WO2015159549A1 - 可用性分析装置、可用性分析方法、及び、可用性分析プログラムが記録された記録媒体 - Google Patents
可用性分析装置、可用性分析方法、及び、可用性分析プログラムが記録された記録媒体 Download PDFInfo
- Publication number
- WO2015159549A1 WO2015159549A1 PCT/JP2015/002092 JP2015002092W WO2015159549A1 WO 2015159549 A1 WO2015159549 A1 WO 2015159549A1 JP 2015002092 W JP2015002092 W JP 2015002092W WO 2015159549 A1 WO2015159549 A1 WO 2015159549A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- state
- failure
- availability
- states
- transition
- Prior art date
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F11/00—Error detection; Error correction; Monitoring
- G06F11/22—Detection or location of defective computer hardware by testing during standby operation or during idle time, e.g. start-up testing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F11/00—Error detection; Error correction; Monitoring
- G06F11/008—Reliability or availability analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F11/00—Error detection; Error correction; Monitoring
- G06F11/07—Responding to the occurrence of a fault, e.g. fault tolerance
- G06F11/0703—Error or fault processing not based on redundancy, i.e. by taking additional measures to deal with the error or fault not making use of redundancy in operation, in hardware, or in data representation
- G06F11/0766—Error or fault reporting or storing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N7/00—Computing arrangements based on specific mathematical models
- G06N7/01—Probabilistic graphical models, e.g. probabilistic networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F11/00—Error detection; Error correction; Monitoring
- G06F11/07—Responding to the occurrence of a fault, e.g. fault tolerance
- G06F11/08—Error detection or correction by redundancy in data representation, e.g. by using checking codes
- G06F11/10—Adding special bits or symbols to the coded information, e.g. parity check, casting out 9's or 11's
- G06F11/1076—Parity data used in redundant arrays of independent storages, e.g. in RAID systems
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N5/00—Computing arrangements using knowledge-based models
- G06N5/01—Dynamic search techniques; Heuristics; Dynamic trees; Branch-and-bound
Definitions
- the present invention relates to an availability analyzer that can analyze availability related to an information processing system and the like.
- Availability is one of the indexes for quantitatively evaluating the reliability (availability) of an IT (Information_Technology) system (hereinafter referred to as “target system”).
- the availability represents the probability that the target system is in a usable state when the state of the target system changes (changes) over time.
- the business operator operating the target system calculates the availability based on the configuration of the target system or information representing the state of the target system.
- the business operator quantitatively evaluates the reliability of the target system based on the calculated availability.
- the business operator searches for defects related to the target system based on the calculated availability.
- the business operator creates an improvement plan based on the calculated availability.
- the availability is calculated based on a state transition (State_Transition) model.
- a procedure for calculating availability based on a stochastic process such as a continuous-time Markov chain includes a procedure 1 and a procedure 2. That is, (Procedure 1) Expressing the state transition related to the target system as a model, (Procedure 2) The probability that the target system is in an available state is calculated by analyzing the stochastic process based on the model.
- Patent Document 1 discloses an apparatus that uses a Markov chain model as a technique for evaluating the availability of a complex target system. That is, the apparatus creates a Markov chain model for the target system using the failure rate and recovery rate for the components of the target system. Next, the apparatus evaluates the availability regarding the target system by analyzing the state transition represented by the created Markov chain model.
- Patent Document 2 discloses a method of expressing a target system as a model by combining a state transition model and a fault tree (Fault_Tree), and analyzing availability regarding the target system based on the model.
- Fault_Tree fault tree
- a model for analyzing availability results in a model related to a continuous-time Markov chain. That is, availability is calculated using a means for analyzing continuous time Markov chains.
- the number of states of the target system increases, the number of transitions between the states increases rapidly according to the combination of those states. For example, when the number of states of the target system is N (where N is a natural number), the matrix Q representing the transition between the states has N square elements. Therefore, a large amount of memory (storage device) is consumed by storing the matrix Q in the storage device.
- the availability evaluation method based on the state transition analysis has a problem that the analysis becomes difficult rapidly as the number of states of the target system increases.
- a main object of the present invention is to provide an availability analysis device and the like that can perform availability analysis even for a large scale target system.
- an availability analysis device includes: (I) component information indicating a transition rate between states of components included in the target system, and (II) a failure state indicating a state in which the target system cannot be operated among a plurality of states that the target system can take.
- Failure information including a condition indicating a state of the component in (III), and (III) recovery information including a transition rate when the target system transitions from the failure state to an operating state indicating the state in which the target system is operating And calculating a value between two states included in the plurality of states, calculating a probability that the target system is in a certain state based on the calculated value between the two states, and calculating the target Analysis for calculating availability related to the target system based on the probability when the system is in the operating state Equipped with a stage.
- the availability analysis method includes: (I) component information indicating a transition rate between states of components included in the target system, and (II) a failure state indicating a state in which the target system cannot be operated among a plurality of states that the target system can take.
- Failure information including a condition indicating a state of the component in (III), and (III) recovery information including a transition rate when the target system transitions from the failure state to an operating state indicating the state in which the target system is operating And calculating a value between two states included in the plurality of states, calculating a probability that the target system is in a certain state based on the calculated value between the two states, and calculating the target Based on the probability when the system is in the operating state, the availability regarding the target system is calculated.
- this object is also realized by such an availability analysis program and a computer-readable recording medium on which the program is recorded.
- availability can be analyzed even for a large target system.
- FIG. 2 is a block diagram illustrating an example of a configuration of a storage system that employs RAID.
- FIG. It is a figure which represents notionally an example of the continuous time Markov chain regarding a memory
- the state in which the target system is operating and the state indicating the failure of the target system transitions.
- an infinitesimal generator matrix (hereinafter referred to as “matrix”) Q.
- the continuous time Markov chain is Continuous_Time_Markov_Chain.
- the infinitesimal generator matrix is Infinitesimal_generator_matrix.
- Each row in matrix Q is associated with one state for the target system in a continuous time Markov chain.
- each column in matrix Q is associated with one state for the target system in a continuous time Markov chain.
- the transition rate (rate) of transition between two different states is expressed as a component related to the matrix Q. When the average transition time is T (where T> 0), the transition rate can be expressed as, for example, “1 ⁇ T”.
- the target system is represented using a first state to an Nth state (where N is a natural number).
- N is a natural number
- the I th row of the matrix Q and the I th column of the matrix Q represent the I th state
- the J th row of the matrix Q and the J th column of the matrix Q represent the J th state.
- the matrix Q is a square matrix, and I is 1 ⁇ I ⁇ N.
- J is 1 ⁇ J ⁇ N.
- the element in the I-th row and the J-th column of the matrix Q represents the transition rate for transition from the I-state to the J-th state.
- the elements in the I-th row and the I-th column of the matrix Q are values calculated according to the definition of the continuous-time Markov chain.
- the state of the target system is associated with a state identifier that can uniquely identify the state. Further, when the target system has a plurality of components, it is assumed that the state of the target system is associated with a combination of states related to the component.
- the target system is composed of a plurality of components (elements).
- the component is an element (component) included in the target system.
- the target system is an information processing apparatus, the component represents, for example, a memory, a hard disk, or the like.
- the target system is a factory, the component represents, for example, a machine, a communication device, or the like in the factory.
- the state in which the component is operating may be referred to as “component operating state”, and the state in which the component has a failure may be referred to as “component failure state”.
- a state related to a component may be expressed as a “component state”.
- a state in which the target system is operating may be referred to as a “system operating state”
- a state in which the target system has a failure and cannot be operated may be referred to as a “system fault state”.
- the state related to the target system may be expressed as “system state”.
- an element in the I-th row and the J-th column of the matrix Q is represented as an (I, J) element.
- the (I, J) element of the matrix Q is represented as Q (I, J).
- Q (I, I) ⁇ ( ⁇ (J ⁇ I) Q (I, J)) (Formula 1).
- a continuous time Markov chain can be analyzed.
- a probability vector ⁇ (numerical string ⁇ ) representing a steady state after a sufficiently long time can be obtained as a solution to the equation shown in Equation 2.
- the availability in the steady state related to the target system is ⁇ 1 .
- FIG. 1 is a block diagram showing the configuration of the availability analysis apparatus 101 according to the first embodiment of the present invention.
- the availability analysis apparatus 101 includes a calculation unit 102 and an analysis unit 103.
- the availability analysis apparatus 101 may further include an input unit 104.
- FIG. 4 is a flowchart showing the flow of processing in the input unit 104.
- the input unit 104 receives component information related to a plurality of components included in the target system that is the target for evaluating the availability 503 (step S201).
- the component represents a component included in the target system.
- the component represents a storage device included in the storage system, a control device that controls the storage device, and the like.
- the target system is software
- the component represents a function, a module, or the like included in the software.
- the component information includes information related to state transitions defined in advance according to the type of the component.
- FIG. 5 is a diagram conceptually illustrating an example of component information.
- the component information may include information regarding a plurality of components.
- ⁇ c represents a transition rate at which the component transitions from the component operating state to the component failure state. That is, ⁇ c represents a transition rate (failure rate) at which the component transitions from the component operating state to the component failure state. Further, mu c represents a transition rate component transitions the component operating state from component failure state (recovery rate).
- the component information includes information such that the first component state relating to the component represents the component operating state and the second component state relating to the component represents the component failure state.
- the component information includes a transition rate of transition from the first component state to the second component state with respect to the component.
- the component information includes, for example, information regarding a transition rate at which the second component state transitions to the first component state.
- the input unit 104 may generate a state transition model related to the target system based on the received component information, and store the state transition model in a storage unit (not shown) (step S202).
- a state related to the target system is represented using nodes, and a transition from the first state to the second state connects the node representing the first state and the node representing the second state. It is expressed using Moreover, the transition rate showing the ease of the transition between a 1st state and a 2nd state may be attached
- the state transition model is conceptually expressed using a graph.
- the input unit 104 receives operation information including one or more operation conditions representing a condition in which the target system is in the system operation state, and stores the operation information in a storage unit (not shown) (step S203).
- the operating condition is expressed using a component state related to a component included in the target system.
- the operating condition is represented by, for example, combining state identifiers representing component states.
- the operation information includes one or more operation conditions.
- the component operating state is represented as 0, and the component failure state is represented as 1.
- the operating condition is expressed as a logical sum of component states related to one or more components. This indicates that the target system is in the system operating state when all the components included in the target system are in the component operating state. In addition, when any one of the components is in a component failure state, the value of the operating condition is 1. In this case, the target system represents a system failure state.
- the operating condition may be whether or not the number of components in a specific component state is less than a predetermined value K.
- the operation condition represents a condition “the target system is in the system operation state when (M ⁇ K) or more components are in the component operation state”.
- M is an integer of 1 or more that represents the number of components that the target system has. Further, 0 ⁇ K ⁇ M.
- the input unit 104 receives failure information including one or more failure conditions indicating a condition in which the target system is in a system failure state, and stores the failure information in a storage unit (not shown) (step S204).
- the failure condition is expressed using a component state relating to a component included in the target system.
- the failure condition is represented by combining state identifiers representing component failure states (hereinafter also referred to as “third state identifiers” for convenience of explanation).
- the failure information includes one or more failure conditions.
- a failure condition is represented as a logical product of component states for one or more components. This indicates that the target system is in a system fault state when all components included in the target system are in a component fault state.
- the failure condition may be whether or not the number of components in a specific component state is a predetermined value K or more.
- the failure condition represents a condition “when the target system is in a system failure state when K or more components are in a component failure state”.
- system state after recovery is the system operating state.
- system state after recovery does not necessarily need to be the system operating state or the system operating state before the transition to the system failure state. The same applies to the following embodiments.
- the input unit 104 receives the recovery information related to the target system, and stores the received recovery information in a storage unit (not shown) (step S205).
- the recovery information the failure condition, the system operating state related to the target system after recovery from the system failure state when the failure condition is satisfied, and the ease of transition when transitioning from the system failure state to the system operating state Is associated with a transition rate representing the length.
- the failure condition included in the recovery information may be a state identifier associated with the failure condition.
- the recovery rate represents a transition rate at which a transition from the system failure state to the system operating state is made.
- the failure condition is represented using a state identifier representing the system failure state.
- the state identifier (that is, the third state identifier) represented by the failure condition, the system operating state, and the transition rate may be associated with each other.
- the third state identifier, the state identifier associated with the system operating state (hereinafter also referred to as “fourth state identifier” for convenience of description), and the transition rate may be associated. Good.
- a state (0, 0) indicating a system operation state recovered from a system failure state when the failure condition A is satisfied, and a transition rate when transitioning from the system failure state to the system operation state are associated with each other.
- the failure condition A is a condition that indicates whether both the component 1 and the component 2 are in the component failure state.
- the failure condition A is whether or not the system state is the state (1, 1).
- the system state satisfies the failure condition A.
- the target system is in a system failure state.
- the system state (1, 0) indicates that the component 1 is in a component failure state and the component 2 is in a component operating state. Therefore, the system state (1, 0) does not satisfy the condition A. For this reason, the target system is not in a system failure state.
- FIG. 2 is a flowchart showing the flow of processing in the availability analysis apparatus 101 according to the first embodiment. This example is an example of a continuous time Markov chain.
- a numerical sequence to be updated is represented as a numerical sequence (vector) ⁇ (k) .
- the calculation unit 102 determines the transition rate (ie, Q (I, J)) when the system state transitions from the I-th system state to the J-th (where 1 ⁇ J ⁇ N) system state, and Q (I Assume that I, I) is calculated.
- the calculation unit 102 does not necessarily need to calculate the transition rate itself, and may be a value calculated based on the transition rate.
- the analysis unit 103 calculates a numerical sequence ⁇ (1) in the first process.
- the numerical sequence ⁇ (1) may be a numerical sequence in which only one element is 1 and the other elements are 0. Further, the numerical sequence ⁇ (1) may be a numerical sequence calculated according to a specific procedure.
- the analysis unit 103 in the processing of the k-th calculates, for calculating the numerical sequence [pi a (k + 1).
- the analysis unit 103 does not update ⁇ i (k) when q ii is 0.
- the analysis unit 103 refers to q ij and q ii in Equation 3. For example, when referring to q ij , the analysis unit 103 calculates i (state identifier, expressed as “first state identifier”) and j (state identifier, expressed as “second state identifier”). 102.
- the calculation unit 102 receives the first state identifier and the second state identifier. Next, the calculation unit 102 determines a value in the case of transition from the I system state represented by the received first state identifier to the J system state represented by the second state identifier, or Q (I, I) according to Equation 1. Is calculated (step S101). The calculation unit 102 transmits the calculated value to the analysis unit 103.
- the analysis unit 103 receives the value calculated by the calculation unit 102, updates the numerical sequence ⁇ (k) according to Equation 3 with the received value as q ij or q ii (step S102).
- the analysis unit 103 transmits i (ie, the first state identifier) and i (ie, the second state identifier) to the calculation unit 102. Similar to the above-described processing, the analysis unit 103 receives the value calculated by the calculation unit 102 according to Equation 1, sets the received value as q ii , and converts the numerical sequence ⁇ (k) into the numerical sequence ⁇ (k + 1 ) according to Equation 3. ) .
- the analysis unit 103 determines that the numerical sequence ⁇ (k) when the difference between the numerical sequence ⁇ (k) and the numerical sequence ⁇ (k + 1) is smaller than the predetermined value ⁇ (that is, the inequality shown in Expression 4 ). The process of updating is terminated.
- the analysis unit 103 calculates a numerical sequence ⁇ (k + 1) .
- the analysis unit 103 calculates availability based on the calculated numerical sequence ⁇ (k + 1) .
- the analysis unit 103 calculates the availability of the target system by calculating the sum of ⁇ I (k + 1) with respect to the I system state representing the system operating state regarding the target system.
- FIG. 3 is a flowchart illustrating a processing flow in the calculation unit 102 according to the first embodiment.
- the calculation unit 102 receives the first state identifier and the second state identifier. Next, the calculation unit 102 determines whether or not the I system state represented by the first state identifier is a system failure state (step S103). For example, the calculation unit 102 executes the determination process shown in step S103 based on whether or not the failure information 501 includes the first state identifier. That is, as described above, since the failure condition is expressed using the state identifier associated with the system failure state, the calculation unit 102 calculates the state identifier associated with the failure state and the first state identifier. Compare.
- calculation unit 102 represents the system operating state associated with the first state identifier from recovery information 502. Read the state identifier and transition rate.
- the system operating state may be a state identifier associated with the operating state.
- the calculation unit 102 determines whether or not the read state identifier indicating the system operating state matches the second state identifier (step S104).
- the calculation unit 102 executes the process shown in step S104 for each system operating state.
- calculation unit 102 transmits a value calculated based on the read transition rate to analysis unit 103. (Step S105).
- calculation unit 102 determines whether or not the first state identifier and the second state identifier match. Determination is made (step S109). When the first state identifier and the second state identifier do not match, the calculation unit 102 calculates 0 as a value, and transmits the calculated 0 to the analysis unit 103 (step S106). When the first state identifier and the second state identifier match, the calculation unit 102 calculates a recovery rate ⁇ ( ⁇ 1) (that is, a value obtained by adding a minus to the recovery rate) as a value, and the calculated value Is transmitted to the analysis unit 103 (step S108). In this case, the recovery rate represents a transition rate at which the system failure state represented by the first state identifier transitions to a state recovered with respect to the system failure state.
- the calculation unit 102 reads a state identifier adjacent to the first state identifier in the state transition model. Being adjacent to a certain state identifier represents a system state that can be shifted directly from the first system state represented by the certain state identifier without passing through a different system state in the state transition model. In this case, the calculation unit 102 performs transition from the I system state represented by the first state identifier to the J system state represented by the second state identifier according to a predetermined calculation procedure (method) based on the component information. A transition rate is calculated (step S107).
- the predetermined calculation procedure is a procedure for calculating the Kronecker sum regarding the state transition model representing the component.
- the predetermined calculation procedure is based on the fact that the generator matrix that represents the transition relating to the system state of the target system including components that are processed independently of each other is the Kronecker sum relating to the generator matrix Q k that represents the transition relating to the component state relating to each component .
- the procedure for calculating the Kronecker sum will be described later.
- calculation unit 102 calculates the value based on the first state identifier and the second state identifier
- calculation unit 102 calculates a value for each second state identifier based on the first state identifier and the plurality of second state identifiers. May be.
- availability can be analyzed even for a large target system. This is because it is not necessary to store a matrix representing a transition from the first system state to the second system state.
- the analysis unit 103 when calculating the availability, the analysis unit 103 requests a value necessary for calculation from the calculation unit 102 and refers to the value calculated by the calculation unit 102. As a result, the availability analyzer 101 does not need to store the value. This is because the calculation unit 102 can calculate the value based on the component information, the failure information, and the recovery information.
- the availability analysis apparatus 101 does not store the matrix in the storage unit. Therefore, the availability analysis apparatus 101 can calculate the availability related to the target system even when the target system includes N or more system states.
- the number of system states of the target system is determined according to the number of components that the target system has and the number of component states of the components. Therefore, according to the availability analysis apparatus 101, even when the number of components increases, it is not necessary to store all the elements of the matrix, so that the availability can be analyzed.
- FIG. 6 is a block diagram showing the configuration of the availability analyzer 111 according to the second embodiment of the present invention.
- FIG. 7 is a flowchart showing the flow of processing in the availability analyzer 111 according to the second embodiment.
- the availability analyzer 111 includes a calculation unit 113 and an analysis unit 103.
- the availability analyzer 111 may further include an input unit 112 and a creation unit 114.
- the calculation unit 113 determines whether or not the non-reachable information includes any of the received state identifiers (step S111).
- Non-reachable information is a state where one or more components have failed from the system failure state (because it is not necessary to consider the reachability for the purpose of availability analysis, it will be referred to as “non-reachable state” hereinafter. ) Is associated with the status identifier.
- the calculation unit 113 may determine whether or not the reachable information includes any of the received state identifiers.
- the reachable information includes a state identifier associated with a system state that is not in a non-reachable state (hereinafter, referred to as “reachable state”).
- the non-reachable state is a failure state in which it is impossible to make a transition from the system operating state to the next.
- the reachable state represents a system state that is not a non-reachable state.
- FIG. 8 is a flowchart illustrating an example of a flow of processing for creating reachable information and the like.
- the availability analyzer 111 receives reachable information or non-reachable information.
- the availability analysis device 111 may include a creation unit 114 that creates reachable information or non-reachable information according to the processing illustrated in FIG.
- the creation unit 114 creates a system state set ⁇ of the target system based on the component state regarding each component of the target system (step S211).
- the creation unit 114 creates the system state of the target system by combining the component states for each component.
- the target system has a component A and a component B.
- the states regarding the component A are assumed to be a component state U a and a component state F a .
- the states regarding the component B are assumed to be a component state U b and a component state F b .
- component state F a will represent the component fault condition component A.
- the component state F b represents a component fault state related to the component B.
- the component state U a is assumed to represent a component operating state related to the component A.
- the component state U b represents a component operating state related to the component B.
- the creation unit 114 creates a set ⁇ of system states related to the target system as shown in Expression 5 by combining the component states related to each component (step S211).
- (U a , U b ), (U a , F b ), (F a , U b ), or (F a , F b ) is an example of a system state.
- the target system is assumed to be in a system fault state.
- the system failure states related to the target system are the system failure state (U a , F b ), the system failure state (F a , U b ), and the system failure state (F a , F b). ).
- the target system when the component B is in a component failure state, the target system is in a system failure state (U a , F b ).
- the target system loses its inherent function in response to a system failure state (falls).
- the target system is subjected to recovery processing according to the recovery procedure.
- the state of the target system does not transit from the system state (U a , U b ) to the system state (F a , F b ) without going through one or more system fault states.
- the non-reachable information is configured using a state identifier representing a system state (F a , F b ).
- the non-reachable information includes a state identifier representing a system failure state that can be transited through one or more system failure states.
- the reachable information is configured using a system identifier (U a , U b ), a system state (U a , F b ), and a state identifier representing the system state (F a , U b ).
- the system fault state related to the target system is a case where three or more types of components are component fault states.
- the non-reachable state regarding the system state is a case where four or more types of components are in a component failure state.
- the creation unit 114 determines whether each element satisfies the failure condition related to the target system by applying the failure condition included in the failure information 501 to the element included in the set ⁇ (step S212).
- the creation unit 114 collects elements (represented as “second elements”) having different component states (represented as “second elements”) constituting the system failure state by a set ⁇ with respect to the elements that are in a system failure state (denoted as “first elements”) Extract from
- the creation unit 114 checks whether the second element satisfies the failure condition.
- the creation unit 114 adds the first element to the non-reachable information when all the extracted second elements satisfy the failure condition (step S213). If there is an element that does not satisfy the failure condition among the extracted second elements, the creation unit 114 adds the first element to the reachable information.
- the creation unit 114 adds the state identifier included in the operation information to the reachable information.
- the input unit 112 receives reachability information about the target system from the outside or the creation unit 114, and stores the reachability information in a storage unit (not shown).
- step S111 The processing after step S111 will be described with reference to FIG.
- calculation unit 113 sets the value to 0 (step S113). Further, when non-reachable information does not include the received state identifier (YES in step S111), calculation unit 113 calculates a value according to the processing shown in steps S103 to S107 shown in FIG. S112).
- the calculation time can be further shortened.
- the reason is Reason 1 and Reason 2. That is, (Reason 1)
- the configuration of the availability analyzer 111 according to the second embodiment includes the configuration of the availability analyzer 101 according to the first embodiment. (Reason 2) This is because processing related to the non-reachable state is reduced.
- the calculation unit 113 first determines whether or not the system state represented by the first state identifier or the second state identifier represents a non-reachable state. The value is 0.
- the calculation unit 113 executes processing related to step S112. Therefore, compared with the availability analysis apparatus 101 according to the first embodiment, processing related to step S112 is reduced. As a result, according to the availability analyzer 111 according to the present embodiment, the calculation time can be further shortened.
- FIG. 9 is a block diagram showing the configuration of the availability analyzer 123 according to the third embodiment of the present invention.
- FIG. 10 is a flowchart showing the flow of processing in the availability analyzer 123 according to the third embodiment.
- the availability analysis device 123 includes a calculation unit 113, an analysis unit 124, a determination unit 121, and a transition information creation unit 122.
- the determination unit 121 determines whether or not the number of state identifiers representing the reachable state included in the reachable information (hereinafter referred to as “reachable state number”) is less than a predetermined number (step S121). .
- transition information creating unit 122 determines the reachable value based on the value calculated by calculating unit 113. Transition information representing a transition state between states is created (step S122). For example, the transition information creation unit 122 transmits a state identifier representing a reachable state to the calculation unit 113. The calculation unit 113 receives the state identifier, calculates a value related to the received state identifier, and transmits the calculated value to the transition information creation unit 122. The transition information creation unit 122 receives the value and stores the received value in the transition information. Transition information can be expressed using the infinitesimal generator matrix described above.
- the analysis unit 124 calculates availability based on the transition information (step S123).
- the transition information created by the transition information creation unit 122 is equivalent to an infinitesimal generator matrix related to the reachable state in the target system.
- analysis unit 124 follows the processing shown in steps S101 and S102 in FIG. Is calculated (step S124).
- the availability analysis device 123 in addition to the effects of the availability analysis device 111 according to the second embodiment, the availability can be calculated at a higher speed.
- the reason is Reason 1 and Reason 2. That is, (Reason 1)
- the configuration of the availability analyzer 123 according to the third embodiment includes the configuration of the availability analyzer 111 according to the second embodiment. (Reason 2) By creating transition information, it is not necessary to repeatedly calculate a transition rate or the like when transitioning from the I system state to the J system state.
- the availability analyzer 123 creates transition information when the number of reachable states is less than a predetermined number. With this process, the availability analyzer 123 creates a situation in which the storage area for storing the transition information is limited and a situation in which the process for repeatedly calculating the transition rate and the like is avoided.
- FIG. 11 is a block diagram showing a configuration of the availability analysis apparatus 133 according to the fourth embodiment of the present invention.
- the availability analysis device 133 includes a calculation unit 113, an analysis unit 124, a determination unit 131, and a transition information creation unit 132.
- the determination unit 131 determines whether the reachable state number included in the reachable information is less than a predetermined number.
- the transition information creation unit 132 creates transition information that represents the state of transitions between reachable states. However, the transition information creation unit 132 processes the system failure state related to the target system as one system failure state. For example, as illustrated in the above-described example, when the target system includes the component A and the component B, the transition information creation unit 132 sets the system state (U a , F b ) and the system state (F a , U b ) are treated as one system failure state.
- the system state (U a , F b ) and the system state (F a , U b ) represent a system failure state related to the target system.
- the transition information creation unit 132 assigns one system state called F s to two system states called (U a , F b ) and (F a , U b ), for example.
- the transition information creation unit 132 further assigns a system state U s to the system operating state (U a , U b ) related to the target system.
- the transition information creation unit 132 since the system state (F a , F b ) is a non-reachable state, the transition information creation unit 132 does not assign a system state to (F a , F b ). That is, the transition information creation unit 132 processes two system states, U s and F s , as the system state of the target system.
- the transition information creation unit 132 is, for example, a value calculated by the calculation unit 113 regarding a transition from the system state (U a , F b ) to a certain state and a calculation unit 113 regarding a transition regarding the system state (F a , U b ) With respect to the values calculated by, an operation described later is applied to the two values.
- the transition information creation unit 132 executes processing with the two system states (U a , F b ) and (F a , U b ) as one system state F s .
- the transition information creation unit 132 creates the matrix Q based on the calculated result, similarly to the transition information creation unit 122 according to the third embodiment.
- FIG. 12 is a block diagram illustrating an example of a configuration of an information system including a storage system 522 that employs RAID.
- the availability analyzer 133 calculates the availability related to the storage system 522 having a plurality of storage devices.
- the storage device is a magnetic disk, a nonvolatile semiconductor memory, or the like.
- the mode of the storage device is not limited to the above example.
- RAID technology is one technology that improves the reliability and performance of storage systems.
- Availability related to a storage system that employs RAID technology includes reliability related to storage devices possessed by RAID, efficiency related to data recovery processing when the storage device is in a failure state, and efficiency related to recovery processing when data is lost. Depends on etc.
- the availability related to the storage system further depends on the RAID level that defines the mode of storing data.
- the storage system calculates a parity regarding the data when storing the data in the storage device.
- the storage system stores the data and the calculated parity in the storage device.
- the storage device in the component failure state is replaced with a new storage device.
- the storage system recovers data stored in the storage device in which the failure has occurred based on the calculated parity and data stored in another storage device, and stores the recovered data in a new storage device.
- a storage system that employs RAID level 5 cannot recover data stored in a storage device having a failure based on parity when two storage devices of the storage device have a failure. In this case, the storage system is reconstructed based on the backup data or the like. The user cannot use the storage system during the period of rebuilding the storage system.
- the storage system 522 includes a RAID (RAID level 5) controller 524, a storage device 525, a storage device 526, and a storage device 527.
- the backup system 523 includes a storage device 528.
- the host computer 521 can communicate with the storage system 522 and the backup system 523.
- the backup system 523 stores the data stored in the RAID-configured storage apparatus configured by the RAID controller 524 in the storage apparatus 528.
- a user who uses the storage system 522 reads and writes data stored in the storage device via the host computer 521. Further, the host computer 521 periodically backs up the data to the backup system 523 in preparation for the loss of data in the storage system 522, for example.
- the host computer 521 analyzes the probability (availability) that the data stored in the storage system 522 can be accessed. That is, it is assumed that the availability analysis device 133 is included in the host computer 521.
- the user inputs operation information regarding the storage system 522, information regarding each component, and the like to the input unit 104 (FIG. 1).
- the input unit 104 generates a state transition model based on components (for example, the storage device 525 to the storage device 527) included in the storage system 522.
- the RAID controller 524 is represented using a continuous-time Markov chain including two states of a component operating state and a component failure state, as illustrated in FIG.
- the failure rate related to the RAID controller 524 is ⁇ c
- the recovery rate related to the RAID controller 524 is ⁇ c
- each of the storage device 525, the storage device 526, and the storage device 527 is represented using a continuous-time Markov chain including two states of a component operating state and a component failure state, as illustrated in FIG.
- FIG. 13 is a diagram conceptually illustrating an example of a continuous time Markov chain related to a storage device.
- the failure rate related to the storage device is ⁇ d
- the recovery rate related to the storage device is ⁇ d .
- component states related to the RAID controller 524, the storage device 525, the storage device 526, and the storage device 527 are represented as x 1 , x 2 , x 3 , and x 4 , respectively.
- the set ⁇ representing the system state relating to the storage system 522 can be represented using a system state (x 1 , x 2 , x 3 , x 4 ) that is a combination of component states relating to each component.
- the input unit 104 receives, for example, the operating condition A shown in Equation 6 as the operating information related to the storage system 522.
- the operation information is not necessarily the logical expression shown in Expression 6.
- the operating condition A represents an operating condition related to the storage system 522, and is 0 when the storage system 522 is in an operating state.
- the system failure state related to the storage system 522 is when the RAID controller 524 is in a component failure state (Equation 7) or when two of the three storage devices are in a component failure state (Equation 8). ).
- the input unit 104 receives Expression 7 and Expression 8 as the failure information 501 regarding the storage system 522.
- Failure condition FC x 1 (Expression 7)
- Failure condition FS x 2 ⁇ x 3 ⁇ x 2 ⁇ x 4 ⁇ x 3 ⁇ x 4 (Expression 8).
- the value of either the failure condition FC or the failure condition FS is 1.
- a recovery rate when recovering from a component fault condition RAID controller 524 to the components operating state and a C representing a recovery rate when recovering from a component fault condition RAID controller 524 to the components operating state and a C.
- a recovery rate when the storage system 522 is reconstructed by restoring data from the backup system 523 when two of the three storage devices are in a failure state is denoted as a S.
- the input unit 104 receives Expressions 9 and 10 as the recovery information 502 related to the storage system 522.
- the input unit 104 may create the recovery information 502.
- the analysis unit 124 generates a numerical sequence ⁇ (1) .
- the numerical sequence ⁇ (1) includes 16 numerical values.
- the numerical analysis method in the analysis unit 124 is, for example, the Jacobian method shown in the first embodiment.
- numeric column [pi (k) is updated to the numerical sequence ⁇ (k + 1), when the difference between the numerical sequence [pi (k) a numeric string ⁇ (k + 1) is sufficiently small, numerical sequence [pi The process of updating (k) is terminated.
- the analysis unit 124 calculates a part of q ij (for example, the matrix Q illustrated in FIG. 14A and FIG. 14B according to the process illustrated in each embodiment of the present invention. Reference only the value of q ij ) for the reachable state.
- 14A and 14B are diagrams illustrating an example of a general matrix Q, which are divided into two drawings due to the illustrated constraints.
- the i-th and j-th column component q ij included in the matrix Q represents a transition rate at which the i-th system state transitions to the j-th system state.
- q ii represents a value obtained by multiplying the sum of transition rates from the i-th system state to a different system state by “ ⁇ 1”.
- q ij is calculated by a calculation unit (for example, the calculation unit 102 and the calculation unit 113) according to a series of processes shown in the flowchart of FIG.
- step S107 in FIG. 3 represents the process which the calculation part 113 calculates based on the Kronecker sum shown to Formula 13 mentioned later.
- q ij and q ii are 0.
- the analysis unit 124 may transmit the values of i and j to the calculation unit 113.
- the calculation unit 113 calculates the value of q ij, and transmits the calculated q ij to the analysis unit 124.
- Analysis unit 124 receives the q ij, based on the received q ij, updates numerical sequence ⁇ a (k).
- the index I of the matrix Q can be obtained, for example, by applying the function illustrated in Expression 11 to the system state (x 1 , x 2 , x 3 , x 4 ) regarding the storage system 522.
- the function may be a function that associates the system state related to the storage system 522 and the value of the index I of the matrix Q so as to correspond one-to-one.
- the value “5” is calculated by applying Expression 11 to the system state (0, 1, 0, 0).
- the system state (0, 1, 0, 0) is associated with the fifth system state, namely the fifth row in the matrix Q and the fifth column in the matrix Q.
- q 5j (where j is an integer) represents a transition rate when transitioning from the fifth system state to the j-th system state.
- q i5 (where i is an integer) represents a transition rate when transitioning from the i-th system state to the fifth system state.
- 14A and 14B include a row whose values are all 0 or a column whose values are all 0. This row and column indicate that the system state corresponding to the index is a non-reachable state.
- the determination unit 131 may calculate the number of reachable states by calculating the non-reachable state according to the failure condition FS, the component states x 1 , x 2 , x 3 , x 4 and Equation 12.
- Expression 12 is 1 when the system state (x 1 , x 2 , x 3 , x 4 ) regarding the storage system 522 is in a non-reachable state.
- two or more storage devices are in a component failure state.
- the system state (1, 1, 1, 0) represents a state in which the RAID controller 524, the storage device 525, and the storage device 526 are in a component failure state.
- the storage system 522 is in a system failure state when the RAID controller 524 is in a component failure state or when two of the three storage devices are in a component failure state. So stop functioning. Therefore, the storage system 522 does not enter the system state (1, 1, 1, 0). In this case, the system state (1, 1, 1, 0) is a non-reachable state.
- the calculation unit 113 calculates 0 as a value when the system state represented by the first state identifier or the system state represented by the second state identifier is a non-reachable state. This corresponds to a row in which all the values are 0 or a column in which all the values are 0 in FIGS. 14A and 14B.
- the transition information creation unit 132 generates the matrix Q by not storing the rows whose values are all 0 and the columns whose values are all 0 as the matrix Q.
- the calculation unit 113 indicates that the system state represented by the first state identifier is a system failure state. It is determined whether or not. For example, in this example, the calculation unit 113 determines whether the storage system 522 is in a system failure state according to Equation 7 and Equation 8.
- the calculation unit 113 calculates a value based on the recovery information 502 when the state represented by the first state identifier is a system failure state. For example, when the system state represented by the first state identifier is a system failure state according to Equation 7 (that is, the failure condition FC), the calculation unit 113 transitions from the recovery information 502 to the transition rate a C associated with the failure condition FC. Read. The calculation unit 113 calculates “ ⁇ a C ” when the first state identifier and the second state identifier match, and calculates the value a C when the first state identifier and the second state identifier do not match. And This process is based on the definition for the matrix Q.
- the calculation unit 113 calculates element values in the matrix Q, for example, according to a procedure for calculating a Kronecker sum disclosed in Non-Patent Document 1 and the like.
- the procedure for calculating the Kronecker sum disclosed in Non-Patent Document 1 and the like is as follows. For a target system including components that operate independently from each other, a generator matrix that represents a state transition represents a Kronecker related to a generator matrix that represents a state transition for each component. Based on being expressed by sum.
- the calculation unit 113 calculates the value of q ij based on the definition related to the Kronecker sum shown in Expression 13 and the matrix elements related to components.
- the calculation unit 113 can calculate a value related to the matrix Q in accordance with the processing described above.
- the analysis unit 124 calculates the availability related to the storage system 522 by calculating the sum related to the system operating state based on the calculated numerical sequence ⁇ (k + 1) (that is, the probability related to the steady state).
- transition information creation unit 132 Next, processing in the transition information creation unit 132 will be described using the above-described example.
- the reachable state is 11 system states among the 16 system states corresponding to 16 rows representing the matrix Q.
- the transition information creating unit 132 creates a matrix R related to reachable states as shown in FIG.
- FIG. 15 is a diagram conceptually illustrating an example of a matrix related to the reachable state. Note that the matrix R illustrated in FIG. 15 represents a matrix including rows corresponding to the reachable state and columns corresponding to the reachable state among the elements of the matrix Q illustrated in FIGS. 14A and 14B.
- the size of the matrix Q is (number of reachable states ⁇ number of reachable states), and is (predetermined number ⁇ predetermined number) at most. If (predetermined number ⁇ predetermined number) is smaller than the capacity of the storage device, the storage device can store the matrix Q.
- the transition information creation unit 132 creates the matrix Q when the storage device can store the matrix Q, and stores the created matrix Q in the storage device.
- the analysis unit 124 may update the numerical sequence ⁇ (k) with reference to the matrix Q in the storage device, for example. Therefore, in the process in which the analysis unit 124 updates the numerical sequence ⁇ (k) , the calculation unit 113 does not need to repeatedly calculate the elements included in the matrix Q.
- the transition information creation unit 132 processes a plurality of system failure states as one system failure state, for example. Even if the system failure states are different from each other in the recovery information 502, when the system failure states are associated with the common system operation state and the common transition rate, the transition information creation unit 132 , The system failure states are collectively processed. This process is performed on the row representing the system fault condition and the column representing the system fault condition in the matrix R.
- a process related to a row representing a system failure state will be described.
- a procedure for calculating the transition rate will be described with reference to the information shown in Expression 10 in the recovery information 502 as an example.
- a system failure state that transitions to a system state (0, 0, 0, 0) representing a state recovered from the system failure state at the transition rate a S is a failure condition FS included in Equation 10 (specifically, Equation 8 ) Is calculated as a system failure state. That is, the system fault state is a system fault state (0, 1, 1, 0), a system fault state (0, 0, 1, 1), and a system fault state (0, 1, 0, 1). .
- the transition information creation unit 132 processes the three system failure states together.
- the transition information creation unit 132 can create the matrix Q illustrated in FIG. 16 by processing the three system failure states together. That is, when creating the matrix Q, the transition information creation unit 132 calculates the sum of the values of the elements constituting the system failure states (in this case, the above three types of system failure states) that are collectively processed. .
- FIG. 16 is a diagram conceptually illustrating an example of a matrix generated when a system failure state to be processed is processed as one system failure state.
- the matrix before change illustrated in FIG. 15 is represented as “matrix R”
- the matrix after change illustrated in FIG. 16 is represented as “matrix Q”.
- the transition information creation unit 132 performs a matrix corresponding to the system failure state from the three types of system failure states according to the procedure for creating the matrix R from the system state described above with reference to Equation 11. Assume that an R index is calculated. For example, in the case of the matrix R illustrated in FIG. 15, the transition information creation unit 132 calculates a value “4” representing an index according to Equation 11 for the system failure state (0, 0, 1, 1). For example, in the case of the matrix R illustrated in FIG. 15, the transition information creation unit 132 calculates a value “6” representing an index according to Equation 11 for the system failure state (0, 1, 0, 1). For example, in the case of the matrix R illustrated in FIG.
- the transition information creation unit 132 calculates a value “7” representing an index according to Equation 11 for the system failure state (0, 1, 1, 0). That is, in the case of the matrix R illustrated in FIG. 15, the system failure state (0, 0, 1, 1) represents the system failure state shown in the fourth row. In the case of the matrix R illustrated in FIG. 15, the system fault state (0, 1, 0, 1) represents the system fault state shown in the sixth row. In the case of the matrix R illustrated in FIG. 15, the system failure state (0, 1, 1, 0) represents the system failure state shown in the seventh row. That is, the index represents the number of rows or columns of the matrix R. Further, in the case of the matrix Q illustrated in FIG. 16, the system failure state processed together in the one represents the system failure state shown in the fourth row.
- the system operating state shown in the first row of the matrix Q illustrated in FIG. 16 represents the system operating state shown in the first row of the matrix R illustrated in FIG.
- the system operating state shown in the second row of the matrix Q illustrated in FIG. 16 represents the system operating state shown in the second row of the matrix R illustrated in FIG.
- the system operating state shown in the third row of the matrix Q illustrated in FIG. 16 represents the system operating state shown in the third row of the matrix R illustrated in FIG.
- the system operating state illustrated in the fifth row of the matrix Q illustrated in FIG. 16 represents the system operating state illustrated in the fifth row of the matrix R illustrated in FIG.
- the elements of the matrix Q illustrated in FIG. 16 are related to one or more types of system failure states that are processed together as one of the elements of the matrix R illustrated in FIG. Can be calculated as a sum of values representing elements calculated for each system failure state. More specifically, the transition information creation unit 132 performs the following processing.
- the transition information creation unit 132 sets a failure condition FS (specifically, Expression 8) that is a specific transition rate a S and transitions to a specific system state in the recovery information 502 as a processing target. The process shown in is executed.
- the transition information creation unit 132 calculates at least one or more system failure states that satisfy the failure condition FS to be processed, and a matrix R corresponding to the calculated system failure state according to the calculation formula illustrated in Equation 11. Are calculated for each individual system failure state.
- the transition information creation unit 132 represents the specific transition rate a S as the value of the column associated with the system state recovered from the system failure state with respect to the row indicated by the calculated index indicating the system failure state. Calculate the value.
- the transition information creation unit 132 transitions from the system failure state to be processed together to the specific system state.
- the transition rate to be calculated is a S, and the transition rate to transition from the system failure state to a state different from the specific system state is calculated as 0.
- the transition information creation unit 132 calculates a value according to the above-described equation 1.
- a row and a column in which a plurality of rows and a plurality of columns indicating the system fault states to be processed together in the matrix R are associated with one row and column of the matrix Q.
- the number of rows and columns of the matrix Q is smaller than the number of rows and columns of the matrix R.
- the decrease number of rows forming the matrix Q and the decrease number of columns forming the matrix Q are: This is the number indicated by the number A. That is, Number A: “(number of system fault states constituting system fault states to be processed together) ⁇ 1” ⁇ 1.
- the number of reductions regarding all “system failure states to be processed together” is the “system failure state to be processed into one” for each row and column. Is the sum of the above-mentioned number A.
- the transition information creation unit 132 adds the transition rates in each column indicated by the index indicating the system failure state to one for the row indicated by the index indicating the system operating state among the indexes calculated according to the above-described processing. To calculate a transition rate representing the sum.
- a set of indexes of the matrix R corresponding to the index J of the matrix Q (that is, the J-th state) is represented as G (J).
- the matrix index set G (4) shown in FIG. 15 is the system failure state included in the system failure state to be combined into one. Is constituted by three elements ⁇ 4, 6, 7 ⁇ representing The three elements are values “4”, “6”, and “7” that represent the indexes obtained previously according to Equation 11.
- indexes calculated according to the equation 11 regarding the system operating state are the same in the matrix R illustrated in FIG. 15 and the matrix Q illustrated in FIG. That is, for index J representing the system operating state, G (J) is assumed to be composed of one element ⁇ J ⁇ .
- an index is not limited to the example mentioned above.
- the transition information creation unit 132 causes the system failure related to the Jth column to be processed together into one system failure.
- the transition rate is calculated according to Equation 14.
- the transition information creation unit 132 calculates a value according to the above-described equation 1.
- the number of columns of the matrix Q is calculated by executing the above-described processing. It becomes smaller than the number of columns of R.
- the number of indexes representing the system operating state in the matrix Q is the same as the number of indexes representing the system operating state in the matrix R
- the number of rows of the matrix Q with respect to the system operating state is Same as the number of lines.
- the number of reductions in the number of columns forming the matrix Q is the sum of the above-mentioned number A relating to each “system failure state to be processed together”.
- the number of rows of the matrix Q is the same as the number of rows of the matrix R.
- the transition information creation unit 132 performs the failure condition illustrated in Expression 7 with respect to the information shown in Expression 9 (the recovery information 502 including the failure condition FC). Based on the FC, a system failure state that satisfies the failure condition FC is calculated. Next, the transition information creation unit 132 obtains an index according to the equation 11 with respect to the calculated system failure state, and the system failure state indicated in the eighth, ninth, tenth, and eleventh rows of the matrix R represented by the obtained index is Treat as one system failure condition. A detailed description of the processing related to the failure condition FC is omitted.
- One index J representing the system operating state in the matrix Q is associated with one index representing the system operating state of the matrix R by the above-described index set G (J).
- one index J representing the system fault state in the matrix Q is associated with a plurality of indexes representing the system fault states to be combined into one by the above-described index set G (J).
- the matrix Q is a square matrix as described at the beginning of the “Description of Embodiments”.
- the relationship that the number of reductions in the number of columns in the matrix Q resulting from the processing relating to the system failure state is equal to the number of decreases in the number of rows in the matrix Q is maintained. That is, when comparing the matrix R and the matrix Q that is the result of the above-described processing, the number of reductions in the number of columns forming the matrix Q is equal to the number of reductions in the number of rows forming the matrix Q.
- the method for determining the number of columns is not limited to the method using the characteristics of the square matrix in the present embodiment.
- the transition information creation unit 132 converts the system failure states shown in the fourth, sixth, and seventh lines in FIG. 15 to one system failure state with respect to the information represented by Expression 10 (recovery information 502 including the failure condition FS). Process as.
- the matrix R illustrated in FIG. 15, in the second row, fourth column and the sixth column of values is lambda d.
- the element in the I-th row and the J-th column of the matrix R represents the transition rate from the I-state to the J-th state.
- the transition rate when transitioning from the system operating state shown to the system failure state shown in the fourth column is ⁇ d .
- the system operating state shown in the second row of the matrix R the transition rate when a transition to a system fault condition shown in the sixth row is a lambda d.
- the calculation unit 113 calculates ⁇ d as a value when the system operating state shown in the second row of the matrix R transitions to the system failure state shown in the fourth column. Further, the transition rate when the system operating state shown in the second row of the matrix R transitions to the system fault state shown in the seventh column is 0.
- the transition information creation unit 132 sets the system failure state shown in the fourth row of the matrix R, the system failure state shown in the sixth row of the matrix R, and the matrix R
- the system failure state shown in the seventh line is processed as a system failure state that is processed together.
- the transition information creation unit 132 determines the transition rate when transitioning from the system operating state shown in the second row of the matrix R to the system fault state that is processed together as one of the above three transition rates. Calculate as the sum.
- the transition information creation unit 132 receives values (in this case, 0 and two ⁇ d ) from the calculation unit 113 for the three transition rates corresponding to the three rows of interest described in the previous paragraph, The sum of the three values (in this case, ⁇ d + ⁇ d +0) is calculated. That is, in FIG. ⁇ Transition rate ⁇ d when transitioning from the system operating state shown in the second row to the system fault state shown in the fourth column, ⁇ Transition rate ⁇ d when transitioning from the system operating state shown in the second row to the system fault state shown in the sixth column, ⁇ Transition rate 0 when the system operating state shown in the second row transitions to the system failure state shown in the seventh column.
- the calculated value (2 ⁇ ⁇ d ) is represented by one value representing the system failure state of the target three rows, and is set in the fourth column of the second row of the matrix Q. This is because the second row of the matrix Q represents the system operating state indicated by the index set G (2), and the fourth row of the matrix Q represents the system failure state indicated by the index set G (4). is there.
- the transition information creation unit 132 relates to the system operating state shown in the third row of the matrix R illustrated in FIG. 15, the system failure state shown in the fourth row, the system failure state shown in the sixth row, and The system fault state shown in the seventh line is processed as a system fault state that is processed together. For this reason, the transition information creation unit 132 sets the transition rate in the case of transitioning from the system operating state shown in the third row of the matrix R illustrated in FIG. Is calculated as the sum of the three transition rates shown in FIG. That is, in FIG.
- the transition information creation unit 132 calculates the transition rate for transitioning from the system operating state shown in the third row in FIG. 15 to the system fault state to be processed as a unit as the sum of the three transition rates described above.
- the transition rate described above is a value calculated by the calculation unit 113.
- the calculated value (2 ⁇ ⁇ d ) is represented by one value representing the system failure state for the three rows of interest, and is set in the fourth column of the third row of the matrix Q. This is because the third row of the matrix Q represents the system operating state indicated by the index set G (3), and the fourth row of the matrix Q represents the system failure state indicated by the index set G (4). is there.
- the transition information creation unit 132 relates to the system operating state shown in the first row of the matrix R illustrated in FIG. 15, the system failure state shown in the fourth row, the system failure state shown in the sixth row, In addition, the system failure state shown in the seventh line is processed as a system failure state that is collectively processed. For this reason, the transition information creation unit 132 sets the transition rate when transitioning from the system operating state shown in the first row of the matrix R illustrated in FIG. Is calculated as the sum of the three transition rates shown in FIG. That is, in FIG.
- the transition information creation unit 132 determines the transition rate from the system operating state shown in the first row of the matrix Q illustrated in FIG. Calculated as the sum of rates.
- the transition rate described above is a value calculated by the calculation unit 113.
- the calculated value (0) is represented by one value indicating the system failure state of the target three rows, and is set in the fourth column of the first row of the matrix Q. This is because the first row of the matrix Q represents the system operating state indicated by the index set G (1), and the fourth row of the matrix Q represents the system failure state indicated by the index set G (4). is there.
- the system failure state related to the transition rate a c shown in the eighth to eleventh rows of the matrix R and the system operating state shown in the fifth row of the matrix R The process is also executed for. However, detailed description is omitted of the processing procedure performed the line as a target with respect to the transition rate a c.
- the transition information creating unit 132 described above changes the matrix R illustrated in FIG. 15 to the matrix Q illustrated in FIG.
- the analysis unit 124 updates the numerical sequence ⁇ (k) while referring to the matrix Q in the storage device. Therefore, in the process in which the analysis unit 124 updates the numerical sequence ⁇ (k) , the calculation unit 113 does not need to repeatedly calculate the elements included in the matrix Q.
- the availability analysis device 133 in addition to the effects of the availability analysis device 123 according to the third embodiment, the availability can be calculated for a large-scale target system.
- the reason is Reason 1 and Reason 2. That is, (Reason 1)
- the configuration of the availability analysis apparatus 133 according to the fourth embodiment includes the configuration of the availability analysis apparatus 123 according to the third embodiment. (Reason 2)
- the size of the matrix Q is further smaller than that of the availability analyzer 123 according to the third embodiment.
- FIG. 17 is a block diagram showing a configuration of the availability analysis apparatus 151 according to the fifth embodiment of the present invention.
- the availability analysis apparatus 151 includes an analysis unit 152.
- the analysis unit 152 calculates a value between two system states among a plurality of system states that can be taken by the target system. That is, (1) Component information representing a transition rate between component states of components included in the target system, (2) Fault information including a condition indicating a component status of a component in a system fault status indicating a system status in which the target system cannot operate among a plurality of system statuses that the target system can take; (3) Recovery information including a transition rate when the target system transitions from a system failure state to a system operating state representing a state in which the target system is operating.
- processing for calculating the value between the two states is the same as the processing in the calculation unit 102 shown in the first embodiment, the calculation unit 113 shown in the second, third, and fourth embodiments. It is.
- the analysis unit 152 calculates the probability that the target system is in a certain system state based on the calculated value between the two states.
- the analysis unit 152 calculates the availability related to the target system based on the probability when the target system is in the system operating state among the calculated transition rates. For example, the analysis unit 152 calculates availability by adding the probabilities when the target system is in the system operating state.
- the process for calculating the transition rate and the process for calculating the availability are shown in the analysis unit 103, the third embodiment, and the fourth embodiment shown in the first and second embodiments. This is the same processing as the processing in the analysis unit 124 or the like.
- availability can be analyzed even for a large target system. This is because it is not necessary to store all the elements of the matrix representing the transition from the first system state to the second system state.
- the availability analysis apparatus may be realized using at least two calculation processing apparatuses physically or functionally. Further, the availability analysis apparatus may be realized as a dedicated apparatus.
- FIG. 18 is a diagram schematically illustrating a hardware configuration of a calculation processing apparatus capable of realizing the availability analysis apparatus according to the first to fifth embodiments.
- the calculation processing device 20 includes a central processing unit (Central_Processing_Unit, hereinafter referred to as “CPU”) 21, a memory 22, a disk 23, a nonvolatile recording medium 24, an input device 25, an output device 26, and a communication interface (hereinafter referred to as “CPU”). Communication IF ”27).
- the calculation processing device 20 can transmit / receive information to / from other calculation processing devices and communication devices via the communication IF 27.
- the non-volatile recording medium 24 is, for example, a compact disk (Compact_Disc), a digital versatile disk (Digital_Versatile_Disc), a universal serial bus memory (USB memory), a solid state drive (Solid_State_Drive), or the like that can be read by a computer.
- the non-volatile recording medium 24 retains such a program without being supplied with power, and can be carried.
- the nonvolatile recording medium 24 is not limited to the above-described medium. Further, the program may be carried via the communication network via the communication IF 27 instead of the nonvolatile recording medium 24.
- the CPU 21 copies a software program (computer program: hereinafter simply referred to as “program”) stored in the disk 23 to the memory 22 and executes arithmetic processing.
- the CPU 21 reads data necessary for program execution from the memory 22. When the display is necessary, the CPU 21 displays the output result on the output device 26. When inputting a program from the outside, the CPU 21 reads the program from the input device 25.
- the CPU 21 executes the availability analysis program (FIGS. 2, 3, and FIG. 2) in the memory 22 corresponding to the function (process) represented by each unit shown in FIG. 1, FIG. 6, FIG. 9, FIG. 4, FIG. 7, FIG. 8, or FIG. 10) is interpreted and executed.
- the CPU 21 sequentially performs the processes described in the above-described embodiments of the present invention.
- the present invention can also be achieved by such an availability analysis program. Furthermore, it can be understood that the present invention can also be realized by a computer-readable non-volatile recording medium in which the availability analysis program is recorded.
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Quality & Reliability (AREA)
- Mathematical Optimization (AREA)
- Artificial Intelligence (AREA)
- Data Mining & Analysis (AREA)
- Mathematical Analysis (AREA)
- Computational Mathematics (AREA)
- Pure & Applied Mathematics (AREA)
- Computing Systems (AREA)
- Evolutionary Computation (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Algebra (AREA)
- Probability & Statistics with Applications (AREA)
- Computer Hardware Design (AREA)
- Debugging And Monitoring (AREA)
- Test And Diagnosis Of Digital Computers (AREA)
Abstract
Description
(手順1)対象システムに関する状態遷移をモデルとして表現する、
(手順2)該モデルに基づき確率過程を分析することにより、対象システムが利用可能な状態にある確率を算出する。
(I)対象システムに含まれるコンポーネントの状態間における遷移率を表すコンポーネント情報と、(II)前記対象システムがとり得る複数の状態のうち、前記対象システムが稼動できない状態を表す障害状態である場合における前記コンポーネントの状態を表す条件を含む障害情報と、(III)前記対象システムが稼動している状態を表す稼動状態に、前記対象システムが前記障害状態から遷移する場合の遷移率を含む復旧情報とに基づき、前記複数の状態に含まれる2つの状態間に関する値を算出し、算出した前記2つの状態間に関する値に基づいて、前記対象システムが、ある状態にある確率を算出し、前記対象システムが前記稼動状態になっている場合の前記確率に基づいて、前記対象システムに関する可用性を算出する解析手段
を備える。
(I)対象システムに含まれるコンポーネントの状態間における遷移率を表すコンポーネント情報と、(II)前記対象システムがとり得る複数の状態のうち、前記対象システムが稼動できない状態を表す障害状態である場合における前記コンポーネントの状態を表す条件を含む障害情報と、(III)前記対象システムが稼動している状態を表す稼動状態に、前記対象システムが前記障害状態から遷移する場合の遷移率を含む復旧情報とに基づき、前記複数の状態に含まれる2つの状態間に関する値を算出し、算出した前記2つの状態間に関する値に基づいて、前記対象システムが、ある状態にある確率を算出し、前記対象システムが前記稼動状態になっている場合の前記確率に基づいて、前記対象システムに関する可用性を算出する。
Q(I、I)=-(Σ(J≠I)Q(I,J))・・・(式1)。
ΣIπI=1・・・(式2)、
(ただし、πIは、対象システムが、定常状態において、第Iシステム状態である確率を表す。また、ΣIは、1乃至Nにて、総和を算出することを表す。#は、行列ベクトル積を表す)。
本実施形態においては、以下の順序にて、可用性分析装置について説明する。尚、カッコ内には、参照する図面が記載されている。
(2)可用性分析装置が有する入力部における処理について(図4)、
(3)対象システムに含まれるコンポーネントのコンポーネント状態について(図5)、
(4)可用性分析装置における処理の流れについて(図2)、
(5)可用性分析装置が有する計算部における処理の流れについて(図3)。
πi (k+1)=-1÷qii×Σ(i≠j)(qij×πj (k))・・・(式3)、
(ただし、πi (k)は、数値列π(k)におけるi番目の数値(すなわち、対象システムが第iシステム状態である確率)を表す。qijは、第iシステム状態から第jシステム状態に遷移する遷移率を表す。Σ(i≠j)は、iとjとが異なる値の場合における和を算出することを表す)。
(ただし、||は絶対値を算出することを表す)。
次に、上述した第1の実施形態を基本とする本発明の第2の実施形態について説明する。
(理由1)第2の実施形態に係る可用性分析装置111が有する構成は、第1の実施形態に係る可用性分析装置101が有する構成を含むからである、
(理由2)非可達状態に関する処理が減るからである。
次に、上述した第2の実施形態を基本とする本発明の第3の実施形態について説明する。
(理由1)第3の実施形態に係る可用性分析装置123が有する構成は、第2の実施形態に係る可用性分析装置111が有する構成を含むからである、
(理由2)遷移情報を作成することにより、第Iシステム状態から第Jシステム状態に遷移する場合の遷移率等を繰り返し算出する必要がないからである。
次に、上述した第3の実施形態を基本とする本発明の第4の実施形態について説明する。
(ただし、∧は、論理積を表す。∨は、論理和を表す)。
障害条件FS:x2∧x3∨x2∧x4∨x3∧x4・・・(式8)。
(障害条件FS、(0、0、0、0)、aS)・・・(式10)。
(ただし、+は、足し算を表す)。
個数A:「(1つにまとめて処理するシステム障害状態を構成するシステム障害状態の状態数)-1」。
(ただし、Σ(G(J)∋K)は、インデックスの集合G(J)に含まれる要素Kに関して総和を算出することを表す)。
○第2行目に示すシステム稼動状態から、第4列目に示すシステム障害状態に遷移する場合の遷移率λd、
○第2行目に示すシステム稼動状態から、第6列目に示すシステム障害状態に遷移する場合の遷移率λd、
○第2行目に示すシステム稼動状態から、第7列目に示すシステム障害状態に遷移する場合の遷移率0。
○第3行目に示すシステム稼動状態から、第4列目に示すシステム障害状態に遷移する場合の遷移率λd、
○第3行目に示すシステム稼動状態から、第6列目に示すシステム障害状態に遷移する場合の遷移率0、
○第3行目に示すシステム稼動状態から、第7列目に示すシステム障害状態に遷移する場合の遷移率λd。
○第1行目に示すシステム稼動状態から、第4列目に示すシステム障害状態に遷移する場合の遷移率0、
○第1行目に示すシステム稼動状態から、第6列目に示すシステム障害状態に遷移する場合の遷移率0、
○第1行目に示すシステム稼動状態から、第7列目に示すシステム障害状態に遷移する場合の遷移率0。
(理由1)第4の実施形態に係る可用性分析装置133が有する構成は、第3の実施形態に係る可用性分析装置123が有する構成を含むからである、
(理由2)複数のシステム障害状態を1つのシステム障害状態として処理する結果、行列Qの大きさが、第3の実施形態に係る可用性分析装置123に比べ、さらに小さくなるからである。
次に、上述した本発明の各実施形態の基本となる本発明の第5の実施形態について説明する。
(1)対象システムに含まれるコンポーネントのコンポーネント状態間における遷移率を表すコンポーネント情報、
(2)対象システムがとり得る複数のシステム状態のうち、対象システムが稼動できないシステム状態を表すシステム障害状態である場合における、コンポーネントのコンポーネント状態を表す条件を含む障害情報、
(3)対象システムが稼動している状態を表すシステム稼動状態に、対象システムがシステム障害状態から遷移する場合の遷移率を含む復旧情報。
上述した本発明の各実施形態における可用性分析装置を、1つの計算処理装置(情報処理装置、コンピュータ)を用いて実現するハードウェア資源の構成例について説明する。
但し、係る可用性分析装置は、物理的または機能的に少なくとも2つの計算処理装置を用いて実現してもよい。また、係る可用性分析装置は、専用の装置として実現してもよい。
102 計算部
103 解析部
104 入力部
501 障害情報
502 復旧情報
503 可用性
111 可用性分析装置
112 入力部
113 計算部
114 作成部
121 判定部
122 遷移情報作成部
123 可用性分析装置
124 解析部
131 判定部
132 遷移情報作成部
133 可用性分析装置
151 可用性分析装置
152 解析部
521 ホストコンピュータ
522 ストレージシステム
523 バックアップシステム
524 RAIDコントローラ
525 記憶装置
526 記憶装置
527 記憶装置
528 記憶装置
20 計算処理装置
21 CPU
22 メモリ
23 ディスク
24 不揮発性記録媒体
25 入力装置
26 出力装置
27 通信IF
Claims (10)
- (I)対象システムに含まれるコンポーネントの状態間における遷移率を表すコンポーネント情報と、(II)前記対象システムがとり得る複数の状態のうち、前記対象システムが稼動できない状態を表す障害状態である場合における前記コンポーネントの状態を表す条件を含む障害情報と、(III)前記対象システムが稼動している状態を表す稼動状態に、前記対象システムが前記障害状態から遷移する場合の遷移率を含む復旧情報とに基づき、前記複数の状態に含まれる2つの状態間に関する値を算出し、算出した前記2つの状態間に関する値に基づいて、前記対象システムが、ある状態にある確率を算出し、前記対象システムが前記稼動状態になっている場合の前記確率に基づいて、前記対象システムに関する可用性を算出する解析手段
を備える可用性分析装置。 - 前記2つの状態に関する値は、第1状態識別子が表す状態から第2状態識別子が表す状態への遷移に関する値であり、
前記解析手段は、前記障害状態を表す第3状態識別子と、前記条件とが関連付けされている前記障害情報に前記第1状態識別子が含まれない場合に、前記コンポーネント情報に基づき前記値を算出する
請求項1に記載の可用性分析装置。 - 前記解析手段は、
(a)前記第3状態識別子と、前記第3状態識別子が表す前記障害状態から遷移する前記稼動状態を表す第4状態識別子と、前記遷移率とが関連付けされている前記復旧情報において、前記第1状態識別子と、前記第2状態識別子とが関連付けされている場合に、前記第1状態識別子及び前記第2状態識別子に関連付けされた前記遷移率を前記値として算出し、
(b)前記第1状態識別子が前記障害状態に含まれ、前記第1状態識別子及び前記第2状態識別子が一致する場合に、前記復旧情報において、前記第1状態識別子に関連付けされた「前記遷移率×(-1)」を前記値として算出し、
(c)前記第1状態識別子が前記障害状態に含まれ、前記(a)及び前記(b)でない場合に、0を前記値として算出する
請求項2に記載の可用性分析装置。 - 前記解析手段は、前記対象システムにおいて達成され得ない状態を表す状態識別子が含まれる非可達情報に、前記第1状態識別子または前記第2状態識別子が含まれる場合に0を前記値として算出し、前記非可達情報が受信した前記状態識別子のいずれも含まない場合に、前記(I)、前記(II)、及び、前記(III)に基づき、前記値を算出する
請求項1乃至請求項3のいずれか一項に記載の可用性分析装置。 - 前記対象システムにおいて達成され得る状態を表す可達状態を識別可能な状態識別子が含まれる可達情報に含まれる状態識別子の個数が、所定の個数以下であるか否かを判定する判定手段と、
前記可達情報に含まれる状態識別子の個数が所定の個数以下である場合に、前記可達状態に関して、前記解析手段が算出する前記値を格納する遷移情報を作成する作成手段と
を備え、
前記解析手段は、前記遷移情報に基づき、前記可用性を算出する
請求項1乃至請求項4のいずれか一項に記載の可用性分析装置。 - 前記判定手段は、前記可達情報に含まれる状態識別子のうち、前記障害状態を1つの状態として設定することにより前記状態識別子の個数を算出し、前記算出した状態識別子の個数が前記所定の個数以下であるか否かを判定し、
前記作成手段は、前記障害状態を1つの状態として、前記遷移情報を作成する
請求項5に記載の可用性分析装置。 - (I)対象システムに含まれるコンポーネントの状態間における遷移率を表すコンポーネント情報と、(II)前記対象システムがとり得る複数の状態のうち、前記対象システムが稼動できない状態を表す障害状態である場合における前記コンポーネントの状態を表す条件を含む障害情報と、(III)前記対象システムが稼動している状態を表す稼動状態に、前記対象システムが前記障害状態から遷移する場合の遷移率を含む復旧情報とに基づき、前記複数の状態に含まれる2つの状態間に関する値を算出し、算出した前記2つの状態間に関する値に基づいて、前記対象システムが、ある状態にある確率を算出し、前記対象システムが前記稼動状態になっている場合の前記確率に基づいて、前記対象システムに関する可用性を算出する可用性分析方法。
- (I)対象システムに含まれるコンポーネントの状態間における遷移率を表すコンポーネント情報と、(II)前記対象システムがとり得る複数の状態のうち、前記対象システムが稼動できない状態を表す障害状態である場合における前記コンポーネントの状態を表す条件を含む障害情報と、(III)前記対象システムが稼動している状態を表す稼動状態に、前記対象システムが前記障害状態から遷移する場合の遷移率を含む復旧情報とに基づき、前記複数の状態に含まれる2つの状態間に関する値を算出し、算出した前記2つの状態間に関する値に基づいて、前記対象システムが、ある状態にある確率を算出し、前記対象システムが前記稼動状態になっている場合の前記確率に基づいて、前記対象システムに関する可用性を算出する解析機能
をコンピュータに実現させる可用性分析プログラムを格納する記録媒体。 - 前記2つの状態に関する値は、第1状態識別子が表す状態から第2状態識別子が表す状態への遷移に関する値であり、
前記解析機能において、前記障害状態を表す第3状態識別子と、前記条件とが関連付けされている前記障害情報に前記第1状態識別子が含まれない場合に、前記コンポーネント情報に基づき前記値を算出する
請求項8に記載の可用性分析プログラムを格納する記録媒体。 - 前記解析機能において、
(a)前記第3状態識別子と、前記第3状態識別子が表す前記障害状態から遷移する前記稼動状態を表す第4状態識別子と、前記遷移率とが関連付けされている前記復旧情報において、前記第1状態識別子と、前記第2状態識別子とが関連付けされている場合に、前記第1状態識別子及び前記第2状態識別子に関連付けされた前記遷移率を前記値として算出し、
(b)前記第1状態識別子が前記障害状態に含まれ、前記第1状態識別子及び前記第2状態識別子が一致する場合に、前記復旧情報において、前記第1状態識別子に関連付けされた「前記遷移率×(-1)」を前記値として算出し、
(c)前記第1状態識別子が前記障害状態に含まれ、前記(a)及び前記(b)でない場合に、0を前記値として算出する
請求項9に記載の可用性分析プログラムを格納する記録媒体。
Priority Applications (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
JP2016513646A JPWO2015159549A1 (ja) | 2014-04-16 | 2015-04-16 | 可用性分析装置、可用性分析方法、及び、可用性分析プログラム |
US15/129,919 US20170147459A1 (en) | 2014-04-16 | 2015-04-16 | Availability analysis device, availability analysis method, and recording medium having availability analysis program recorded therein |
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
JP2014-084087 | 2014-04-16 | ||
JP2014084087 | 2014-04-16 |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2015159549A1 true WO2015159549A1 (ja) | 2015-10-22 |
Family
ID=54323770
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/JP2015/002092 WO2015159549A1 (ja) | 2014-04-16 | 2015-04-16 | 可用性分析装置、可用性分析方法、及び、可用性分析プログラムが記録された記録媒体 |
Country Status (3)
Country | Link |
---|---|
US (1) | US20170147459A1 (ja) |
JP (1) | JPWO2015159549A1 (ja) |
WO (1) | WO2015159549A1 (ja) |
Families Citing this family (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP6864992B2 (ja) * | 2016-04-28 | 2021-04-28 | 日立Astemo株式会社 | 車両制御システム検証装置及び車両制御システム |
-
2015
- 2015-04-16 US US15/129,919 patent/US20170147459A1/en not_active Abandoned
- 2015-04-16 WO PCT/JP2015/002092 patent/WO2015159549A1/ja active Application Filing
- 2015-04-16 JP JP2016513646A patent/JPWO2015159549A1/ja active Pending
Non-Patent Citations (1)
Title |
---|
FUMIO MACHIDA ET AL.: "Candy: Component-based Availability Modeling Framework for Cloud Service Management Using SysML", PROCEEDINGS OF THE 30TH IEEE SYMPOSIUM ON RELIABLE DISTRIBUTED SYSTEMS, pages 209 - 218, XP032021998 * |
Also Published As
Publication number | Publication date |
---|---|
US20170147459A1 (en) | 2017-05-25 |
JPWO2015159549A1 (ja) | 2017-04-13 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US9280416B1 (en) | Selection of erasure code parameters for no data repair | |
US10157105B2 (en) | Method for data protection for cloud-based service system | |
US7925927B2 (en) | Simulator for determining data loss in a fault tolerant system | |
US11048839B2 (en) | Adaptive error correction in quantum computing | |
US20090327195A1 (en) | Root cause analysis optimization | |
CN103544202A (zh) | 用于安排数据处理的方法和系统 | |
KR102109015B1 (ko) | 부분 접속 복구가 가능하고 중복 인코딩이 용이한 데이터 저장 방법 및 시스템 | |
JP2021521504A (ja) | 階層型ストレージを使用したコンピュータ上の量子回路のシミュレーション | |
US10592344B1 (en) | Generation and verification of erasure encoded fragments | |
CN106575244A (zh) | 确保云应用的高可用性的修补过程 | |
Mohror et al. | Detailed modeling and evaluation of a scalable multilevel checkpointing system | |
CN115185736B (zh) | 基于图卷积神经网络的微服务调用链异常检测方法及装置 | |
JP7059214B2 (ja) | 演算装置 | |
US9489254B1 (en) | Verification of erasure encoded fragments | |
Barenco et al. | Effects of noise on quantum error correction algorithms | |
JP2019053474A (ja) | クラウドベースサービスのデータ保護方法 | |
WO2015159549A1 (ja) | 可用性分析装置、可用性分析方法、及び、可用性分析プログラムが記録された記録媒体 | |
US9552254B1 (en) | Verification of erasure encoded fragments | |
CN103812719B (zh) | 集群系统的失效预测方法及装置 | |
US9489252B1 (en) | File recovery using diverse erasure encoded fragments | |
Zwingmann et al. | Optimal disassembly sequencing strategy using constraint programming approach | |
CN113297338B (zh) | 产品推荐路径生成方法、装置、设备及存储介质 | |
US20210279575A1 (en) | Information processing apparatus, information processing method, and storage medium | |
Levitin et al. | Performance distribution of a fault-tolerant system in the presence of failure correlation | |
Galinanes et al. | Ensuring data durability with increasingly interdependent content |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 15779400 Country of ref document: EP Kind code of ref document: A1 |
|
ENP | Entry into the national phase |
Ref document number: 2016513646 Country of ref document: JP Kind code of ref document: A |
|
WWE | Wipo information: entry into national phase |
Ref document number: 15129919 Country of ref document: US |
|
NENP | Non-entry into the national phase |
Ref country code: DE |
|
122 | Ep: pct application non-entry in european phase |
Ref document number: 15779400 Country of ref document: EP Kind code of ref document: A1 |