit 36 outputs information indicating that fact to the importance analyzing unit 37. Further, when the calculated system operating rate is equal to or greater than the system operating rate reference value, the system operating rate determining unit 36 outputs information indicating the fact to the cost calculating unit 41.

<Importance analysis part>
When it is determined that the calculated system operating rate has not reached the system operating rate reference value, the importance analysis unit 37 determines that the non-operating rate corresponding to the basic event (= 1-operating rate for the basic event) and the top event Is a basic event extracting unit that extracts a basic event related to an increase in the system operating rate based on the relationship with the system non-operating rate (= 1-system operating rate) corresponding to. The specific explanation is as follows.

  The importance level analysis unit 37 applies each basic event (hardware) to the system non-operation rate (= 1-system operation rate) calculated by the operation rate calculation unit 35 based on the input information indicating the above. The extent to which the failure mode of the hardware component and the failure mode of the software component have an influence is analyzed. For example, the importance level analysis unit 37 calculates the degree of contribution (ratio) of the unusable rate (= 1−operating rate) of each basic event to the system unusable rate (= 1−system operating rate) of the top event. Based on the calculated contribution, the importance analysis unit 37 extracts a basic event that is important for a decrease in the system non-operation rate, that is, an increase (improvement) in the system operation rate, and responds to the basic event. Extract components.

  Further, the importance analysis unit 37 extracts a basic event (for example, an OS failure in the AP server 2 software of the AP server 2) that contributes to an increase in the operating rate corresponding to the highest event. Then, the importance analysis unit 37 extracts a component (for example, AP server 2 software) corresponding to the basic event. The information extracted by the importance analysis unit 37 is sent to the change possibility determination unit 38.

<Changeability determination unit>
Based on the basic events and components extracted by the importance analysis unit 37, the change possibility determination unit 38 can increase the operating rate corresponding to the basic event (decreasing the non-operating rate corresponding to the basic event). It is determined whether or not. The specific explanation is as follows.

  The system management database 2 stores failure probability list data (hereinafter referred to as failure probability list data) corresponding to the failure mode of each component. The system management database 2 stores average repair time list data (hereinafter referred to as average repair time list data) corresponding to the failure modes of the respective components. If the failure probability corresponding to the failure mode of a predetermined component is recorded in the failure probability list data, it is assumed that there is a component that operates with the failure probability. Further, when the average repair time corresponding to the failure mode of a predetermined component is recorded in the average repair time list data, it is assumed that the failure can be recovered with the average repair time.

  The change possibility determination unit 38 refers to the failure probability list data sent via the information extraction / input unit 11, and refers to the failure probability corresponding to the extracted basic event (for example, the AP server 1 software X1 Failure probability corresponding to the failure of the OS A1) and failure probability corresponding to the failure mode of the component corresponding to the basic event recorded in the failure probability list data (for example, the failure of the OS of the AP server 1 software X2) Are compared with the failure probability B1 corresponding to the failure of the OS of the AP server 1 software X3. The AP server 1 software X1, X2, and X3 have the same function, but have different manufacturers (for example, different performance for realizing the function).

  Whether the change possibility determination unit 38 has a failure probability corresponding to the failure mode of the component corresponding to the basic event included in the failure probability list data that is smaller than the current failure probability corresponding to the basic event. Determine whether or not. If there is a change probability determination unit 38 that is smaller than the current failure probability corresponding to the basic event, the change in the operating rate corresponding to the basic event (decreasing the non-operating rate corresponding to the basic event) Is output to the failure information resetting unit 40, and the failure probability corresponding to the failure mode of the component corresponding to the basic event related to the increase in the operation rate (decrease in the non-operation rate) is output. On the other hand, if there is no change probability determination unit 38 that is smaller than the current failure probability corresponding to the basic event, the change possibility determination unit 38 performs the following processing.

  The change possibility determination unit 38 refers to the average repair time list data sent via the information extraction / input unit 11, the current average repair time of the failure mode corresponding to the extracted basic event, Compared to the average repair time corresponding to the failure mode of the component corresponding to the basic event recorded in the list data, and the average repair corresponding to the failure mode of the component corresponding to the basic event included in the list data It is determined whether there is any time smaller than the current average repair time of the failure mode corresponding to the basic event.

  If there is a small change, the change possibility determination unit 38 determines that the operating rate corresponding to the basic event can be increased (decrease in the non-operating rate corresponding to the basic event), and the failure information is restored. The setting unit 40 outputs the average repair time of the failure mode corresponding to the basic event related to the increase in the operation rate (decrease in the non-operation rate).

  On the other hand, the change possibility determination unit 38 has no failure probability list data smaller than the current failure probability corresponding to the basic event, and the failure corresponding to the basic event in the average repair time list data. If there is no mode smaller than the current average repair time, information indicating that the downtime corresponding to the extracted basic event cannot be reduced is output to the system configuration resetting unit 39.

<System configuration resetting section>
The system configuration resetting unit 39 responds to new system configuration information and a new failure mode when it is determined by the changeability determining unit 38 that the downtime corresponding to the extracted basic event cannot be reduced. The failure information regarding the failure to be reset is reset in the information extraction / input unit 11. The system configuration resetting unit 39 constitutes a part of the first resetting unit. The specific explanation is as follows.

  The system configuration resetting unit 39 includes a new system configuration, new system function configuration information 21, new reliability related information 22, new system reference information 24, and new cost related information input by the system evaluator. 25 is reset in the information extraction / input unit 11. Thereafter, as described above, the processing after the processing by the field data analysis unit 31 is performed.

<Failure information resetting unit>
When it is determined that the downtime corresponding to the extracted basic event can be reduced, the failure information resetting unit 40 determines a new failure probability or a new failure mode corresponding to the extracted basic event. The average repair time is reset in the operation rate calculation unit 35. The failure information resetting unit 40 constitutes a part of the first resetting unit. The specific explanation is as follows.

  The failure information resetting unit 40 resets the failure probability in the operation rate calculating unit 35 when the failure probability corresponding to the failure mode of the component corresponding to the basic event is input from the change possibility determination unit 38. In this case, the operation rate calculation unit 35 holds a failure probability corresponding to each basic event, and is reset when the failure probability corresponding to the basic event is reset by the failure information resetting unit 40. Based on the failure probability corresponding to the basic event and the failure probability of the basic event other than the basic event, the system operation rate is calculated. Thereafter, the processing after the calculation of the system operation rate as described above is performed again.

  Further, the failure information resetting unit 40 resets the average repair time in the operation rate calculating unit 35 when the average repair time corresponding to the failure mode corresponding to the basic event is input from the change possibility determination unit 38. To do. In this case, the operation rate calculation unit 35 holds an average repair time corresponding to each basic event, and resets when the average repair time corresponding to the basic event is reset by the failure information resetting unit 40 The system operation rate is calculated based on the average repair time corresponding to the basic event and the average repair time of basic events other than the basic event. Thereafter, the processing after the calculation of the system operation rate as described above is performed again.

<Cost calculation unit>
The cost calculation unit 41 calculates the sum of the facility costs of the components included in the information system when the system operation rate determination unit 36 determines that the calculated system operation rate has reached the system operation rate reference value. In addition, the expected operating loss is calculated based on the system operating rate calculated by the operating rate calculating unit 35, the system stoppage loss, and the operation period of the information system, and the above sum and the operating loss expected value are calculated. Is calculated as a cost. The specific explanation is as follows.

  The cost calculation unit 41 performs the following processing based on the cost related information 25 and the system operation rate reference value sent from the information extraction / input unit 11. The cost calculation unit 41 first calculates the sum of the facility costs of each component included in the cost related information 25. Next, the cost calculation unit 41 calculates the expected value of the operating loss due to the system stop when the information system is operated for a predetermined period based on the calculated system operation rate, the system stop loss amount, and the operation period. Calculate the expected value of operating loss. The cost calculation unit 41 outputs the calculated cost to the cost determination unit 42. The expected value of operating loss is expressed, for example, by the following equation (4).

[Equation 4]
Expected value of operating loss = (1-calculated system availability) x system outage loss x operating period

<Cost determination unit>
The cost determination unit 42 compares the calculated cost with the cost allowable value sent from the information extraction / input unit 11, and when the cost is smaller than the cost allowable value, the system operation rate reference value and the cost allowable value Information such as a system configuration that satisfies the conditions is output to the information output unit 13. Thereby, information such as the system configuration output from the information output unit 13 is displayed by the user terminal 3. If the cost is greater than the allowable cost value, the cost determining unit 42 outputs information indicating that to the cost information resetting unit 43.

<Cost information resetting section>
When it is determined that the calculated cost exceeds the allowable cost value, the cost information resetting unit 43 sets a new allowable cost value, a new system availability reference value, new system configuration information (for example, New system function configuration information 21, new system configuration of system reference information 24, new cost related information 25), and failure information related to a failure corresponding to a new failure mode (for example, new reliability related information 22) The second resetting unit resets one or more pieces of information in the information extraction / input unit 11. The specific explanation is as follows.

  The cost information resetting unit 43, for example, gives information that instructs the information output unit 13 to input any of the new cost tolerance, the system operating rate reference value, and the new system information. Output. When the user inputs any information using the user terminal 3, the information is output to the cost information resetting unit 43.

  The cost information resetting unit 43 resets the new cost allowable value in the information extraction / input unit 11 when a new cost allowable value is input. As a result, the new cost allowable value is sent to the cost determining unit 42, and the process by the cost determining unit 42 is performed again based on the new cost allowable value.

  The cost information resetting unit 43 resets the new system availability reference value in the information extraction / input unit 11 when a new system availability reference value is input. As a result, the new system operation rate reference value is sent to the system operation rate determination unit 36, and processing subsequent to the process by the system operation rate determination unit 36 is performed based on the new system operation rate reference value.

  The cost information resetting unit 43 resets this information in the information extraction / input unit 11 when information about a new system is input. Then, as described above, the processing after the processing by the field data analysis unit 31 is performed.

(Reliability evaluation method)
Next, a reliability evaluation method using the reliability evaluation system having the above-described configuration will be described below. This reliability evaluation is performed during operation of the information system. In the description of the reliability evaluation method, the same description as that in the above-described reliability evaluation system is omitted.

  First, the evaluator who evaluates the reliability of the information system periodically inputs the failure interval and the repair time of the failure mode of the hardware component and the software component using the user terminal 3. Enter the measured value of the time fluctuation of the number of business processes in the information system.

  The information processing / calculating unit 12 stores the measured values of the failure interval and repair time in the failure mode in the system management database 2 as reliability related information 22. Further, the information processing / calculation unit 12 stores the actual measurement value of the time variation of the number of business processes in the information system as the business process related information 23 in the system management database 2.

  In parallel with the above-described processing, the following reliability evaluation is performed.

  FIG. 15 is a flowchart for explaining a reliability evaluation method using the reliability evaluation system having the above-described configuration.

  As shown in FIG. 15, in step S <b> 10, the evaluator inputs an instruction for reliability evaluation of the information system using the user terminal 3. At this time, information specifying the information system is also input.

  In step S11, the information extraction / input unit 11 extracts business process related information 23 corresponding to information for identifying the information system from the system management database 2 based on an instruction for reliability evaluation of the information system. As a result, the business process related information 23 is input to the information extraction / input unit 11.

  In step S12, the business process count data analysis unit 34 uses the hardware component (or software configuration) as the analysis result of the information regarding the time variation data of the business process count per time interval of the information system based on the input information. The weighting factor concerning the failure occurrence probability per time interval at each time in the failure mode of (element) is calculated. The business process number data analysis unit 34 holds the information input from the information extraction / input unit 11 and also holds the weight factor related to the failure occurrence probability per time interval calculated at each time.

  In step S15, the information extraction / input unit 11 receives from the system management database 2 system function configuration information 21 corresponding to information for specifying the information system, reliability related information, based on an instruction for reliability evaluation of the information system. 22, System reference information 24 and cost related information 25 are extracted. As a result, information such as the system function configuration information 21 is input to the information extraction / input unit 11.

  In step S20, the field data analysis unit 31 calculates the failure probability and the average repair time in the failure mode as the analysis result of the failure information in the failure mode in the hardware component and software component based on the input information. . The field data analysis unit 31 holds information input from the information extraction / input unit 11 and holds the calculated failure probability and average repair time.

  In step S <b> 25, the fault tree creation unit 32 creates the above-described fault tree based on the system configuration included in the system function configuration information 21 and the system reference information 24.

  In step S30, the fault tree-logic conversion unit 33 uses the generated fault tree, Formula 2 and Formula 3, and converts the fault tree data into a predetermined relational expression related to the Boolean variable Y representing the state of the top event. Convert to

  In step S35, the operation rate calculation unit 35 calculates the hardware component (or software) calculated by the failure probability of each failure mode calculated by the field data analysis unit 31, the average repair time, and the business processing number data analysis unit 34. The failure rate (1-operation rate) corresponding to the failure mode of the hardware component and the software component is calculated based on the weighting factor applied to the failure occurrence probability per time interval of each failure mode of the component). To do. Then, the operating rate calculation unit 35 calculates the system operating rate corresponding to the top event based on the Boolean variable Yj that represents the state of the basic event and the predetermined relational expression related to the Boolean variable Y that represents the state of the top event. . Here, the operating rate calculation unit 35 holds the calculated inoperability rate and system operating rate corresponding to each basic event, failure probability and average repair time corresponding to each basic event.

  In step S40, the system operation rate determination unit 36 determines whether the calculated system operation rate is equal to or greater than the system operation rate reference value. If it is determined that the calculated system operating rate is smaller than the system operating rate reference value, the process of step S45 is performed. If it is determined that the calculated system operating rate is greater than or equal to the system operating rate reference value, step S45 is performed. The process of S70 is performed.

  In step S45, the importance analysis unit 37 extracts a basic event related to an increase in the system operation rate, and also extracts components corresponding to the basic event.

  In step S50, the changeability determination unit 38 can increase the operating rate corresponding to the basic event (decreasing the non-operating rate corresponding to the basic event) based on the extracted basic event and component. It is determined whether or not. If it is determined that the operating rate can be increased, the process of step S55 is performed. If it is determined that the operating rate cannot be increased, the process of step S60 is performed.

  In step S <b> 55, the failure information resetting unit 40 resets a new failure probability or a new average repair time corresponding to the extracted basic event in the operation rate calculating unit 35. When a new failure probability corresponding to the extracted basic event is reset in the operating rate calculation unit 35, the operating rate calculation unit 35 extracts the failure probability corresponding to each held basic event. The failure probability corresponding to the basic event is rewritten with a new failure probability. Then, based on the new failure probability corresponding to the extracted basic event and the failure probability (information held) of the basic event other than the extracted basic event, the operation rate calculating unit 35 again performs step S35. The system operation rate calculation process is performed. Then, the process after step S40 is performed.

  When a new average repair time corresponding to the extracted basic event is reset in the operation rate calculation unit 35, the operation rate calculation unit 35 includes the average repair time corresponding to each held basic event. The average repair time corresponding to the extracted basic event is rewritten to a new average repair time. Based on the new average repair time corresponding to the extracted basic event and the average repair time (information held) of the basic events other than the extracted basic event, the operation rate calculating unit 35 again A system operation rate calculation process in step S35 is performed. Then, the process after step S40 is performed.

  In step S60, the system configuration resetting unit 39 sets the new system configuration, the new system function configuration information 21, the new reliability related information 22, the new system reference information 24, and the new cost related information input by the evaluator. The information 25 is reset in the information extraction / input unit 11. Specifically, the system configuration resetting unit 39 displays information instructing the user terminal 3 to input information such as a new system configuration. When the evaluator inputs information such as a new system configuration using the user terminal 3, the information is output to the system configuration resetting unit 39.

  When information such as a new system configuration is reset in the information extraction / input unit 11, the information extraction / input unit 11 reads the reset information (new system configuration, new system function configuration information 21, new Reliability related information 22, new system reference information 24, and new cost related information 25) are output to the field data analysis unit 31. And the process after step S20 is performed again.

  In step S70, the cost calculation unit 41 calculates the cost described above. FIG. 17 is a detailed flowchart of the calculation process performed by the cost calculation unit 41. As shown in FIG. 17, in step S <b> 71, the cost calculation unit 41 calculates the sum of the facility costs of each component included in the cost related information 25. In step S72, the cost calculation unit 41 calculates an expected value of business loss. In step S <b> 73, the cost calculation unit 41 calculates the sum of the total equipment cost of each component and the expected operating loss amount as a cost.

  In step S75, the cost determination unit 42 determines whether or not the calculated cost exceeds the allowable cost value, and if the calculated cost does not exceed the allowable cost value, the process of step S90 is performed. If the calculated cost exceeds the allowable cost value, the process of step S80 is performed.

  In step S80, the cost information resetting unit 43 sets the new cost tolerance, the system operation rate reference value, and the information on the new system (new system configuration and new system function configuration information 21, new The information extraction / input unit 11 is reset with any one of the reliability related information 22, the new system reference information 24, and the new cost related information 25.

  When a new cost allowable range is reset in the information extraction / input unit 11, the information extraction / input unit 11 outputs the new cost allowable range to the cost determination unit 42. The cost determination unit 42 performs the determination process using the calculated cost and the new cost allowable range again.

  When a new system operation rate reference value is reset in the information extraction / input unit 11, the information extraction / input unit 11 outputs the new system operation rate reference value to the system operation rate determination unit 36. The system availability determination unit 36 performs the determination process using the calculated system availability and the new system availability reference value again. Thereafter, the processing after step S45 or the processing after step S70 is performed.

  When information on a new system is input, this information is reset in the information extraction / input unit 11. And as mentioned above, based on the reset information, the process after step S20 is performed.

  In step S <b> 90, the information output unit 13 outputs information such as system function configuration information 21, system reference information 24, and business process related information 23 to the user terminal 3. As a result, information such as system function configuration information 21 and system reference information 24 (information on the configuration of the information system that satisfies the system operation rate reference value and the cost tolerance), and business processing related information 23 are displayed on the user terminal 3. Is done.

(Operational effect according to the embodiment of the present invention)
According to the embodiment of the present invention, the information necessary for the reliability evaluation input to the information extraction / input unit 11 includes, as the business process related information 23, information related to the time variation data of the number of business processes in the information system. It is. Based on the information input to the information extraction / input unit 11, processing from the field data analysis unit 31 to the system operation rate determination unit 36 is performed. At this time, the business process count data analysis unit 34 determines the time of each time in the failure mode of the hardware component (or software component) based on the information on the time variation data of the number of business processes per time interval of the information system. A weighting factor to be applied to the failure occurrence probability per interval is calculated. Then, the operation rate calculating unit 35 analyzes the analysis result (the calculated failure probability of each failure mode and the average repair time) by the field data analyzing unit 31 and the analysis result (the calculated hardware configuration by the business process number data analyzing unit 34). Based on the failure factor of each element in the failure mode of each element (or software component), the operating rate corresponding to the top event is calculated as the system operating rate. As described above, the system operation rate can be calculated in consideration of the fluctuation of the failure occurrence probability due to the time fluctuation of the number of business processes in the system.

  Thereafter, when the system operation rate determination unit 36 determines that the calculated system operation rate has not reached the system operation rate reference value, the processing from the importance analysis unit 37 to the change possibility determination unit 38 is performed. Thereafter, processing by the system configuration resetting unit 39 or the failure information resetting unit 40 is performed. When the processing of the system configuration resetting unit 39 is performed, the processing from the field data analysis unit 31 to the system availability determination unit 36 is performed based on the reset information. Further, when the process of the failure information resetting unit 40 is performed, the process from the operating rate calculating unit 35 to the system operating rate determining unit 36 is performed based on the reset information. Then, the above-described processing can be repeated until the calculated system operation rate reaches the system operation rate reference value.

  In this way, system availability is evaluated in consideration of fluctuations in the probability of failure due to time fluctuations in the number of business processes in the information system. It is possible to evaluate the rapid change in system availability when changing in the vicinity of the band interval. Thus, when the information system configuration is changed and operated in accordance with the system function configuration information 21 and the system reference information 24 corresponding to the evaluated system operation rate, the system operation rate falls below the system operation rate reference value. This situation is avoided. As a result, the situation where the number of redesigns of the information system is increased is avoided, and the burden on the information system designer is reduced. Therefore, according to the present embodiment, it is possible to improve the system operation rate and reduce the burden on the designer who designs the information system.

  Also, if the person who evaluates the system reliability understands the program contents that make up the software related to the operation of the information system, it can know the relationship between the program contents and the failure of the information system. Although we were able to evaluate the operating rate, there are actually few cases where the contents of the software program can be understood. For this reason, system availability is usually not evaluated in consideration of software, and when system availability is evaluated in consideration of software failure events, the person who evaluates system reliability It is necessary to understand the program contents of the software. For this reason, although the burden on the person who performs the above evaluation increases, according to this embodiment, not only the hardware failure event but also the software failure event is considered even if the program contents constituting the software are not known. Therefore, it is possible to reduce the burden on the person who evaluates the system reliability.

  Furthermore, according to the present embodiment, the field data analysis unit 31 extracts failure information (for example, failure information) related to a failure corresponding to the failure mode among the information input by the information extraction / input unit 11, the device group, and Since each failure mode is classified and the above analysis is performed based on the classification result, when performing the analysis processing for each failure mode, it is not necessary to perform processing related to information unrelated to the analysis processing, and the analysis processing is performed quickly. It becomes possible.

  In addition, according to the present embodiment, the field data analysis unit 31 calculates the failure probability of the failure mode using the Weibull distribution, so that the process of calculating the failure probability of the failure mode is quickly performed, and as a result, It is possible to quickly calculate the non-operation rate and the system operation rate of the constituent elements. In addition, it is possible to calculate failure probability of failure mode without distinguishing between software and hardware as a component, so when calculating failure probability of failure mode by distinguishing software and hardware. In comparison, the load on the field data analysis unit 31 can be reduced.

  Furthermore, according to the present embodiment, when the system operation rate determination unit 36 determines that the calculated system operation rate has reached the system operation rate reference value, the cost calculation unit 41 is included in the information system. Calculate the sum of the component equipment costs and the expected operating loss amount, and calculate the sum of the calculated operating loss expected value and the sum of the component equipment costs included in the information system as the cost. The determination unit 42 determines whether or not the calculated cost exceeds the allowable cost value, and when it is determined that the calculated cost exceeds the allowable cost value, the cost information resetting unit 43 Information extraction / input unit for new cost tolerance, new system operation rate reference value, new system configuration information, and failure information related to failure corresponding to new failure mode To re-set to 1. Then, based on the reset information, processing by the system operating rate determination unit 36 and processing by the cost determining unit 42 are performed, and the calculated system operating rate reaches the system operating rate reference value and is calculated. The above process can be repeated until the cost is kept within the allowable cost value. For this reason, it is possible to provide a system that improves the system operation rate and suppresses the cost required for realizing and operating the system.

(Modification of the embodiment of the present invention)
(1) In the embodiment described above, the field data analysis unit 31 may calculate the failure probability of the failure mode as follows. In the system management database 2, for each failure mode, the failure interval of the failure mode (τ ₁ , τ ₂ ,...) And the failure probability (f (τ ₁ ), f () associated with this failure interval. τ ₂ ),...) measured value data (however, data calculated by simulation or the like may be stored). The field data analysis unit 31 has a failure mode failure interval (τ ₁ , τ ₂ ,...) Input by the information extraction / input unit 11 and a failure probability (f (τ ₁ ) associated with the failure interval. ), F (τ ₂ ),...)) Are estimated and the failure mode distribution function (F (τ)) is estimated to calculate the failure mode failure probability. Also good.

  According to the present modification example, the field data analysis unit 31 calculates the failure probability of the failure mode using data indicating the relationship between the failure interval of the failure mode and the failure probability (for example, actual measurement data). The failure probability of the failure mode is calculated more accurately, and as a result, the unavailability of the component can be calculated more accurately.

(2) Further, in the above-described embodiment, the above-described information system reliability evaluation system operates during the operation of the information system. However, the present invention is not limited to this, and the information system reliability evaluation system operates in the information system design stage. May be. In this case, in the system reference information 24, the system configuration is a system configuration proposal, and the system operation rate reference value is a design reference value of the system operation rate. Further, the reliability related information 22 is not a measured value acquired during operation of the information system but a predicted value obtained by a preset specification value or simulation.

(3) Furthermore, in the above-described embodiment, the information extraction / input unit 11 extracts information necessary for reliability evaluation from the system management database 2, but the present invention is not limited to this, and the information extraction / input unit 11, information necessary for reliability evaluation may be input by the user terminal 3. Further, the failure information included in the classification tables shown in FIGS. 4 and 5 may include information related to the maintenance of the failure mode. And based on the information regarding this maintenance, the above-mentioned average repair time may be calculated.

(4) In the above-described embodiment, the change possibility determination unit 38 may perform the following processing. The change possibility determination unit 38 causes the user terminal 3 to display information inquiring whether there is a component having a failure probability lower than the current failure probability corresponding to the extracted component of the basic event.

  When there is a component having a failure probability lower than the current failure probability, the user terminal 3 receives information indicating that, a component, and a failure probability. When there is no component having a failure probability lower than the current failure probability, the user terminal 3 receives information indicating that. Based on the information input to the user terminal 3, the change possibility determination unit 38 determines whether there is a component having a failure probability lower than the current failure probability corresponding to the component corresponding to the extracted basic event. You may make it determine.

  Similarly, the change possibility determination unit 38 receives information asking whether there is a component having an average repair time shorter than the current average repair time corresponding to the failure mode of the component corresponding to the extracted basic event. Based on the information displayed on the terminal 3 and input to the user terminal 3 in the same manner as described above, the change possibility determination unit 38 determines whether the current failure mode of the component corresponding to the extracted basic event corresponds to the current failure mode. It may be determined whether there is a component having an average repair time shorter than the average repair time.

(5) Still further, in the above-described embodiment, the above-described various information related to hardware components and software components (system function configuration information 21, information related to component failure modes, cost-related information 25) is used. Although reliability evaluation is performed, the present invention is not limited to this, and the above-described reliability evaluation is performed using various pieces of information related to a component group composed of hardware components and a component group composed of software components. May be performed.

  For example, when the hardware components are the CPU, memory, and hard disk of the AP server 1, the component groups configured by the hardware components are the AP server 1 and the AP server group. For example, when the software components are AP server 1 software 1, AP server 1 software 2,... Constituting the AP server 1 software, the component group constituted by the software components is AP server 1 software. That is. In addition, various types of information (system function configuration information 21, information about the failure mode of the component, cost-related information 25, etc.) and various types of information about the component group (system function configuration information 21, component element group information) The above-described reliability evaluation may be performed based on information on the failure mode and cost-related information 25).

  In this way, even when there is no data on failure in units of components, if there is data on failures in units of component groups, the above-described reliability evaluation becomes possible, so detailed information on failures for reliability evaluation is possible. It is possible to avoid a situation where reliability evaluation cannot be performed when there is no information.

(6) In the above-described embodiment, the cost calculation unit 41 calculates the sum of the total equipment cost of each component and the expected operating loss amount as the cost. However, the present invention is not limited to this. Only the sum of the equipment costs of each component may be calculated. Then, the cost information resetting unit 43 includes a new cost tolerance, information on a new system (new system configuration and new system function configuration information 21, new reliability related information 22, new system reference information 24, Any one or more pieces of information in the new cost related information 25) may be reset in the information extraction / input unit 11.

(7) In the above-described embodiment, the time variation data of the number of business processes per time interval is used as the business process related information 23. However, the present invention is not limited to this. For example, the number of accesses per time interval and the time interval Time variation data such as the amount of data processing per hit may be used.

(8) In the above-described embodiment, the fault tree is used as the diagram. However, the present invention is not limited to this. From the top event indicating the failure mode of the information system to the basic event indicating the failure mode of the component or component group. It is only necessary to use a diagram showing a hierarchical logical relationship, and a reliability block diagram may be used as shown in FIG.

  Here, FIG. 17 is a schematic diagram showing an example of a fault tree relating to a loss of function of the WEB server 1 software created by the fault tree creation unit 32 of the present embodiment shown in FIG. 7 as a reliability block diagram. is there. In FIG. 17, the OS of the WEB server 1 software, the application of the WEB server 1 software, the middleware of the WEB server 1 software, and the user input of the WEB server 1 software are connected in series. The normal operation of the WEB server 1 software is the logical relationship achieved when the OS of the WEB server 1 software, the application of the WEB server 1 software, the middleware of the WEB server 1 software, and the user input of the WEB server 1 software all operate normally. It is shown.

  That is, in the reliability block diagram shown in FIG. 17, a failure or a mistake occurs in any of the OS of the WEB server 1 software, the application of the WEB server 1 software, the middleware of the WEB server 1 software, or the user input of the WEB server 1 software. If the function is lost, it means that the function of the WEB server 1 software is lost, and the logical relationship is equivalent to the fault tree shown in FIG.

(Reliability evaluation program)
The computer can perform processing in the reliability evaluation system of the information system of the present embodiment and its modified example by executing a reliability evaluation program of the information system (hereinafter referred to as a reliability evaluation program). That is, the reliability evaluation program is a program that causes a computer to execute the functions of the information extraction / input unit 11, the information processing / calculation unit 12, and the information output unit 13 of the present embodiment and the modified example thereof. According to such a reliability evaluation program, it is possible to easily realize a reliability evaluation system that exhibits the operational effects shown in the present embodiment and its modified examples with a general-purpose computer.

  The reliability evaluation program can be recorded on a computer-readable recording medium. As this recording medium, there are, for example, a hard disk 1100, a flexible disk 1200, a compact disk 1300, an IC chip 1400, and a cassette tape 1500 as shown in FIG. According to the recording medium in which such a reliability evaluation program is recorded, the program can be easily stored, transported, sold, and the like.

  As mentioned above, although an example of the embodiment of the present invention and an example of the change was explained, it is only a specific example, does not limit the present invention in particular, and the concrete composition of each part can be changed suitably. . Further, the actions and effects described in the embodiment and its modified examples are merely a list of the most preferable actions and effects resulting from the present invention, and the actions and effects according to the present invention are not limited to the embodiments of the present invention and modifications thereof. It is not limited to those described in the examples.

It is a block diagram which shows the function structure of the reliability evaluation system which concerns on this embodiment. It is a figure which shows the information input into the information extraction and input part which comprises the system evaluation part with which the reliability evaluation system shown in FIG. 1 is provided. It is a block diagram which shows the processing function structure of the information processing and calculating part which comprises the system evaluation part with which the reliability evaluation system shown in FIG. 1 is provided. It is a figure which shows the classification table by which the failure information corresponding to the failure mode of a hardware component was matched and matched with the server group and failure mode. It is a figure which shows the classification table by which the failure information corresponding to the failure mode of a software component was matched and matched with the server group and failure mode. It is a figure which shows an example of the fault tree created by the fault tree creation part which comprises the information processing and calculating part shown in FIG. It is a figure which shows a part of fault tree shown in FIG. It is a figure which shows a part of fault tree shown in FIG. It is a figure which shows a part of fault tree shown in FIG. It is a figure which shows a part of fault tree shown in FIG. It is a figure which shows a part of fault tree shown in FIG. It is a figure which shows a part of fault tree shown in FIG. It is a figure which shows the time variation table of the number of business processings produced | generated by the business processing number data analysis part which comprises the information processing and calculating part shown in FIG. It is a figure which shows the time fluctuation table of the weighting factor computed by the business processing number data analysis part which comprises the information processing and calculating part shown in FIG. It is a flowchart for demonstrating the reliability evaluation method which concerns on this embodiment. It is a flowchart for demonstrating the calculation process of the cost which concerns on this embodiment. It is the figure which showed a part of fault tree shown in FIG. 7 with the reliability block diagram. It is a figure which shows the recording medium which stores the program of this embodiment and its modification. It is a figure which shows the time fluctuation data of the number of business processing of a stock online transaction system as an example of the time fluctuation data of the number of business processing of an information system.

Explanation of symbols

DESCRIPTION OF SYMBOLS 11 ... Information input part (information extraction / input part), 21 ... System function structure information, 22 ... Reliability related information, 23 ... Business processing related information, 24 ... System reference information, 25 ... Cost related information, 31 ... Failure information Analysis unit (field data analysis unit), 32 ... Diagram creation unit (fault tree creation unit), 34 ... Business process information analysis unit (business process number data analysis unit), 35 ... Operation rate calculation unit, 36 ... Operation rate determination unit (System operation rate determination unit), 37 ... basic event extraction unit (importance analysis unit), 38 ... change possibility determination unit, 39, 40 ... first resetting unit (system configuration resetting unit, failure information resetting unit) 41... Cost calculation unit 42. Cost determination unit 43 43 Second resetting unit (cost information resetting unit)

Claims (19)

As information necessary for reliability evaluation of an information system including a device, a component indicating each of software for executing the function of the device and hardware configuring the device, or a configuration configured by the component Failure information related to a failure corresponding to a failure mode of an element group, system configuration information indicating information related to a configuration of the information system using the device and the component or the component group, and an operating rate of the entire information system An information input unit for inputting information including at least a reference value of a system operation rate indicating business processing information regarding business processing of the information system;
A failure information analysis unit that analyzes the failure information for each failure mode and acquires failure analysis information as an analysis result;
Creation of a diagram showing a hierarchical logical relationship from a top event indicating a failure mode of the information system to a basic event indicating a failure mode of the component or group of components based on the system configuration information And
Analyzing the business process information, and obtaining a business process analysis information as an analysis result;
The operating rate of the basic event is calculated based on the failure analysis information and the business process analysis information, and the operating rate of the top event is calculated based on the failure analysis information, the business process analysis information, and the diagram. An operation rate calculation unit to calculate as
An operation rate determination unit that determines whether or not the calculated system operation rate has reached the reference value;
When it is determined that the calculated system operating rate has not reached the reference value, the system is based on the relationship between the operating rate corresponding to the basic event and the system operating rate corresponding to the top event. A basic event extraction unit that extracts the basic event related to an increase in availability,
A change possibility determination unit that determines whether or not the unavailability corresponding to the extracted basic event can be reduced;
If it is determined that the downtime corresponding to the extracted basic event can be reduced, the failure analysis information corresponding to the extracted basic event is reset in the uptime calculating unit. First resetting that resets the new failure information and the new system configuration information in the information input unit when it is determined that the unavailability corresponding to the extracted basic event cannot be reduced And
An information system reliability evaluation system comprising: The business process information includes time variation data of the number of business processes per time interval of the information system,
The business process information analysis unit analyzes the time variation data, calculates a weighting factor to be applied to a failure occurrence probability per time interval of the failure mode of the component or the component group, and acquires the business process analysis information The information system reliability evaluation system according to claim 1, wherein: The failure information analysis unit acquires the failure occurrence probability as the failure analysis information,
The operating rate calculation unit multiplies the failure occurrence probability acquired by the failure information analysis unit by the weight factor acquired by the business process information analysis unit, and uses the failure occurrence probability multiplied by the weight factor. The information system reliability evaluation system according to claim 2, wherein an operation rate of the basic event and the system operation rate are calculated.   The failure information analysis unit includes the failure information input to the information input unit for each of the device group and the failure mode indicating a set of devices in which the plurality of devices included in the information system are classified according to processing functions. The information system reliability evaluation system according to claim 1, wherein the failure analysis information is acquired based on the classification result. The failure information input to the information input unit includes a failure interval of a failure mode of the component or the component group,
The failure information analysis unit is configured based on the failure interval of the failure mode of the component or the component group input to the information input unit, and a Weibull distribution indicating a distribution of failure probability with respect to the failure interval. 5. The reliability evaluation system for an information system according to claim 1, wherein a failure probability of a failure mode of an element or the component group is calculated and the failure analysis information is acquired. The failure information input to the information input unit includes a failure interval of a failure mode of the component or the component group, and a failure probability corresponding to the failure interval,
The failure information analysis unit is configured to calculate a failure probability for the failure interval based on the failure interval of the failure mode of the component or the component group input to the information input unit and the failure probability corresponding to the failure interval. 5. The distribution according to claim 1, wherein the failure probability of the failure mode of the component or the component group is calculated based on the failure probability distribution, and the failure analysis information is acquired. The reliability evaluation system of the information system as described in any one. In the information input unit, a facility cost indicating a cost required for the component or the component group, and a cost allowable value indicating an allowable range of a cost required for the information system are input,
When it is determined that the calculated system operation rate has reached the reference value, a cost calculation unit that calculates a total of equipment costs of the component or the component group as a cost,
A cost determination unit that determines whether the calculated cost exceeds the cost allowable value;
When it is determined that the calculated cost exceeds the allowable cost value, the information input unit stores at least one of the new failure information, the new system configuration information, and the new allowable cost value. A second resetting unit resetting to
The reliability evaluation system for an information system according to any one of claims 1 to 6, further comprising: In the information input unit, an operating loss amount indicating an operating loss amount due to the information system being stopped for a unit period, and an operation period of the information system are input.
The cost calculating unit calculates an operating loss expected value indicating an expected value of the operating loss based on the calculated system operating rate, the operating loss, and the operation period, and calculates the operating loss calculated Calculate the sum of the expected value and the sum of the equipment costs as the cost,
The second resetting unit, when it is determined that the calculated cost exceeds the cost tolerance, the new failure information, the new system configuration information, the new reference value, and the new 8. The reliability evaluation system for an information system according to claim 7, wherein any one or more information of an allowable cost value is reset in the information input unit.   The information system reliability evaluation system according to any one of claims 1 to 8, wherein the diagram is a fault tree or a reliability block diagram. As information necessary for reliability evaluation of an information system including a device, a component indicating each of software for executing the function of the device and hardware configuring the device, or a configuration configured by the component Failure information related to a failure corresponding to a failure mode of an element group, system configuration information indicating information related to a configuration of the information system using the device and the component or the component group, and an operating rate of the entire information system Input information to input information including at least the reference value of the system operation rate indicating business processing information related to the business processing of the information system,
Analyzing the business process information, performing business process information analysis to obtain business process analysis information as an analysis result,
Analyzing the failure information for each failure mode, performing failure information analysis to obtain failure analysis information as an analysis result,
Creation of a diagram showing a hierarchical logical relationship from a top event indicating a failure mode of the information system to a basic event indicating a failure mode of the component or group of components based on the system configuration information And
The operating rate of the basic event is calculated based on the failure analysis information and the business process analysis information, and the operating rate of the top event is calculated based on the failure analysis information, the business process analysis information, and the diagram. Calculate the operating rate calculated as
Performing an operation rate determination to determine whether the calculated system operation rate has reached the reference value;
When it is determined that the calculated system operating rate has not reached the reference value, the system is based on the relationship between the operating rate corresponding to the basic event and the system operating rate corresponding to the top event. Perform basic event extraction to extract the basic events related to the increase in availability,
Performing a change possibility determination to determine whether or not the unavailability corresponding to the extracted basic event can be reduced;
If it is determined that the downtime corresponding to the extracted basic event can be reduced, the failure analysis information corresponding to the extracted basic event is reset in the uptime calculating unit. First resetting that resets the new failure information and the new system configuration information in the information input unit when it is determined that the unavailability corresponding to the extracted basic event cannot be reduced And
When new failure analysis information corresponding to the extracted basic event is reset, processing after the operation rate calculation is performed, and new failure information and new system configuration information are reset A method for evaluating the reliability of an information system, wherein the processing after the failure information analysis is performed. The business process information includes time variation data of the number of business processes per time interval of the information system,
In the business process information analysis, the time variation data is analyzed, a weight factor applied to a failure occurrence probability per time interval of the failure mode of the component or the component group is calculated, and the business process analysis information is acquired. The reliability evaluation method for an information system according to claim 10. In the failure information analysis, the failure occurrence probability is acquired as the failure analysis information,
In the operation rate calculation, the failure occurrence probability obtained by the failure information analysis unit is multiplied by the weight factor obtained by the business process information analysis unit, and the failure occurrence probability obtained by multiplying the weight factor is used. 12. The information system reliability evaluation method according to claim 11, wherein the operating rate of the basic event and the system operating rate are calculated.   In the failure information analysis, the failure information input to the information input unit is classified for each device group indicating a set of devices in which the plurality of devices included in the information system are classified by processing function and for each failure mode. 13. The information system reliability evaluation method according to claim 10, wherein the failure analysis information is acquired based on the classification result. The failure information input to the information input unit includes a failure interval of a failure mode of the component or the component group,
In the failure information analysis, based on the failure interval of the failure mode of the component or the group of components input to the information input unit and the Weibull distribution indicating the distribution of failure probability with respect to the failure interval, the component The reliability evaluation method for an information system according to any one of claims 10 to 13, wherein a failure probability of a failure mode of the component group is calculated and the failure analysis information is acquired. The failure information input to the information input unit includes a failure interval of a failure mode of the component or the component group, and a failure probability corresponding to the failure interval,
In the failure information analysis, based on the failure interval of the failure mode of the component or the component group input to the information input unit and the failure probability corresponding to the failure interval, the distribution of failure probability with respect to the failure interval 14. The failure probability of the failure mode of the component or the component group is calculated based on the failure probability distribution, and the failure analysis information is acquired. The reliability evaluation method of the information system as described in kaichi. In the information input, an equipment cost indicating a cost required for the component or the group of components and a cost allowable value indicating an allowable range of a cost required for the information system are input.
When it is determined that the calculated system operation rate has reached the reference value, a cost calculation is performed to calculate a total of equipment costs of the component or the component group as a cost,
A cost determination is performed to determine whether or not the calculated cost exceeds the allowable cost value,
When it is determined that the calculated cost exceeds the allowable cost value, the information on one or more of the new failure information, the new system configuration information, and the new allowable cost value is reset. 2 Re-set,
When the new cost tolerance is reset, the processing after the cost calculation is performed, and when the new system configuration information and the new failure information are reset, the processing after the information analysis is performed. The method for evaluating reliability of an information system according to any one of claims 10 to 15, wherein: In the information input, an operating loss amount indicating an operating loss amount due to the information system being stopped for a unit period and an operation period of the information system are input.
In the cost calculation, an operating loss expected value indicating an expected value of the operating loss is calculated based on the calculated system operation rate, the operating loss, and the operation period, and the calculated operating loss expected Calculate the sum of the value and the sum of the equipment costs as the cost,
In the second resetting, when it is determined that the calculated cost exceeds the allowable cost value, the new failure information, the new system configuration information, the new reference value, and the new cost Reset any one or more of the tolerance values,
17. The information system reliability evaluation method according to claim 16, wherein when the new reference value is reset, the processing after the operation rate determination is performed.   The information system reliability evaluation method according to any one of claims 10 to 17, wherein the diagram is a fault tree or a reliability block diagram. As information necessary for reliability evaluation of an information system including a device, a component indicating each of software for executing the function of the device and hardware configuring the device, or a configuration configured by the component Failure information related to a failure corresponding to a failure mode of an element group, system configuration information indicating information related to a configuration of the information system using the device and the component or the component group, and an operating rate of the entire information system Input information to input information including at least the reference value of the system operation rate indicating business processing information related to the business processing of the information system,
Analyzing the business process information, performing business process information analysis to obtain business process analysis information as an analysis result,
Analyzing the failure information for each failure mode, performing failure information analysis to obtain failure analysis information as an analysis result,
Creation of a diagram showing a hierarchical logical relationship from a top event indicating a failure mode of the information system to a basic event indicating a failure mode of the component or group of components based on the system configuration information And
The operating rate of the basic event is calculated based on the failure analysis information and the business process analysis information, and the operating rate of the top event is calculated based on the failure analysis information, the business process analysis information, and the diagram. Calculate the operating rate calculated as
Performing an operation rate determination to determine whether the calculated system operation rate has reached the reference value;
When it is determined that the calculated system operating rate has not reached the reference value, the system is based on the relationship between the operating rate corresponding to the basic event and the system operating rate corresponding to the top event. Perform basic event extraction to extract the basic events related to the increase in availability,
Performing a change possibility determination to determine whether or not the unavailability corresponding to the extracted basic event can be reduced;
If it is determined that the downtime corresponding to the extracted basic event can be reduced, the failure analysis information corresponding to the extracted basic event is reset in the uptime calculating unit. First resetting that resets the new failure information and the new system configuration information in the information input unit when it is determined that the unavailability corresponding to the extracted basic event cannot be reduced And
When new failure analysis information corresponding to the extracted basic event is reset, processing after the operation rate calculation is performed, and new failure information and new system configuration information are reset A reliability evaluation program for an information system, which causes a computer to execute a function for performing processing after the failure information analysis.

Images