WO2013114830A1 - Process prediction execution device and process prediction execution method - Google Patents

Process prediction execution device and process prediction execution method Download PDF

Info

Publication number
WO2013114830A1
WO2013114830A1 PCT/JP2013/000349 JP2013000349W WO2013114830A1 WO 2013114830 A1 WO2013114830 A1 WO 2013114830A1 JP 2013000349 W JP2013000349 W JP 2013000349W WO 2013114830 A1 WO2013114830 A1 WO 2013114830A1
Authority
WO
WIPO (PCT)
Prior art keywords
event
prediction
knowledge
combination
unit
Prior art date
Application number
PCT/JP2013/000349
Other languages
French (fr)
Japanese (ja)
Inventor
山本　浩
Original Assignee
日本電気株式会社
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 日本電気株式会社 filed Critical 日本電気株式会社
Publication of WO2013114830A1 publication Critical patent/WO2013114830A1/en

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"

Definitions

  • the present invention relates to a process prediction execution apparatus, a process prediction execution method, and a program for predicting and executing a process to be executed next by a system.
  • Patent Document 1 A technique for predicting and executing system processing is described in Patent Document 1.
  • the business process execution method described in Patent Document 1 statistically analyzes executed business process history information. Next, based on the analysis result, the business process execution method predicts a business process that is likely to transition to the execution state next among the business process instances in the standby state.
  • the business process execution method is characterized in that a business process instance previously executes an activity to be executed in the future.
  • the business process execution method described in Patent Document 1 can effectively use CPU time and reduce the execution time of a business process by adopting the above-described configuration.
  • Patent Document 1 it is not possible to perform predictive execution of processes for which processes are not defined in advance, such as a flow composed of a plurality of operations executed by a system administrator. The reason is that the technology described in Patent Document 1 is capable of predicting and executing an activity to be executed in the future because the business process definition is a target with a predetermined flow.
  • an object of the present invention is to provide a technique capable of accurately performing prediction even for an unknown flow for which a process is not defined in advance.
  • the system prediction execution apparatus of the present invention uses a combination of events having continuity as a combination of a current event that is an event currently being executed and a prediction event that is an event predicted from the current event.
  • a prediction event that is an event predicted from the current event.
  • the system prediction execution method uses a combination of events having continuity as a current event that is a currently executed event and a prediction event that is an event predicted from the current event.
  • a prediction event that is an event predicted from the current event.
  • the program according to the present invention stores a combination of events having continuity as a combination of a current event that is currently being executed and a predicted event that is an event predicted from the current event.
  • an event generated by a user operation matches the current event stored in the predicted knowledge storage unit, an operation corresponding to the predicted event combined with the current event is executed. To run.
  • the process prediction execution apparatus According to the process prediction execution apparatus, the process prediction execution method, and the program of the present invention, it is possible to perform prediction execution with high accuracy even for an unknown flow in which no process is defined in advance.
  • FIG. 1 is a diagram for explaining an example of a software arrangement method.
  • FIG. 2 is a diagram showing the configuration of the process prediction execution system according to the first embodiment of the present invention.
  • FIG. 3 is a diagram schematically showing the logical structure of the program of this embodiment that operates in the parent distribution server 1100.
  • FIG. 4 is a diagram illustrating an example of operation definition.
  • FIG. 5 is a diagram illustrating an example of a process definition.
  • FIG. 6 is a diagram illustrating an example of an operation flow definition.
  • FIG. 7 is a diagram illustrating an example of data stored in the prediction knowledge storage unit 1152.
  • FIG. 8 is a flowchart showing the flow of processing in the process prediction execution apparatus 1100 according to the first embodiment.
  • FIG. 8 is a flowchart showing the flow of processing in the process prediction execution apparatus 1100 according to the first embodiment.
  • FIG. 9 is a flowchart showing the flow of the prediction execution process executed by the prediction execution unit 1130.
  • FIG. 10 is a flowchart showing a flow of event continuity determination processing executed by the event continuity determination unit 1121.
  • FIG. 11 is a flowchart showing the flow of prediction knowledge creation processing executed by the prediction knowledge creation unit 1140.
  • FIG. 12 is a flowchart showing the flow of the parameter variable process executed by the parameter variable unit 1141.
  • FIG. 13 is a diagram illustrating a state where there is no prediction knowledge.
  • FIG. 14 is a diagram illustrating a state in which new knowledge is registered as predicted knowledge.
  • FIG. 15 is a diagram illustrating a state in which new knowledge is registered as predicted knowledge.
  • FIG. 16 is a diagram illustrating a state in which new knowledge is registered as predicted knowledge.
  • FIG. 17 is a diagram showing a configuration of a process prediction execution apparatus 2000 according to the second embodiment of the present invention.
  • FIG. 18 is a block diagram illustrating an example of a
  • One of the distribution methods using a software distribution tool is a method of communicating with a client program operating on the OS of a distribution target machine using TCP / IP, etc., transmitting software, and installing on the OS.
  • a distribution method using a software distribution tool there is a method in which a distribution target machine is network booted and software such as an OS is directly installed on a disk of the distribution target machine via a boot program.
  • System requirements necessary for distributing software differ depending on the distribution method.
  • the system requirement is that an OS is installed on the distribution target machine, and that the client program is installed on the OS.
  • the system requirements are that the distribution target machine can be network booted and that the network boot server (distribution server) is located in the same network as the distribution target machine. become.
  • cloud computing In recent years, the use of a large-scale data center via a network (cloud computing) has attracted attention, but software distribution work is also required when building and managing a cloud computing environment. In that case, it is assumed that a software distribution tool is used to make the software distribution work more efficient.
  • the cloud computing environment is characterized in that a plurality of people and companies are mixed as users (tenants) using the system, and that a network between tenants needs to be separated from the viewpoint of security.
  • the system requirements for software distribution tools also depend on the system network environment. For this reason, in the case of a configuration in which a large number of networks exist as in a cloud computing environment, the whole cannot be managed with a centralized configuration (one distribution server is prepared and distributed to all machines from there). Therefore, for example, it is necessary to devise a method for configuring the software distribution tool, such as arranging a component (child distribution server) having a distribution server function for each network.
  • the software itself such as the OS, patches, and applications to be distributed only needs to be managed centrally.
  • the component child distribution server
  • the distribution server function that exists in the same network as the distribution target machine It is necessary to consider the location of the software, such as copying the software to be used.
  • FIG. 1 is a diagram for explaining an example of a software arrangement method.
  • the main software is arranged in the parent distribution server 5100 and the parent distribution server 5100 to the child distribution server at the timing when the software needs to be distributed to the distribution target machine 6100. Copy software to 5200.
  • the software arrangement method as shown in FIG. 1 has a problem in that the distribution speed is slow because software is sequentially copied as necessary.
  • FIG. 2 is a diagram showing the configuration of the process prediction execution system according to the first embodiment.
  • the process prediction execution system includes a parent distribution server 1100, a child distribution server 1200 to a child distribution server 1300, a distribution target machine 2100 to a distribution target machine 2500, a network device 3100 to a network device 3200, and a management network 4100.
  • a management network 4300 is a diagram showing the configuration of the process prediction execution system according to the first embodiment.
  • the process prediction execution system includes a parent distribution server 1100, a child distribution server 1200 to a child distribution server 1300, a distribution target machine 2100 to a distribution target machine 2500, a network device 3100 to a network device 3200, and a management network 4100.
  • a management network 4300 is a diagram showing the configuration of the process prediction execution system according to the first embodiment.
  • the process prediction execution system includes a parent distribution server 1100, a child distribution server 1200 to a child distribution server 1300, a distribution target machine 2100 to
  • the parent distribution server 1100 is hardware equipped with software for managing and controlling the child distribution server 1200 to the child distribution server 1300 via the management network 4100 to the management network 4300.
  • the child distribution server 1200 is hardware equipped with software for managing and controlling the distribution target machine 2100 to the distribution target machine 2300 via the management network 4200.
  • the child distribution server 1300 is hardware equipped with software for managing and controlling the distribution target machine 2400 to the distribution target machine 2500 via the management network 4300.
  • the network device 3100 to the network device 3200 and the management network 4100 to the management network 4300 are configured such that the parent distribution server 1100, the child distribution server 1200 to the child distribution server 1300, and the distribution target machine 2100 to the distribution target machine 2500 communicate with each other. It is a network infrastructure that goes through.
  • FIG. 2 is an illustration for schematically explaining the present embodiment, and does not limit the system configuration. That is, the child distribution server and the distribution target machine can be configured as an arbitrary number according to the system to be managed and controlled.
  • FIG. 3 is a diagram schematically showing the logical structure of the program of this embodiment that operates in the parent distribution server 1100.
  • the parent distribution server 1100 includes a working area 1110, an event processing unit 1120, a prediction execution unit 1130, a prediction knowledge creation unit 1140, and a data storage area 1150.
  • the parent distribution server 1100 is also referred to as a process prediction execution apparatus 1100.
  • the working area 1110 corresponds to a volatile storage area such as a main memory.
  • the event processing unit 1120, the prediction execution unit 1130, and the prediction knowledge creation unit 1140 correspond to program modules executed by a CPU (Central Processing Unit) or the like.
  • the data storage area 1150 corresponds to a nonvolatile storage device such as a magnetic disk, and a DBMS (DataBase Management System) that manages data stored therein.
  • DBMS DataBase Management System
  • FIG. 18 is a block diagram showing an example of a non-volatile recording medium 1170 that records (stores) the code of the above-described program module.
  • the non-volatile recording medium 1170 may be supplied to the parent distribution server 1100, and the CPU (not shown) of the parent distribution server 1100 may read and execute the program code stored in the non-volatile recording medium 1170.
  • the parent distribution server is, for example, a computer.
  • the CPU may store the code of the program stored in the nonvolatile recording medium 1170 in a storage unit (not shown). That is, the present embodiment includes an embodiment of a nonvolatile recording medium 1170 that stores a program (software) executed by a computer (CPU).
  • the working area 1110 includes “past events for storing past events”, “current events for storing detected events”, and “event processing unit 1120, prediction execution unit 1130”.
  • a queue for storing temporary data in the process of the prediction knowledge creating unit 1140 is stored.
  • the queue is a storage area having a feature that first stored (enqueued) data is first extracted (dequeued).
  • the event processing unit 1120 includes an event continuity determination unit 1121 inside.
  • the event processing unit 1120 detects a management operation (operation) executed by the system administrator as an event, the event processing unit 1120 requests the prediction execution unit 1130 to perform prediction execution of an operation predicted from the event.
  • the event processing unit 1120 uses the event continuity determination unit 1121 to determine whether there is continuity between past events and detected events. Next, when it is determined that there is continuity, the event processing unit 1120 requests the prediction knowledge creation unit 1140 to create prediction knowledge.
  • the event processing unit 1120 is a program module that executes the above operations.
  • the prediction execution unit 1130 is a program module that predicts and executes an operation predicted from an event detected by the event processing unit 1120 with reference to a prediction knowledge storage unit 1152 described later in response to a request from the event processing unit 1120. .
  • the prediction knowledge creation unit 1140 is a program module that includes a parameter variable conversion unit 1141 therein and creates prediction knowledge in the prediction knowledge storage unit 1152 in response to a request from the event processing unit 1120.
  • the data storage area 1150 includes an operation definition storage unit 1151 and a prediction knowledge storage unit 1152.
  • the operation definition storage unit 1151 stores an operation definition, a process definition, and an operation flow definition.
  • Operation definition defines management operations performed by the system administrator.
  • FIG. 4 is a diagram illustrating an example of operation definition. As shown in a tree structure at the top of FIG. 4, the operations include machine registration, image registration, image distribution, and the like. As shown below the tree structure in FIG. 4, the operation definition defines a class corresponding to the type of operation, an attribute corresponding to an argument of the operation, and the like.
  • FIG. 5 is a diagram illustrating an example of a process definition.
  • the processes include machine registration processing, image registration processing, image distribution processing, and image transmission processing.
  • the process definition defines a class corresponding to the type of process, an attribute corresponding to internal data, and the like.
  • the process definition may define process arguments, methods corresponding to the actual processing of the process, and the like.
  • each process becomes a prediction execution target when “true” is defined as the value of the attribute “prediction execution target” of each process, and does not become a prediction execution target when “false” is defined.
  • the value of “prediction execution target” may be defined in advance by the user.
  • Operation flow definition defines what process each operation is composed of.
  • FIG. 6 is a diagram illustrating an example of an operation flow definition. As shown in FIG. 6, the flow is defined by a process as an element.
  • the prediction knowledge storage unit 1152 stores knowledge for predicting an event (operation) that will occur next when a certain event (operation) occurs.
  • FIG. 7 is a diagram illustrating an example of data stored in the prediction knowledge storage unit 1152.
  • the data stored in the prediction knowledge storage unit 1152 includes the current event and its parameters, the prediction event and its parameters, and the number of executions.
  • This predictive knowledge expresses that when an event detected by the event processing unit 1120 matches the current event and its parameters, an operation corresponding to the predictive event and parameters may be executed in the near future. Yes.
  • FIG. 8 is a flowchart showing a process flow in the process prediction execution apparatus 1100 according to the first embodiment.
  • the event processing unit 1120 refers to the operation definition stored in the operation definition storage unit 1151 as an event. Detect (Step A1).
  • the event processing unit 1120 calls the prediction execution unit 1130 with the event as an input, and executes the prediction execution process (step A2).
  • the event processing unit 1120 uses the event continuity determination unit 1121 to execute an event continuity determination process for determining whether the detected event is continuous with a past event (step S1). A3).
  • the event continuity is determined by, for example, storing the event detected last time and determining that the occurrence interval between the event and the event detected this time is within a certain time. You may judge.
  • the event processing unit 1120 calls the prediction knowledge creating unit 1140 and executes the prediction knowledge creating process (Step A5).
  • Step A4 If it is determined that the detected event is not continuous with the past event (N in Step A4), the process prediction execution apparatus 1100 ends the event processing and waits for the next event to occur.
  • FIG. 9 is a flowchart showing the flow of the prediction execution process executed by the prediction execution unit 1130. As shown in FIG. 9, the prediction execution unit 1130 stores the input event as a current event in the working area 1110 (step B1).
  • the prediction execution unit 1130 When the current event matches the current event of the prediction knowledge stored in the prediction knowledge storage unit 1152 (Y in Step B2), the prediction execution unit 1130 next executes the operation corresponding to the prediction event of the entry.
  • the operation is considered as a highly likely operation (step B3).
  • the prediction execution unit 1130 identifies a matching entry by determining that the one with the highest number of executions is matched.
  • the processes that are actually executed are only processes defined as prediction execution targets (“image transmission process” in FIG. 5, “image transmission process” in the “image distribution” operation in FIG. 6, etc.).
  • the prediction execution unit 1130 determines that there is no predicted operation and ends the process.
  • the prediction execution unit 1130 stores the event of the operation that has been predicted execution in the current event of the working area 1110, and performs the processing from Step B2. repeat.
  • the prediction execution unit 1130 ends the prediction execution process.
  • the determination as to whether or not to continue the prediction execution may be performed by, for example, providing the number of prediction executions and controlling the number of prediction execution processes.
  • FIG. 10 is a flowchart showing the flow of event continuity determination processing executed by the event continuity determination unit 1121. As shown in FIG. 10, the event continuity determination unit 1121 determines whether or not the time interval between the past event and the detected event is within a predetermined time (step C1).
  • Step C2 If it is determined that the time interval is within the predetermined time (Y in Step C1), the event continuity determination unit 1121 determines that the events are continuous (Step C2).
  • the event continuity determination unit 1121 determines that the events are not continuous (Step C3).
  • the determination method of event continuity shown in FIG. 10 is an example, and the present invention is not limited to this.
  • the event continuity determination unit 1121 may determine event continuity by any other method as long as it is a reasonable method for determining event continuity.
  • FIG. 11 is a flowchart showing a flow of predicted knowledge creation processing executed by the predicted knowledge creation unit 1140. As illustrated in FIG. 11, the prediction knowledge creating unit 1140 enqueues a combination of a past event and a detected event into a queue in the working area 1110 (Step D1).
  • the prediction knowledge creation unit 1140 calls the parameter variable conversion unit 1141. Thereby, the parameter variable conversion unit 1141 executes the parameter variable conversion process (step D2).
  • the prediction knowledge creating unit 1140 dequeues the combination of the past event and the detected event from the queue in the working area 1110 (Step D3).
  • the predicted knowledge creating unit 1140 compares the dequeued combination with the predicted knowledge and, if there is a matching entry (Y in step D4), increments the number of times the entry is executed (step D5). If they do not match (N in step D4), the predicted knowledge creating unit 1140 registers the new knowledge in the predicted knowledge (step D6). The prediction knowledge creating unit 1140 repeats these processes as many times as the number of queue entries (step D7).
  • FIG. 12 is a flowchart showing the flow of the parameter variable process executed by the parameter variable unit 1141.
  • the parameter variable converting unit 1141 confirms whether or not the combination of the past event and the detected event matches the class of the entry (current event, predicted event) in the predicted knowledge. (Step E1).
  • step E3 If the classes match (Y in step E1), if the parameter of the past event is different from the parameter value of the current event in the prediction knowledge (N in step E2), the parameter variableizing unit 1141 changes the parameter that does not match the variable A combination of the converted past event and the detected event is enqueued in the queue of the working area 1110 (step E3).
  • the parameter variableizing unit 1141 excludes combinations that have not been effective as predictive knowledge, such as combinations in which all parameters have been converted into variables, and combinations in which past events are the same as detected events. You may make it do.
  • Step E4 If the classes do not match (N in Step E1), or if the past event completely matches the current event and its parameters in the prediction knowledge (Y in Step E2), the parameter variableizing unit 1141 enters the next entry Is compared (step E4).
  • the process prediction execution apparatus 1100 creates prediction knowledge suitable for the system environment and the operation environment by repeating the above-described processing described with reference to FIGS. 8 to 12, and performs prediction execution according to the created knowledge.
  • machine_1 is registered as a distribution target machine and starts from a state where there is no predictive knowledge as shown in FIG. Further, as an operation of the event continuity determination unit 1121, if events occur continuously when the occurrence interval between the event detected last time and the event detected this time is within a certain time (referred to as “certain time TI”). Suppose you decide.
  • Image registration (patch_image_1) after elapse of a certain time TI 2.
  • Image distribution (machine_1, patch_image_1) within a certain time TI 3.
  • Image registration (patch_image_2) after elapse of a certain time TI 4).
  • Image distribution (machine_1, patch_image_2) within a certain time TI 5.
  • Image registration (patch_image_3) after elapse of a certain time TI 6).
  • Image distribution (machine_1, patch_image_3) within a certain time TI (The event of 1. occurs)
  • the event processing unit 1120 detects the event “image registration (patch_image_1)” (step A1 in FIG. 8).
  • the event processing unit 1120 calls the prediction execution unit 1130 with the event as an input (step A2 in FIG. 8, FIG. 9).
  • the prediction execution unit 1130 stores the input event as a current event in the working area 1110 (step B1 in FIG. 9). Next, the prediction execution unit 1130 determines whether or not the current event matches the current event of the prediction knowledge stored in the prediction knowledge storage unit 1152 (step B2 in FIG. 9). Since there is no matching prediction knowledge (FIG. 13), the prediction execution unit 1130 ends the prediction execution process.
  • the event processing unit 1120 uses the event continuity determination unit 1121 to determine whether the detected event is continuous with past events (step A3 in FIG. 8, FIG. 10). . Since the previous event occurs before the predetermined time TI (N in Step C1 in FIG. 10), the event continuity determination unit 1121 determines that the events are not continuous (Step in FIG. 10). C3).
  • the process prediction execution apparatus 1100 ends the event processing and waits for the next event.
  • Event 2 occurs
  • the event processing unit 1120 detects the event “image distribution (machine_1, patch_image_1)” ( Step A1) in FIG.
  • the event processing unit 1120 calls the prediction execution unit 1130 with the event as an input (step A2 in FIG. 8, FIG. 9). Since there is no prediction knowledge in the prediction knowledge storage unit 1152 (FIG. 13), the prediction execution unit 1130 ends the prediction execution process.
  • the event processing unit 1120 uses the event continuity determination unit 1121 to determine whether the detected event is continuous with past events (step A3 in FIG. 8, FIG. 10). . Since the previous event “image registration (patch_image_1)” has occurred within a predetermined time TI (Y in step C1 in FIG. 10), the event continuity determination unit 1121 determines that the events are continuous (FIG. 10). 10 step C2).
  • the event processing unit 1120 calls the prediction knowledge creating unit 1140 (step A5 in FIG. 8, FIG. 11).
  • the prediction knowledge creation unit 1140 enqueues the combination of the previous event “image registration (patch_image_1)” and the detected event “image distribution (machine_1, patch_image_1)” in the queue of the working area 1110 (step D1 in FIG. 11). Next, the prediction knowledge creation unit 1140 calls the parameter variable conversion unit 1141 (step D2 in FIG. 11, FIG. 12).
  • the parameter variable conversion unit 1141 performs the process of comparing the entry and the class in the prediction knowledge for the combination of the previous event “image registration (patch_image_1)” and the detected event “image delivery (machine_1, patch_image_1)” (FIG. 12). Step E1). However, in this case, since there is no prediction knowledge (FIG. 13), the comparison results do not match (N in step E1 in FIG. 12), and the parameter variableizing unit 1141 ends the parameter variableizing process.
  • the prediction knowledge creating unit 1140 dequeues the combination of the previous event “image registration (patch_image_1)” and the detected event “image delivery (machine_1, patch_image_1)” from the queue in the working area 1110 (FIG. 11 step D3).
  • FIG. 14 is a diagram illustrating a state in which new knowledge is registered as prediction knowledge in FIG. 13.
  • Ending the prediction knowledge creation process end the event process and wait for the next event.
  • the event processing unit 1120 detects the event “image registration (patch_image_2)” (step A1 in FIG. 8). Since the detected event does not match the current event of the prediction knowledge, the prediction execution unit 1130 does not perform the prediction execution process (step A2 in FIG. 8). Next, since the predetermined time TI has passed since the previous event, the event continuity determination unit 1121 does not determine that the event is a continuous event (N in Step A4 in FIG. 8), and the process prediction execution apparatus 1100 End the process and wait for the next event.
  • Event 4 occurs
  • the event processing unit 1120 detects the event “image distribution (machine_1, patch_image_2)” (Ste A1) in FIG. Since the detected event does not match the current event of the prediction knowledge (FIG. 14), the prediction execution unit 1130 does not perform the prediction execution process (step A2 in FIG. 8).
  • the event processing unit 1120 uses the event continuity determination unit 1121 to determine whether the detected event is continuous with past events (step A3 in FIG. 8, FIG. 10). . Since the previous event “image registration (patch_image — 2)” has occurred within a certain time TI (Y in step C1 in FIG. 10), the event continuity determination unit 1121 determines that the events are continuous (FIG. The event processing unit 1120 calls the prediction knowledge creating unit 1140 (step A5 in FIG. 8).
  • the prediction knowledge creating unit 1140 enqueues the combination of the previous event “image registration (patch_image_2)” and the detected event “image delivery (machine_1, patch_image_2)” in the queue of the working area 1110 (step D1 in FIG. 11).
  • the parameter variableizing unit 1141 is called (step D2 in FIG. 11, FIG. 12).
  • the parameter variable conversion unit 1141 performs a process of comparing the entry and the class in the prediction knowledge for the combination of the previous event “image registration (patch_image_2)” and the detected event “image delivery (machine_1, patch_image_2)” (FIG. 12). Step E1). As a result of the comparison processing, the classes match in “image registration” and “image distribution” (Y in step E1 in FIG. 12), and there are parameters that do not match (N in step E2 in FIG. 12). The unit 1141 enqueues “image registration (variable)” and “image distribution (machine_1, variable)”, which are combinations of the parameters that do not match, into a queue (step E3 in FIG. 12).
  • the parameter variable conversion unit 1141 ends the parameter variable conversion process.
  • the prediction knowledge creating unit 1140 dequeues the combination of the previous event “image registration (patch_image_2)” and the detected event “image delivery (machine_1, patch_image_2)” from the queue in the working area 1110 (FIG. 11 step D3). Since the dequeued combination does not match the predicted knowledge (N in step D4 in FIG. 11), the predicted knowledge creating unit 1140 registers the new knowledge in the predicted knowledge (step D6 in FIG. 11).
  • FIG. 15 is a diagram illustrating a state in which new knowledge is registered as prediction knowledge in FIG. 14.
  • FIG. 16 is a diagram illustrating a state in which new knowledge is registered as prediction knowledge in FIG. 15.
  • the process prediction execution apparatus 1100 After completion of the prediction knowledge creation process, the process prediction execution apparatus 1100 ends the event process and waits for the next event.
  • Event 5 occurs
  • the event processing unit 1120 detects the event “image registration (patch_image — 3)” (step A1 in FIG. 8).
  • the event processing unit 1120 calls the prediction execution unit 1130 with the event as an input (step A2 in FIG. 8, FIG. 9).
  • the prediction execution unit 1130 stores the input event as a current event in the working area 1110 (step B1 in FIG. 9).
  • the prediction execution unit 1130 determines whether or not the current event matches the current event of the prediction knowledge stored in the prediction knowledge storage unit 1152 (step B2 in FIG. 9).
  • the prediction execution unit 1130 predictively executes the prediction event “image distribution (machine_1, (variable))” in the same entry (step in FIG. 9). Y of B2, B3).
  • the prediction execution target process in the image distribution operation is “image transmission processing”, and therefore only “image transmission processing” is executed.
  • the process prediction execution apparatus 1100 in the first embodiment it is possible to perform prediction execution with high accuracy even for an unknown flow in which no process is defined in advance.
  • the reason is that the process prediction execution apparatus 1100 according to the first embodiment is capable of performing prediction execution without incurring a special design cost for each individual system even for an unknown flow in which no process is defined in advance. This is because data can be automatically created.
  • the event processing unit 1120 detects an operation as an event.
  • the prediction execution unit 1130 executes the prediction execution process with the event as an input.
  • the event continuity determination unit 1121 determines whether the detected event is continuous with past events.
  • the predicted knowledge creating unit 1140 executes a predicted knowledge creating process.
  • the process prediction execution apparatus 1100 automatically enters a state in which accurate prediction execution is possible.
  • the reason is that the process prediction execution apparatus 1100 automatically extracts a series of operations continuously executed by the system administrator as prediction execution candidates, and manages statistical data of the repeatedly executed operations to improve prediction accuracy. It is because it is letting.
  • an unknown prediction execution candidate can be created. This is because, as an event serving as a trigger (trigger) for predictive execution, not only an event whose parameter value also matches, but also a parameter that is converted into a variable and a matching event is used as the candidate.
  • a trigger for predictive execution, not only an event whose parameter value also matches, but also a parameter that is converted into a variable and a matching event is used as the candidate.
  • FIG. 17 is a diagram illustrating a configuration of the process prediction execution apparatus 2000 according to the second embodiment. As illustrated in FIG. 17, the process prediction execution device 2000 includes a prediction knowledge storage unit 2010 and a prediction execution unit 2020.
  • the prediction knowledge storage unit 2010 stores a combination of events having continuity as a combination of a current event that is an event currently being executed and a prediction event that is an event predicted from the current event.
  • the prediction execution unit 2020 executes an operation corresponding to the prediction event combined with the current event when the event generated by the user operation matches the current event stored in the prediction knowledge storage unit.
  • the process prediction execution apparatus 2000 in the second embodiment it is possible to perform prediction execution with high accuracy even for an unknown flow in which no process is defined in advance.
  • the program of the present invention may be a program that causes a computer to execute each operation described so far.

Landscapes

  • Engineering & Computer Science (AREA)
  • Business, Economics & Management (AREA)
  • Human Resources & Organizations (AREA)
  • Economics (AREA)
  • Strategic Management (AREA)
  • Marketing (AREA)
  • Game Theory and Decision Science (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Development Economics (AREA)
  • Operations Research (AREA)
  • Quality & Reliability (AREA)
  • Tourism & Hospitality (AREA)
  • Physics & Mathematics (AREA)
  • General Business, Economics & Management (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Debugging And Monitoring (AREA)

Abstract

The invention provides a system prediction execution device that enables precise execution of a prediction even for an unknown flow for which a process has not been defined in advance. The system prediction execution device contains a prediction knowledge storage means and a prediction execution means. Said prediction knowledge storage means stores a combination of events having continuity, said combination of events being stored as a combination of: a current event that is an event currently executing; and a prediction event that is an event predicted from the current event. If an event generated by a user operation matches the current event stored by the prediction knowledge storage means, said prediction execution means executes an operation corresponding to the prediction event that is combined with the current event.

Description

プロセス予測実行装置及びプロセス予測実行方法Process prediction execution apparatus and process prediction execution method
　本発明は、システムが次に実行する処理を予測して実行するプロセス予測実行装置、プロセス予測実行方法、及びプログラムに関する。 The present invention relates to a process prediction execution apparatus, a process prediction execution method, and a program for predicting and executing a process to be executed next by a system.
　システムの処理を予測実行する技術が特許文献１に記載されている。 A technique for predicting and executing system processing is described in Patent Document 1.
　特許文献１に記載のビジネスプロセス実行方法は、実行済みのビジネスプロセス履歴情報を統計的に分析する。次に、そのビジネスプロセス実行方法は、その分析結果に基づいて、待機状態にあるビジネスプロセスインスタンスの中で、次に実行状態に遷移する可能性の高いビジネスプロセスを予測する。そして、そのビジネスプロセス実行方法は、ビジネスプロセスインスタンスが、将来実行すべきアクティビティを予め実行しておくことを特徴とする。 The business process execution method described in Patent Document 1 statistically analyzes executed business process history information. Next, based on the analysis result, the business process execution method predicts a business process that is likely to transition to the execution state next among the business process instances in the standby state. The business process execution method is characterized in that a business process instance previously executes an activity to be executed in the future.
　特許文献１に記載のビジネスプロセス実行方法は、上述の構成を採用することにより、ＣＰＵ時間を有効に活用でき、ビジネスプロセスの実行時間を短縮することができる。 The business process execution method described in Patent Document 1 can effectively use CPU time and reduce the execution time of a business process by adopting the above-described configuration.
特開２０１０－１９８３２４号公報JP 2010-198324 A
　しかしながら特許文献１に記載の技術では、システム管理者が実行する複数のオペレーションで構成されるフローのように、予めプロセスが定義されていない処理の予測実行を行うことはできない。その理由は、特許文献１に記載の技術は、ビジネスプロセス定義という予めフローが決まっている対象だからこそ、将来実行されるアクティビティを予測実行することが可能になるからである。 However, with the technique described in Patent Document 1, it is not possible to perform predictive execution of processes for which processes are not defined in advance, such as a flow composed of a plurality of operations executed by a system administrator. The reason is that the technology described in Patent Document 1 is capable of predicting and executing an activity to be executed in the future because the business process definition is a target with a predetermined flow.
　以上より、本発明の目的は、予めプロセスが定義されていない未知のフローに対しても、精度良く予測実行を行うことが可能となる技術を提供することである。 As described above, an object of the present invention is to provide a technique capable of accurately performing prediction even for an unknown flow for which a process is not defined in advance.
　上記目的を達成するため、本発明におけるシステム予測実行装置は、連続性があるイベントの組合せを、現在実行中のイベントであるカレントイベント及び当該カレントイベントから予測されるイベントである予測イベントの組合せとして格納する予測知識格納手段と、ユーザの操作により発生したイベントが、前記予測知識格納手段が格納する前記カレントイベントと一致する場合、当該カレントイベントと組み合わせにされている前記予測イベントに対応するオペレーションを実行する予測実行手段と、を含む。 In order to achieve the above object, the system prediction execution apparatus of the present invention uses a combination of events having continuity as a combination of a current event that is an event currently being executed and a prediction event that is an event predicted from the current event. When the predictive knowledge storage means to store and the event generated by the user's operation match the current event stored by the predictive knowledge storage means, an operation corresponding to the predictive event combined with the current event is performed. Predictive execution means to be executed.
　また、上記目的を達成するため、本発明におけるシステム予測実行方法は、連続性があるイベントの組合せを、現在実行中のイベントであるカレントイベント及び当該カレントイベントから予測されるイベントである予測イベントの組合せとして格納し、ユーザの操作により発生したイベントが、前記予測知識格納手段が格納する前記カレントイベントと一致する場合、当該カレントイベントと組み合わせにされている前記予測イベントに対応するオペレーションを実行する。 In order to achieve the above object, the system prediction execution method according to the present invention uses a combination of events having continuity as a current event that is a currently executed event and a prediction event that is an event predicted from the current event. When an event generated by a user operation stored as a combination matches the current event stored in the prediction knowledge storage means, an operation corresponding to the prediction event combined with the current event is executed.
　また、上記目的を達成するため、本発明におけるプログラムは、連続性があるイベントの組合せを、現在実行中のイベントであるカレントイベント及び当該カレントイベントから予測されるイベントである予測イベントの組合せとして格納し、ユーザの操作により発生したイベントが、前記予測知識格納手段が格納する前記カレントイベントと一致する場合、当該カレントイベントと組み合わせにされている前記予測イベントに対応するオペレーションを実行する、処理をコンピュータに実行させる。 In order to achieve the above object, the program according to the present invention stores a combination of events having continuity as a combination of a current event that is currently being executed and a predicted event that is an event predicted from the current event. When an event generated by a user operation matches the current event stored in the predicted knowledge storage unit, an operation corresponding to the predicted event combined with the current event is executed. To run.
　本発明におけるプロセス予測実行装置、プロセス予測実行方法、及びプログラムによれば、予めプロセスが定義されていない未知のフローに対しても、精度良く予測実行を行うことが可能となる。 According to the process prediction execution apparatus, the process prediction execution method, and the program of the present invention, it is possible to perform prediction execution with high accuracy even for an unknown flow in which no process is defined in advance.
図１は、ソフトウェアの配置方法の一例を説明するための図である。FIG. 1 is a diagram for explaining an example of a software arrangement method. 図２は、本発明の第１実施形態に係るプロセス予測実行システムの構成を表した図である。FIG. 2 is a diagram showing the configuration of the process prediction execution system according to the first embodiment of the present invention. 図３は、親配信サーバ１１００内で動作する本実施形態のプログラムの論理的な構造を模式的に表した図である。FIG. 3 is a diagram schematically showing the logical structure of the program of this embodiment that operates in the parent distribution server 1100. 図４は、オペレーション定義の例を示す図である。FIG. 4 is a diagram illustrating an example of operation definition. 図５は、プロセス定義の例を示す図である。FIG. 5 is a diagram illustrating an example of a process definition. 図６は、オペレーションフロー定義の例を示す図である。FIG. 6 is a diagram illustrating an example of an operation flow definition. 図７は、予測知識格納部１１５２が格納するデータの例を示す図である。FIG. 7 is a diagram illustrating an example of data stored in the prediction knowledge storage unit 1152. 図８は、第１実施形態に係るプロセス予測実行装置１１００における処理の流れを示すフローチャート図である。FIG. 8 is a flowchart showing the flow of processing in the process prediction execution apparatus 1100 according to the first embodiment. 図９は、予測実行部１１３０が実行する予測実行処理の流れを示すフローチャート図である。FIG. 9 is a flowchart showing the flow of the prediction execution process executed by the prediction execution unit 1130. 図１０は、イベント連続性判断部１１２１が実行するイベント連続性判断処理の流れを示すフローチャート図である。FIG. 10 is a flowchart showing a flow of event continuity determination processing executed by the event continuity determination unit 1121. 図１１は、予測知識作成部１１４０が実行する予測知識作成処理の流れを示すフローチャート図である。FIG. 11 is a flowchart showing the flow of prediction knowledge creation processing executed by the prediction knowledge creation unit 1140. 図１２は、パラメータ変数化部１１４１が実行するパラメータ変数化処理の流れを示すフローチャート図である。FIG. 12 is a flowchart showing the flow of the parameter variable process executed by the parameter variable unit 1141. 図１３は、予測知識がない状態を示す図である。FIG. 13 is a diagram illustrating a state where there is no prediction knowledge. 図１４は、予測知識として新規知識が登録された状態を示す図である。FIG. 14 is a diagram illustrating a state in which new knowledge is registered as predicted knowledge. 図１５は、予測知識として新規知識が登録された状態を示す図である。FIG. 15 is a diagram illustrating a state in which new knowledge is registered as predicted knowledge. 図１６は、予測知識として新規知識が登録された状態を示す図である。FIG. 16 is a diagram illustrating a state in which new knowledge is registered as predicted knowledge. 図１７は、本発明の第２実施形態におけるプロセス予測実行装置２０００の構成を示す図である。FIG. 17 is a diagram showing a configuration of a process prediction execution apparatus 2000 according to the second embodiment of the present invention. 図１８は、不揮発性記録媒体の例を示すブロック図である。FIG. 18 is a block diagram illustrating an example of a nonvolatile recording medium.
　＜第１実施形態＞
　まず、本発明の実施形態の理解を容易にするために、本発明の背景を説明する。 <First Embodiment>
First, in order to facilitate understanding of the embodiments of the present invention, the background of the present invention will be described.
　業務サービスを動作させるサーバ、又は業務端末（ＰＣ（Ｐｅｒｓｏｎａｌ　Ｃｏｍｐｕｔｅｒ））で構成されるシステムを構築する場合、ＯＳ（Ｏｐｅｒａｔｉｎｇ　Ｓｙｓｔｅｍ）の初期インストール、累積パッチのインストール、業務アプリケーションのインストールなどソフトウェアのインストール作業が必要になる。この時、ソフトウェア配信ツールを使用してこれらの作業をリモートから一括して実行することで、作業コストを削減することが一般的に行われている。また、システム構築後についても、セキュリティパッチの定期的な適用、業務アプリケーションのアップデートなどを行う場合も、ソフトウェア配信ツールを使用して作業を効率化することが多い。 When building a system consisting of a server that operates business services or business terminals (PCs (Personal Computers)), software installation work such as OS (Operating System) initial installation, cumulative patch installation, business application installation, etc. Is required. At this time, it is a common practice to reduce the work cost by collectively executing these operations remotely using a software distribution tool. Even after system construction, software distribution tools are often used to make work more efficient when security patches are regularly applied and business applications are updated.
　ソフトウェア配信ツールによる配信方式の一つに、配信対象マシンのＯＳ上で動作するクライアントプログラムとＴＣＰ／ＩＰなどで通信を行い、ソフトウェアを送信してＯＳ上でインストールする方式がある。また、ソフトウェア配信ツールによる配信方式の一つに、配信対象マシンをネットワークブートさせてブートプログラム経由でＯＳなどのソフトウェアを配信対象マシンのディスクに直接インストールする方式もある。 One of the distribution methods using a software distribution tool is a method of communicating with a client program operating on the OS of a distribution target machine using TCP / IP, etc., transmitting software, and installing on the OS. In addition, as a distribution method using a software distribution tool, there is a method in which a distribution target machine is network booted and software such as an OS is directly installed on a disk of the distribution target machine via a boot program.
　ソフトウェアを配信するために必要なシステム要件は配信方式によって異なる。例えば、クライアントプログラムと通信する方式では、配信対象マシンにＯＳがインストールされていること、そのＯＳ上にクライアントプログラムがインストールされていることなどがシステム要件になる。一方、配信対象マシンをネットワークブートさせる配信方式では、配信対象マシンがネットワークブート可能なこと、ネットワークブート用のサーバ（配信サーバ）が配信対象マシンと同一ネットワーク内に配置されていることなどがシステム要件になる。 ∙ System requirements necessary for distributing software differ depending on the distribution method. For example, in a method of communicating with a client program, the system requirement is that an OS is installed on the distribution target machine, and that the client program is installed on the OS. On the other hand, in the distribution method in which the distribution target machine is network booted, the system requirements are that the distribution target machine can be network booted and that the network boot server (distribution server) is located in the same network as the distribution target machine. become.
　近年、大規模データセンタをネットワーク経由で使用すること（クラウドコンピューティング）が注目されてきているが、クラウドコンピューティング環境を構築及び管理する場合にもソフトウェア配信作業は必要になる。その場合、ソフトウェア配信作業を効率化するためにソフトウェア配信ツールが使用されることが想定される。しかし、クラウドコンピューティング環境ではシステムを使用するユーザ（テナント）として複数の人及び会社が混在すること、セキュリティ上の観点からテナント間のネットワークは分離させられる必要があること、などの特徴がある。 In recent years, the use of a large-scale data center via a network (cloud computing) has attracted attention, but software distribution work is also required when building and managing a cloud computing environment. In that case, it is assumed that a software distribution tool is used to make the software distribution work more efficient. However, the cloud computing environment is characterized in that a plurality of people and companies are mixed as users (tenants) using the system, and that a network between tenants needs to be separated from the viewpoint of security.
　ソフトウェア配信ツールのシステム要件は、システムのネットワーク環境にも依存する。このため、クラウドコンピューティング環境のように多数のネットワークが存在する構成の場合、集中的な構成（配信サーバを１台用意し、そこから全マシンに配信する）で全体を管理することができない。したがって、例えば、ネットワーク毎に配信サーバの機能を持ったコンポーネント（子配信サーバ）を配置するなど、ソフトウェア配信ツールの構成方法に工夫が必要になる。 The system requirements for software distribution tools also depend on the system network environment. For this reason, in the case of a configuration in which a large number of networks exist as in a cloud computing environment, the whole cannot be managed with a centralized configuration (one distribution server is prepared and distributed to all machines from there). Therefore, for example, it is necessary to devise a method for configuring the software distribution tool, such as arranging a component (child distribution server) having a distribution server function for each network.
　一方、配信するＯＳ、パッチ、アプリケーションなどのソフトウェアそのものについては全体で一元的に管理されていればよい。しかし、前述のような配信サーバ機能を持ったコンポーネントの配置の都合上、実際に配信する際には配信対象マシンと同一ネットワーク内に存在する配信サーバ機能を持ったコンポーネント（子配信サーバ）に配信するソフトウェアをコピーしておくなど、ソフトウェアの配置場所を考慮する必要がある。 On the other hand, the software itself such as the OS, patches, and applications to be distributed only needs to be managed centrally. However, due to the placement of components with the distribution server function as described above, when actually distributing, it is distributed to the component (child distribution server) with the distribution server function that exists in the same network as the distribution target machine It is necessary to consider the location of the software, such as copying the software to be used.
　図１は、ソフトウェアの配置方法の一例を説明するための図である。図１に示すソフトウェアの配置方法では、大元となるソフトウェアは親配信サーバ５１００に配置しておき、配信対象マシン６１００にソフトウェアを配信する必要が生じたタイミングで、親配信サーバ５１００から子配信サーバ５２００にソフトウェアをコピーする。但し、図１に示すようなソフトウェアの配置方法では、必要に応じて逐次ソフトウェアをコピーするため配信速度が遅くなるという問題がある。 FIG. 1 is a diagram for explaining an example of a software arrangement method. In the software arrangement method shown in FIG. 1, the main software is arranged in the parent distribution server 5100 and the parent distribution server 5100 to the child distribution server at the timing when the software needs to be distributed to the distribution target machine 6100. Copy software to 5200. However, the software arrangement method as shown in FIG. 1 has a problem in that the distribution speed is slow because software is sequentially copied as necessary.
　図１に示したソフトウェアの配置方法以外の例として、親配信サーバだけでなく子配信サーバにも全ソフトウェアのコピーを配置する方法などがある。但し、このような配置方法をとった場合、子配信サーバにも全ソフトウェアを格納する容量が必要になるためストレージコストがかかってしまうという問題がある。 As an example other than the software arrangement method shown in FIG. 1, there is a method of arranging copies of all software not only on the parent distribution server but also on the child distribution server. However, when such an arrangement method is adopted, there is a problem in that a storage cost is required because the child distribution server also needs a capacity for storing all the software.
　そのため、全ソフトウェアを子配信サーバに格納するストレージコストをかけられないが、配信実行速度は可能な限り向上させたい場合、必要に応じてソフトウェアの子配信サーバへのコピー処理を予め予測実行しておくことが必要となる。しかしながらソフトウェア配信作業時のソフトウェアコピーの予測実行のような、システム管理者が実行する複数のオペレーションで構成されるフローのようにプロセスが予め定義されていない処理を予測実行することが可能な技術は存在しなかった。 For this reason, the storage cost of storing all software in the child distribution server cannot be applied, but if the distribution execution speed is to be improved as much as possible, the copy processing of the software to the child distribution server is predicted and executed as necessary. It is necessary to keep it. However, a technology capable of predicting and executing a process whose process is not defined in advance, such as a flow including a plurality of operations executed by a system administrator, such as predictive execution of software copy at the time of software distribution work Did not exist.
　本発明によれば、予めプロセスが定義されていない未知のフローに対しても、精度良く予測実行を行うことが可能となり、上述の問題が解決される。 According to the present invention, it is possible to perform predictive execution with high accuracy even for an unknown flow whose process is not defined in advance, and the above-described problem is solved.
　図２は、第１実施形態に係るプロセス予測実行システムの構成を表した図である。図２に示すように、プロセス予測実行システムは、親配信サーバ１１００、子配信サーバ１２００～子配信サーバ１３００、配信対象マシン２１００～配信対象マシン２５００、ネットワーク機器３１００～ネットワーク機器３２００並びに管理用ネットワーク４１００～管理用ネットワーク４３００を含む。 FIG. 2 is a diagram showing the configuration of the process prediction execution system according to the first embodiment. As shown in FIG. 2, the process prediction execution system includes a parent distribution server 1100, a child distribution server 1200 to a child distribution server 1300, a distribution target machine 2100 to a distribution target machine 2500, a network device 3100 to a network device 3200, and a management network 4100. A management network 4300.
　親配信サーバ１１００は、管理用ネットワーク４１００～管理用ネットワーク４３００を経由して、子配信サーバ１２００～子配信サーバ１３００を管理及び制御するソフトウェアを搭載したハードウェアである。 The parent distribution server 1100 is hardware equipped with software for managing and controlling the child distribution server 1200 to the child distribution server 1300 via the management network 4100 to the management network 4300.
　子配信サーバ１２００は、管理用ネットワーク４２００を経由して、配信対象マシン２１００～配信対象マシン２３００を管理及び制御するソフトウェアを搭載したハードウェアである。 The child distribution server 1200 is hardware equipped with software for managing and controlling the distribution target machine 2100 to the distribution target machine 2300 via the management network 4200.
　子配信サーバ１３００は、管理用ネットワーク４３００を経由して、配信対象マシン２４００～配信対象マシン２５００を管理・制御するソフトウェアを搭載したハードウェアである。 The child distribution server 1300 is hardware equipped with software for managing and controlling the distribution target machine 2400 to the distribution target machine 2500 via the management network 4300.
　ネットワーク機器３１００～ネットワーク機器３２００、及び管理用ネットワーク４１００～管理用ネットワーク４３００は、親配信サーバ１１００、子配信サーバ１２００～子配信サーバ１３００、配信対象マシン２１００～配信対象マシン２５００が相互に通信する場合に経由するネットワークインフラである。 The network device 3100 to the network device 3200 and the management network 4100 to the management network 4300 are configured such that the parent distribution server 1100, the child distribution server 1200 to the child distribution server 1300, and the distribution target machine 2100 to the distribution target machine 2500 communicate with each other. It is a network infrastructure that goes through.
　なお、図２は、本実施形態を模式的に説明するための例示であって、システム構成を制限するものではない。すなわち、子配信サーバ及び配信対象マシンは、管理及び制御するシステムに応じて任意の台数として構成可能である。 In addition, FIG. 2 is an illustration for schematically explaining the present embodiment, and does not limit the system configuration. That is, the child distribution server and the distribution target machine can be configured as an arbitrary number according to the system to be managed and controlled.
　図３は、親配信サーバ１１００内で動作する本実施形態のプログラムの論理的な構造を模式的に表した図である。図３に示すように、親配信サーバ１１００は、ワーキングエリア１１１０、イベント処理部１１２０、予測実行部１１３０、予測知識作成部１１４０及びデータ格納領域１１５０を含む。ここで、親配信サーバ１１００は、プロセス予測実行装置１１００とも言う。 FIG. 3 is a diagram schematically showing the logical structure of the program of this embodiment that operates in the parent distribution server 1100. As shown in FIG. 3, the parent distribution server 1100 includes a working area 1110, an event processing unit 1120, a prediction execution unit 1130, a prediction knowledge creation unit 1140, and a data storage area 1150. Here, the parent distribution server 1100 is also referred to as a process prediction execution apparatus 1100.
　ここで、ワーキングエリア１１１０は、メインメモリ等の揮発性の記憶領域に相当する。イベント処理部１１２０、予測実行部１１３０及び予測知識作成部１１４０は、ＣＰＵ（Ｃｅｎｔｒａｌ　Ｐｒｏｃｅｓｓｉｎｇ　Ｕｎｉｔ）等によって実行されるプログラムモジュールに相当する。データ格納領域１１５０は、磁気ディスク等の不揮発性の記憶装置、及びそこに格納されるデータを管理するＤＢＭＳ（ＤａｔａＢａｓｅ　Ｍａｎａｇｅｍｅｎｔ　Ｓｙｓｔｅｍ）に相当する。 Here, the working area 1110 corresponds to a volatile storage area such as a main memory. The event processing unit 1120, the prediction execution unit 1130, and the prediction knowledge creation unit 1140 correspond to program modules executed by a CPU (Central Processing Unit) or the like. The data storage area 1150 corresponds to a nonvolatile storage device such as a magnetic disk, and a DBMS (DataBase Management System) that manages data stored therein.
　図１８は、上述のプログラムモジュールのコードを記録（記憶）する不揮発性記録媒体１１７０の例を示すブロック図である。不揮発性記録媒体１１７０が、親配信サーバ１１００に供給され、親配信サーバ１１００のＣＰＵ（不図示）は、不揮発性記録媒体１１７０に格納されたプログラムのコードを読み出して実行するようにしてもよい。ここで、親配信サーバは、例えばコンピュータである。或いは、そのＣＰＵは、不揮発性記録媒体１１７０に格納されたプログラムのコードを、記憶部（不図示）に格納するようにしてもよい。すなわち、本実施形態は、コンピュータ（ＣＰＵ）が実行するプログラム（ソフトウェア）を記憶する不揮発性記録媒体１１７０の実施形態を含む。 FIG. 18 is a block diagram showing an example of a non-volatile recording medium 1170 that records (stores) the code of the above-described program module. The non-volatile recording medium 1170 may be supplied to the parent distribution server 1100, and the CPU (not shown) of the parent distribution server 1100 may read and execute the program code stored in the non-volatile recording medium 1170. Here, the parent distribution server is, for example, a computer. Alternatively, the CPU may store the code of the program stored in the nonvolatile recording medium 1170 in a storage unit (not shown). That is, the present embodiment includes an embodiment of a nonvolatile recording medium 1170 that stores a program (software) executed by a computer (CPU).
　ワーキングエリア１１１０には、「過去に発生したイベントを記憶しておくための過去のイベント」、「検出したイベントを記憶しておくためのカレントイベント」、並びに「イベント処理部１１２０、予測実行部１１３０及び予測知識作成部１１４０の処理における一時的なデータを格納するためのキュー」が格納される。ここで、キューとは、最初に格納した（エンキューした）データが最初に取り出される（デキューされる）特徴を持った記憶領域である。 The working area 1110 includes “past events for storing past events”, “current events for storing detected events”, and “event processing unit 1120, prediction execution unit 1130”. In addition, a queue for storing temporary data in the process of the prediction knowledge creating unit 1140 is stored. Here, the queue is a storage area having a feature that first stored (enqueued) data is first extracted (dequeued).
　イベント処理部１１２０は、内部にイベント連続性判断部１１２１を含む。イベント処理部１１２０は、システム管理者によって実行された管理操作（オペレーション）をイベントとして検出した時に、そのイベントから予測されるオペレーションの予測実行を予測実行部１１３０に依頼する。イベント処理部１１２０は、イベント連続性判断部１１２１を使用して、過去のイベントと検出したイベントとに連続性があるかどうかを判断する。次に、連続性があると判断した場合、イベント処理部１１２０は、予測知識作成部１１４０に予測知識の作成を依頼する。イベント処理部１１２０は、以上の動作を実行するプログラムモジュールである。 The event processing unit 1120 includes an event continuity determination unit 1121 inside. When the event processing unit 1120 detects a management operation (operation) executed by the system administrator as an event, the event processing unit 1120 requests the prediction execution unit 1130 to perform prediction execution of an operation predicted from the event. The event processing unit 1120 uses the event continuity determination unit 1121 to determine whether there is continuity between past events and detected events. Next, when it is determined that there is continuity, the event processing unit 1120 requests the prediction knowledge creation unit 1140 to create prediction knowledge. The event processing unit 1120 is a program module that executes the above operations.
　予測実行部１１３０は、イベント処理部１１２０からの依頼に対して、後述する予測知識格納部１１５２を参照しながら、イベント処理部１１２０が検出したイベントから予測されるオペレーションを予測実行するプログラムモジュールである。 The prediction execution unit 1130 is a program module that predicts and executes an operation predicted from an event detected by the event processing unit 1120 with reference to a prediction knowledge storage unit 1152 described later in response to a request from the event processing unit 1120. .
　予測知識作成部１１４０は、内部にパラメータ変数化部１１４１を含み、イベント処理部１１２０からの依頼に対して、予測知識格納部１１５２内に予測知識を作成するプログラムモジュールである。 The prediction knowledge creation unit 1140 is a program module that includes a parameter variable conversion unit 1141 therein and creates prediction knowledge in the prediction knowledge storage unit 1152 in response to a request from the event processing unit 1120.
　データ格納領域１１５０は、オペレーション定義格納部１１５１及び予測知識格納部１１５２を含む。 The data storage area 1150 includes an operation definition storage unit 1151 and a prediction knowledge storage unit 1152.
　オペレーション定義格納部１１５１には、オペレーション定義、プロセス定義及びオペレーションフロー定義が格納される。 The operation definition storage unit 1151 stores an operation definition, a process definition, and an operation flow definition.
　オペレーション定義とは、システム管理者によって行われる管理操作を定義したものである。図４は、オペレーション定義の例を示す図である。図４の上部にツリー構造で示すように、オペレーションには、マシン登録、イメージ登録、イメージ配信等がある。また、図４のツリー構造の下に示すように、オペレーション定義には、オペレーションの種類に相当するクラス、オペレーションの引数に相当する属性などが定義される。 Operation definition defines management operations performed by the system administrator. FIG. 4 is a diagram illustrating an example of operation definition. As shown in a tree structure at the top of FIG. 4, the operations include machine registration, image registration, image distribution, and the like. As shown below the tree structure in FIG. 4, the operation definition defines a class corresponding to the type of operation, an attribute corresponding to an argument of the operation, and the like.
　プロセス定義とは、オペレーションが行われた時に実際に実行される処理を定義したものである。図５は、プロセス定義の例を示す図である。図５の上部にツリー構造で示すように、プロセスには、マシン登録処理、イメージ登録処理、イメージ配信処理、イメージ送信処理がある。また、図５のツリー構造の下に示すように、プロセス定義には、プロセスの種類に相当するクラス、内部データに相当する属性などが定義される。なお、図５に示した以外に、プロセス定義には、プロセスの引数や、プロセスの実際の処理に相当するメソッドなどが定義されていても良い。ここで、各プロセスは、各プロセスの属性「予測実行対象」の値として「真」を定義した場合は予測実行対象となり、「偽」を定義した場合は予測実行対象とはならない。「予測実行対象」の値は予めユーザにより定義されていても良い。 Process definition defines the processing that is actually executed when an operation is performed. FIG. 5 is a diagram illustrating an example of a process definition. As shown in the tree structure at the top of FIG. 5, the processes include machine registration processing, image registration processing, image distribution processing, and image transmission processing. Further, as shown below the tree structure in FIG. 5, the process definition defines a class corresponding to the type of process, an attribute corresponding to internal data, and the like. In addition to the processes shown in FIG. 5, the process definition may define process arguments, methods corresponding to the actual processing of the process, and the like. Here, each process becomes a prediction execution target when “true” is defined as the value of the attribute “prediction execution target” of each process, and does not become a prediction execution target when “false” is defined. The value of “prediction execution target” may be defined in advance by the user.
　オペレーションフロー定義とは、各オペレーションがどのようなプロセスで構成されるのかを定義したものである。図６は、オペレーションフロー定義の例を示す図である。図６に示すように、プロセスを要素としたフローで定義される。 Operation flow definition defines what process each operation is composed of. FIG. 6 is a diagram illustrating an example of an operation flow definition. As shown in FIG. 6, the flow is defined by a process as an element.
　予測知識格納部１１５２は、あるイベント（オペレーション）が発生した時に次に発生するであろうイベント（オペレーション）を予測するための知識を格納する。 The prediction knowledge storage unit 1152 stores knowledge for predicting an event (operation) that will occur next when a certain event (operation) occurs.
　図７は、予測知識格納部１１５２が格納するデータの例を示す図である。図７に示すように、予測知識格納部１１５２が格納するデータには、カレントイベント及びそのパラメータと、予測イベント及びそのパラメータと、実行回数とが含まれる。この予測知識は、イベント処理部１１２０が検出したイベントがカレントイベント及びそのパラメータにマッチした場合に、その予測イベント及びパラメータに対応するオペレーションが近い将来に実行される可能性があることを表現している。 FIG. 7 is a diagram illustrating an example of data stored in the prediction knowledge storage unit 1152. As shown in FIG. 7, the data stored in the prediction knowledge storage unit 1152 includes the current event and its parameters, the prediction event and its parameters, and the number of executions. This predictive knowledge expresses that when an event detected by the event processing unit 1120 matches the current event and its parameters, an operation corresponding to the predictive event and parameters may be executed in the near future. Yes.
　次に、図８～１２を参照して、第１実施形態に係るプロセス予測実行装置１１００の動作について説明する。 Next, the operation of the process prediction execution apparatus 1100 according to the first embodiment will be described with reference to FIGS.
　図８は、第１実施形態に係るプロセス予測実行装置１１００における処理の流れを示すフローチャート図である。図８に示すように、システム管理者による管理操作（オペレーション）が実行されると、イベント処理部１１２０は、オペレーション定義格納部１１５１に格納されているオペレーション定義を参照しながら、そのオペレーションをイベントとして検出する（ステップＡ１）。 FIG. 8 is a flowchart showing a process flow in the process prediction execution apparatus 1100 according to the first embodiment. As shown in FIG. 8, when a management operation (operation) by the system administrator is executed, the event processing unit 1120 refers to the operation definition stored in the operation definition storage unit 1151 as an event. Detect (Step A1).
　イベントを検出すると、イベント処理部１１２０は、そのイベントを入力として予測実行部１１３０を呼び出し、予測実行処理を実行する（ステップＡ２）。 When an event is detected, the event processing unit 1120 calls the prediction execution unit 1130 with the event as an input, and executes the prediction execution process (step A2).
　予測実行処理が終了すると、イベント処理部１１２０は、イベント連続性判断部１１２１を用いて、検出したイベントが過去のイベントと連続性があるかどうかを判断するイベント連続性判断処理を実行する（ステップＡ３）。 When the prediction execution process ends, the event processing unit 1120 uses the event continuity determination unit 1121 to execute an event continuity determination process for determining whether the detected event is continuous with a past event (step S1). A3).
　イベント連続性の判断は、例えば、前回検出したイベントを記憶しておいて、そのイベントと今回検出したイベントとの発生間隔が一定時間以内であった場合に連続していると判断するなどして判断しても良い。 The event continuity is determined by, for example, storing the event detected last time and determining that the occurrence interval between the event and the event detected this time is within a certain time. You may judge.
　検出したイベントが過去のイベントと連続性があると判断される場合（ステップＡ４のＹ）、イベント処理部１１２０は、予測知識作成部１１４０を呼び出し、予測知識作成処理を実行する（ステップＡ５）。 When it is determined that the detected event is continuous with the past event (Y in Step A4), the event processing unit 1120 calls the prediction knowledge creating unit 1140 and executes the prediction knowledge creating process (Step A5).
　検出したイベントが過去のイベントと連続性がないと判断される場合（ステップＡ４のＮ）、プロセス予測実行装置１１００はイベント処理を終了して次のイベントの発生を待つ。 If it is determined that the detected event is not continuous with the past event (N in Step A4), the process prediction execution apparatus 1100 ends the event processing and waits for the next event to occur.
　図９は、予測実行部１１３０が実行する予測実行処理の流れを示すフローチャート図である。図９に示すように、予測実行部１１３０は、入力されたイベントをカレントイベントとしてワーキングエリア１１１０に格納する（ステップＢ１）。 FIG. 9 is a flowchart showing the flow of the prediction execution process executed by the prediction execution unit 1130. As shown in FIG. 9, the prediction execution unit 1130 stores the input event as a current event in the working area 1110 (step B1).
　カレントイベントが予測知識格納部１１５２に格納されている予測知識のカレントイベントとマッチする場合（ステップＢ２のＹ）、予測実行部１１３０は、そのエントリの予測イベントに対応するオペレーションを次に実行される可能性の高いオペレーションとみなして実行する（ステップＢ３）。 When the current event matches the current event of the prediction knowledge stored in the prediction knowledge storage unit 1152 (Y in Step B2), the prediction execution unit 1130 next executes the operation corresponding to the prediction event of the entry. The operation is considered as a highly likely operation (step B3).
　この場合、複数のエントリがマッチする場合、予測実行部１１３０は最も実行回数が高いものをマッチすると判断するなどして、マッチするエントリを特定する。また、実際に実行される処理は、予測実行対象として定義されているプロセス（図５の「イメージ送信処理」、図６の「イメージ配信」オペレーションの「イメージ送信処理」等）のみである。 In this case, when a plurality of entries match, the prediction execution unit 1130 identifies a matching entry by determining that the one with the highest number of executions is matched. In addition, the processes that are actually executed are only processes defined as prediction execution targets (“image transmission process” in FIG. 5, “image transmission process” in the “image distribution” operation in FIG. 6, etc.).
　カレントイベントが予測知識格納部１１５２に格納されている予測知識のカレントイベントとマッチしない場合（ステップＢ２のＮ）、予測実行部１１３０は、予測されるオペレーションがないと判断して処理を終了する。 If the current event does not match the current event of the predicted knowledge stored in the predicted knowledge storage unit 1152 (N in Step B2), the prediction execution unit 1130 determines that there is no predicted operation and ends the process.
　予測実行を行った後、さらに予測実行を続ける場合（ステップＢ４のＹ）、予測実行部１１３０は、予測実行したオペレーションのイベントをワーキングエリア１１１０のカレントイベントに格納して、ステップＢ２からの処理を繰り返す。 When the prediction execution is further continued after performing the prediction execution (Y in Step B4), the prediction execution unit 1130 stores the event of the operation that has been predicted execution in the current event of the working area 1110, and performs the processing from Step B2. repeat.
　予測実行をやめる場合（ステップＢ４のＮ）、予測実行部１１３０は、予測実行処理を終了する。なお、予測実行を続けるかどうかの判断は、例えば、予測実行回数を設け、予測実行処理の回数を制御しても良い。 When the prediction execution is stopped (N in Step B4), the prediction execution unit 1130 ends the prediction execution process. The determination as to whether or not to continue the prediction execution may be performed by, for example, providing the number of prediction executions and controlling the number of prediction execution processes.
　図１０は、イベント連続性判断部１１２１が実行するイベント連続性判断処理の流れを示すフローチャート図である。図１０に示すように、イベント連続性判断部１１２１は、過去のイベントと検出したイベントとの時間間隔が所定の時間以内か否かを判定する（ステップＣ１）。 FIG. 10 is a flowchart showing the flow of event continuity determination processing executed by the event continuity determination unit 1121. As shown in FIG. 10, the event continuity determination unit 1121 determines whether or not the time interval between the past event and the detected event is within a predetermined time (step C1).
　時間間隔が所定の時間以内だと判定すると（ステップＣ１のＹ）、イベント連続性判断部１１２１はイベントが連続していると判断する（ステップＣ２）。 If it is determined that the time interval is within the predetermined time (Y in Step C1), the event continuity determination unit 1121 determines that the events are continuous (Step C2).
　時間間隔が所定の時間以内ではないと判定すると（ステップＣ１のＮ）、イベント連続性判断部１１２１はイベントが連続していないと判断する（ステップＣ３）。 If it is determined that the time interval is not within the predetermined time (N in Step C1), the event continuity determination unit 1121 determines that the events are not continuous (Step C3).
　図１０に示したイベントの連続性の判断方法は一例であり、これに限定されない。イベントの連続性を判断するのに妥当な方法であれば、イベント連続性判断部１１２１は他のいかなる方法によってイベントの連続性を判断しても良い。 The determination method of event continuity shown in FIG. 10 is an example, and the present invention is not limited to this. The event continuity determination unit 1121 may determine event continuity by any other method as long as it is a reasonable method for determining event continuity.
　図１１は、予測知識作成部１１４０が実行する予測知識作成処理の流れを示すフローチャート図である。図１１に示すように、予測知識作成部１１４０は、過去のイベントと検出したイベントとの組み合わせをワーキングエリア１１１０のキューにエンキューする（ステップＤ１）。 FIG. 11 is a flowchart showing a flow of predicted knowledge creation processing executed by the predicted knowledge creation unit 1140. As illustrated in FIG. 11, the prediction knowledge creating unit 1140 enqueues a combination of a past event and a detected event into a queue in the working area 1110 (Step D1).
　予測知識作成部１１４０は、パラメータ変数化部１１４１を呼び出す。これにより、パラメータ変数化部１１４１はパラメータ変数化処理を実行する（ステップＤ２）。 The prediction knowledge creation unit 1140 calls the parameter variable conversion unit 1141. Thereby, the parameter variable conversion unit 1141 executes the parameter variable conversion process (step D2).
　パラメータ変数化処理後、予測知識作成部１１４０は、過去のイベントと検出したイベントとの組み合わせをワーキングエリア１１１０のキューからデキューする（ステップＤ３）。 After the parameter variable processing, the prediction knowledge creating unit 1140 dequeues the combination of the past event and the detected event from the queue in the working area 1110 (Step D3).
　予測知識作成部１１４０は、デキューした組み合わせを予測知識と比較して一致するエントリがあれば（ステップＤ４のＹ）、そのエントリの実行回数をインクリメントする（ステップＤ５）。一致しない場合（ステップＤ４のＮ）、予測知識作成部１１４０は、新規知識として予測知識内に登録する（ステップＤ６）。予測知識作成部１１４０は、これらの処理をキューのエントリの数だけ繰り返す（ステップＤ７）。 The predicted knowledge creating unit 1140 compares the dequeued combination with the predicted knowledge and, if there is a matching entry (Y in step D4), increments the number of times the entry is executed (step D5). If they do not match (N in step D4), the predicted knowledge creating unit 1140 registers the new knowledge in the predicted knowledge (step D6). The prediction knowledge creating unit 1140 repeats these processes as many times as the number of queue entries (step D7).
　図１２は、パラメータ変数化部１１４１が実行するパラメータ変数化処理の流れを示すフローチャート図である。図１２に示すように、パラメータ変数化部１１４１は、過去のイベントと検出したイベントとの組み合わせについて、それらが予測知識内のエントリのクラス（カレントイベント、予測イベント）と一致するかどうかの確認をする（ステップＥ１）。 FIG. 12 is a flowchart showing the flow of the parameter variable process executed by the parameter variable unit 1141. As shown in FIG. 12, the parameter variable converting unit 1141 confirms whether or not the combination of the past event and the detected event matches the class of the entry (current event, predicted event) in the predicted knowledge. (Step E1).
　クラスが一致する場合（ステップＥ１のＹ）、過去のイベントのパラメータが予測知識内のカレントイベントのパラメータの値と異なれば（ステップＥ２のＮ）、パラメータ変数化部１１４１は、一致しないパラメータを変数化した過去のイベントと検出したイベントとの組み合わせをワーキングエリア１１１０のキューにエンキューする（ステップＥ３）。 If the classes match (Y in step E1), if the parameter of the past event is different from the parameter value of the current event in the prediction knowledge (N in step E2), the parameter variableizing unit 1141 changes the parameter that does not match the variable A combination of the converted past event and the detected event is enqueued in the queue of the working area 1110 (step E3).
　この場合、パラメータ変数化部１１４１は、パラメータが全て変数化されてしまった組み合わせ、過去のイベントと検出したイベントとが同じになってしまった組み合わせなど、予測知識として効果が出ないものについては除外するようにしても良い。 In this case, the parameter variableizing unit 1141 excludes combinations that have not been effective as predictive knowledge, such as combinations in which all parameters have been converted into variables, and combinations in which past events are the same as detected events. You may make it do.
　クラスが一致しない場合（ステップＥ１のＮ）、又は、過去のイベントが予測知識内のカレントイベント及びそのパラメータと完全に一致する場合（ステップＥ２のＹ）、パラメータ変数化部１１４１は、次のエントリとの比較処理を行う（ステップＥ４）。 If the classes do not match (N in Step E1), or if the past event completely matches the current event and its parameters in the prediction knowledge (Y in Step E2), the parameter variableizing unit 1141 enters the next entry Is compared (step E4).
　プロセス予測実行装置１１００は、図８～図１２を参照して説明した上記の処理を繰り返すことにより、システム環境、運用環境に適した予測知識を作成するとともに、作成した知識に従って予測実行を行う。 The process prediction execution apparatus 1100 creates prediction knowledge suitable for the system environment and the operation environment by repeating the above-described processing described with reference to FIGS. 8 to 12, and performs prediction execution according to the created knowledge.
　次に、図１３～１６を参照して、第１実施形態に係るプロセス予測実行装置１１００の動作を具体的に説明する。 Next, the operation of the process prediction execution apparatus 1100 according to the first embodiment will be specifically described with reference to FIGS.
　前提として、配布対象マシンとしてｍａｃｈｉｎｅ＿１が登録されており、図１３のように予測知識がない状態から開始されるとする。また、イベント連続性判断部１１２１の動作として、前回検出したイベントと今回検出したイベントとの発生間隔が一定時間（「一定時間ＴＩ」と呼ぶ）以内であった場合にイベントが連続していると判断するとする。 As a premise, it is assumed that machine_1 is registered as a distribution target machine and starts from a state where there is no predictive knowledge as shown in FIG. Further, as an operation of the event continuity determination unit 1121, if events occur continuously when the occurrence interval between the event detected last time and the event detected this time is within a certain time (referred to as “certain time TI”). Suppose you decide.
　ここでは、典型的な例として、次に示す順序でシステム管理操作が行われた場合の動作を説明する。
１．一定時間ＴＩ以上経過後、イメージ登録（ｐａｔｃｈ＿ｉｍａｇｅ＿１）
２．一定時間ＴＩ以内に、イメージ配信（ｍａｃｈｉｎｅ＿１，ｐａｔｃｈ＿ｉｍａｇｅ＿１）
３．一定時間ＴＩ以上経過後、イメージ登録（ｐａｔｃｈ＿ｉｍａｇｅ＿２）
４．一定時間ＴＩ以内に、イメージ配信（ｍａｃｈｉｎｅ＿１，ｐａｔｃｈ＿ｉｍａｇｅ＿２）
５．一定時間ＴＩ以上経過後、イメージ登録（ｐａｔｃｈ＿ｉｍａｇｅ＿３）
６．一定時間ＴＩ以内に、イメージ配信（ｍａｃｈｉｎｅ＿１，ｐａｔｃｈ＿ｉｍａｇｅ＿３）
　（１．のイベントが発生）
　システム管理者によってイメージ「ｐａｔｃｈ＿ｉｍａｇｅ＿１」が登録されると、イベント処理部１１２０は、イベント「イメージ登録（ｐａｔｃｈ＿ｉｍａｇｅ＿１）」を検出する（図８のステップＡ１）。 Here, as a typical example, an operation when a system management operation is performed in the following order will be described.
1. Image registration (patch_image_1) after elapse of a certain time TI
2. Image distribution (machine_1, patch_image_1) within a certain time TI
3. Image registration (patch_image_2) after elapse of a certain time TI
4). Image distribution (machine_1, patch_image_2) within a certain time TI
5. Image registration (patch_image_3) after elapse of a certain time TI
6). Image distribution (machine_1, patch_image_3) within a certain time TI
(The event of 1. occurs)
When the image “patch_image_1” is registered by the system administrator, the event processing unit 1120 detects the event “image registration (patch_image_1)” (step A1 in FIG. 8).
　イベントを検出すると、イベント処理部１１２０は、そのイベントを入力として予測実行部１１３０を呼び出す（図８のステップＡ２、図９）。 When an event is detected, the event processing unit 1120 calls the prediction execution unit 1130 with the event as an input (step A2 in FIG. 8, FIG. 9).
　予測実行部１１３０は、入力されたそのイベントをカレントイベントとしてワーキングエリア１１１０に格納する（図９のステップＢ１）。次に、予測実行部１１３０は、カレントイベントが予測知識格納部１１５２に格納されている予測知識のカレントイベントとマッチするか判断する（図９のステップＢ２）。予測実行部１１３０は、マッチする予測知識がないため（図１３）、予測実行処理を終了する。 The prediction execution unit 1130 stores the input event as a current event in the working area 1110 (step B1 in FIG. 9). Next, the prediction execution unit 1130 determines whether or not the current event matches the current event of the prediction knowledge stored in the prediction knowledge storage unit 1152 (step B2 in FIG. 9). Since there is no matching prediction knowledge (FIG. 13), the prediction execution unit 1130 ends the prediction execution process.
　予測実行処理が終わると、イベント処理部１１２０は、イベント連続性判断部１１２１を用いて、検出したイベントが過去のイベントと連続性があるかどうかを判断する（図８のステップＡ３、図１０）。前回のイベントは一定時間ＴＩ以上前に発生しているものであるため（図１０のステップＣ１のＮ）、イベント連続性判断部１１２１は、イベントが連続していないと判断する（図１０のステップＣ３）。 When the prediction execution process ends, the event processing unit 1120 uses the event continuity determination unit 1121 to determine whether the detected event is continuous with past events (step A3 in FIG. 8, FIG. 10). . Since the previous event occurs before the predetermined time TI (N in Step C1 in FIG. 10), the event continuity determination unit 1121 determines that the events are not continuous (Step in FIG. 10). C3).
　検出したイベントが過去のイベントと連続性がないため（図８のステップＡ４のＮ）、プロセス予測実行装置１１００は、イベント処理を終了して次のイベントを待つ。 Since the detected event is not continuous with the past event (N in step A4 in FIG. 8), the process prediction execution apparatus 1100 ends the event processing and waits for the next event.
　（２．のイベントが発生）
　次のシステム管理操作として、一定時間ＴＩ以内に、マシン「ｍａｃｈｉｎｅ＿１」に対してイメージ「ｐａｔｃｈ＿ｉｍａｇｅ＿１」が配信されると、イベント処理部１１２０は、イベント「イメージ配信（ｍａｃｈｉｎｅ＿１，ｐａｔｃｈ＿ｉｍａｇｅ＿１）」を検出する（図８のステップＡ１）。 (Event 2 occurs)
As the next system management operation, when the image “patch_image_1” is distributed to the machine “machine_1” within a predetermined time TI, the event processing unit 1120 detects the event “image distribution (machine_1, patch_image_1)” ( Step A1) in FIG.
　イベントを検出すると、イベント処理部１１２０は、そのイベントを入力として予測実行部１１３０を呼び出す（図８のステップＡ２、図９）。予測知識格納部１１５２には予測知識がないため（図１３）、予測実行部１１３０は、予測実行処理を終了する。 When an event is detected, the event processing unit 1120 calls the prediction execution unit 1130 with the event as an input (step A2 in FIG. 8, FIG. 9). Since there is no prediction knowledge in the prediction knowledge storage unit 1152 (FIG. 13), the prediction execution unit 1130 ends the prediction execution process.
　予測実行処理が終わると、イベント処理部１１２０は、イベント連続性判断部１１２１を用いて、検出したイベントが過去のイベントと連続性があるかどうかを判断する（図８のステップＡ３、図１０）。前回のイベント「イメージ登録（ｐａｔｃｈ＿ｉｍａｇｅ＿１）」が一定時間ＴＩ以内に発生しているため（図１０のステップＣ１のＹ）、イベント連続性判断部１１２１は、イベントが連続していると判断する（図１０のステップＣ２）。 When the prediction execution process ends, the event processing unit 1120 uses the event continuity determination unit 1121 to determine whether the detected event is continuous with past events (step A3 in FIG. 8, FIG. 10). . Since the previous event “image registration (patch_image_1)” has occurred within a predetermined time TI (Y in step C1 in FIG. 10), the event continuity determination unit 1121 determines that the events are continuous (FIG. 10). 10 step C2).
　検出したイベントが過去のイベントと連続性があるため（図８のステップＡ４のＹ）、イベント処理部１１２０は、予測知識作成部１１４０を呼び出す（図８のステップＡ５、図１１）。 Since the detected event is continuous with the past event (Y in step A4 in FIG. 8), the event processing unit 1120 calls the prediction knowledge creating unit 1140 (step A5 in FIG. 8, FIG. 11).
　予測知識作成部１１４０は、前回のイベント「イメージ登録（ｐａｔｃｈ＿ｉｍａｇｅ＿１）」と検出したイベント「イメージ配信（ｍａｃｈｉｎｅ＿１，ｐａｔｃｈ＿ｉｍａｇｅ＿１）」との組み合わせをワーキングエリア１１１０のキューにエンキューする（図１１のステップＤ１）。次に、予測知識作成部１１４０は、パラメータ変数化部１１４１を呼び出す（図１１のステップＤ２、図１２）。 The prediction knowledge creation unit 1140 enqueues the combination of the previous event “image registration (patch_image_1)” and the detected event “image distribution (machine_1, patch_image_1)” in the queue of the working area 1110 (step D1 in FIG. 11). Next, the prediction knowledge creation unit 1140 calls the parameter variable conversion unit 1141 (step D2 in FIG. 11, FIG. 12).
　パラメータ変数化部１１４１は、前回のイベント「イメージ登録（ｐａｔｃｈ＿ｉｍａｇｅ＿１）」と検出したイベント「イメージ配信（ｍａｃｈｉｎｅ＿１，ｐａｔｃｈ＿ｉｍａｇｅ＿１）」との組み合わせについて、予測知識内のエントリとクラスの比較処理を行う（図１２のステップＥ１）。しかし、この場合、予測知識がないため（図１３）、比較結果は一致せず（図１２のステップＥ１のＮ）、パラメータ変数化部１１４１は、パラメータ変数化処理を終了する。 The parameter variable conversion unit 1141 performs the process of comparing the entry and the class in the prediction knowledge for the combination of the previous event “image registration (patch_image_1)” and the detected event “image delivery (machine_1, patch_image_1)” (FIG. 12). Step E1). However, in this case, since there is no prediction knowledge (FIG. 13), the comparison results do not match (N in step E1 in FIG. 12), and the parameter variableizing unit 1141 ends the parameter variableizing process.
　パラメータ変数化処理後、予測知識作成部１１４０は、前回のイベント「イメージ登録（ｐａｔｃｈ＿ｉｍａｇｅ＿１）」と検出したイベント「イメージ配信（ｍａｃｈｉｎｅ＿１，ｐａｔｃｈ＿ｉｍａｇｅ＿１）」との組み合わせをワーキングエリア１１１０のキューからデキューする（図１１のステップＤ３）。 After the parameter variable processing, the prediction knowledge creating unit 1140 dequeues the combination of the previous event “image registration (patch_image_1)” and the detected event “image delivery (machine_1, patch_image_1)” from the queue in the working area 1110 (FIG. 11 step D3).
　デキューした組み合わせが予測知識と一致しないため（図１１のステップＤ４のＮ）、予測知識作成部１１４０は、その組み合わせを新規知識として予測知識内に登録し（図１１のステップＤ６）、予測知識作成処理を終了する。図１４は、図１３に予測知識として新規知識が登録された状態を示す図である。 Since the dequeued combination does not match the predicted knowledge (N in step D4 in FIG. 11), the predicted knowledge creating unit 1140 registers the combination as new knowledge in the predicted knowledge (step D6 in FIG. 11), and creates the predicted knowledge. End the process. FIG. 14 is a diagram illustrating a state in which new knowledge is registered as prediction knowledge in FIG. 13.
　予測知識作成処理終了後、イベント処理を終了して次のイベントを待つ。 Ending the prediction knowledge creation process, end the event process and wait for the next event.
　（３．のイベントが発生）
　次のシステム管理操作として、一定時間ＴＩ以上経過後、イメージ「ｐａｔｃｈ＿ｉｍａｇｅ＿２」が登録されると、イベント処理部１１２０は、イベント「イメージ登録（ｐａｔｃｈ＿ｉｍａｇｅ＿２）」を検出する（図８のステップＡ１）。検出したイベントは予測知識のカレントイベントとマッチしないため、予測実行部１１３０は、予測実行処理を行わない（図８のステップＡ２）。次に、前回のイベントから一定時間ＴＩ以上経過しているため、イベント連続性判断部１１２１は、連続イベントと判断せずに（図８のステップＡ４のＮ）、プロセス予測実行装置１１００は、イベント処理を終了して次のイベントを待つ。 (The event of 3. occurs)
As the next system management operation, when the image “patch_image_2” is registered after the elapse of a predetermined time TI, the event processing unit 1120 detects the event “image registration (patch_image_2)” (step A1 in FIG. 8). Since the detected event does not match the current event of the prediction knowledge, the prediction execution unit 1130 does not perform the prediction execution process (step A2 in FIG. 8). Next, since the predetermined time TI has passed since the previous event, the event continuity determination unit 1121 does not determine that the event is a continuous event (N in Step A4 in FIG. 8), and the process prediction execution apparatus 1100 End the process and wait for the next event.
　（４．のイベントが発生）
　次のシステム管理操作として、一定時間ＴＩ以内に、マシン「ｍａｃｈｉｎｅ＿１」に対してイメージ「ｐａｔｃｈ＿ｉｍａｇｅ＿２」が配信されると、イベント処理部１１２０は、イベント「イメージ配信（ｍａｃｈｉｎｅ＿１，ｐａｔｃｈ＿ｉｍａｇｅ＿２）」を検出する（図８のステップＡ１）。検出したイベントは予測知識のカレントイベントとマッチしないため（図１４）、予測実行部１１３０は、予測実行処理を行わない（図８のステップＡ２）。 (Event 4 occurs)
As the next system management operation, when the image “patch_image_2” is distributed to the machine “machine_1” within a predetermined time TI, the event processing unit 1120 detects the event “image distribution (machine_1, patch_image_2)” ( Step A1) in FIG. Since the detected event does not match the current event of the prediction knowledge (FIG. 14), the prediction execution unit 1130 does not perform the prediction execution process (step A2 in FIG. 8).
　予測実行処理が終わると、イベント処理部１１２０は、イベント連続性判断部１１２１を用いて、検出したイベントが過去のイベントと連続性があるかどうかを判断する（図８のステップＡ３、図１０）。前回のイベント「イメージ登録（ｐａｔｃｈ＿ｉｍａｇｅ＿２）」が一定時間ＴＩ以内に発生しているため（図１０のステップＣ１のＹ）、イベント連続性判断部１１２１は、イベントが連続していると判断し（図１０のステップＣ２、図８のステップＡ４のＹ）、イベント処理部１１２０は、予測知識作成部１１４０を呼び出す（図８のステップＡ５）。 When the prediction execution process ends, the event processing unit 1120 uses the event continuity determination unit 1121 to determine whether the detected event is continuous with past events (step A3 in FIG. 8, FIG. 10). . Since the previous event “image registration (patch_image — 2)” has occurred within a certain time TI (Y in step C1 in FIG. 10), the event continuity determination unit 1121 determines that the events are continuous (FIG. The event processing unit 1120 calls the prediction knowledge creating unit 1140 (step A5 in FIG. 8).
　予測知識作成部１１４０は、前回のイベント「イメージ登録（ｐａｔｃｈ＿ｉｍａｇｅ＿２）」と検出したイベント「イメージ配信（ｍａｃｈｉｎｅ＿１，ｐａｔｃｈ＿ｉｍａｇｅ＿２）」との組み合わせをワーキングエリア１１１０のキューにエンキューし（図１１のステップＤ１）、パラメータ変数化部１１４１を呼び出す（図１１のステップＤ２、図１２）。 The prediction knowledge creating unit 1140 enqueues the combination of the previous event “image registration (patch_image_2)” and the detected event “image delivery (machine_1, patch_image_2)” in the queue of the working area 1110 (step D1 in FIG. 11). The parameter variableizing unit 1141 is called (step D2 in FIG. 11, FIG. 12).
　パラメータ変数化部１１４１は、前回のイベント「イメージ登録（ｐａｔｃｈ＿ｉｍａｇｅ＿２）」と検出したイベント「イメージ配信（ｍａｃｈｉｎｅ＿１，ｐａｔｃｈ＿ｉｍａｇｅ＿２）」との組み合わせについて、予測知識内のエントリとクラスの比較処理を行う（図１２のステップＥ１）。その比較処理の結果、「イメージ登録」及び「イメージ配信」でクラスが一致し（図１２のステップＥ１のＹ）、一致しないパラメータが存在するため（図１２のステップＥ２のＮ）、パラメータ変数化部１１４１は、一致しないパラメータを変数化した組み合わせの「イメージ登録（変数）」「イメージ配信（ｍａｃｈｉｎｅ＿１，変数）」をキューにエンキューする（図１２のステップＥ３）。 The parameter variable conversion unit 1141 performs a process of comparing the entry and the class in the prediction knowledge for the combination of the previous event “image registration (patch_image_2)” and the detected event “image delivery (machine_1, patch_image_2)” (FIG. 12). Step E1). As a result of the comparison processing, the classes match in “image registration” and “image distribution” (Y in step E1 in FIG. 12), and there are parameters that do not match (N in step E2 in FIG. 12). The unit 1141 enqueues “image registration (variable)” and “image distribution (machine_1, variable)”, which are combinations of the parameters that do not match, into a queue (step E3 in FIG. 12).
　予測知識内で比較していないエントリは存在しないため（図１２のステップＥ４のＮ）、パラメータ変数化部１１４１は、パラメータ変数化処理を終了する。 Since there is no entry that is not compared in the prediction knowledge (N in step E4 in FIG. 12), the parameter variable conversion unit 1141 ends the parameter variable conversion process.
　パラメータ変数化処理後、予測知識作成部１１４０は、前回のイベント「イメージ登録（ｐａｔｃｈ＿ｉｍａｇｅ＿２）」と検出したイベント「イメージ配信（ｍａｃｈｉｎｅ＿１，ｐａｔｃｈ＿ｉｍａｇｅ＿２）」との組み合わせをワーキングエリア１１１０のキューからデキューする（図１１のステップＤ３）。デキューした組み合わせが予測知識と一致しないため（図１１のステップＤ４のＮ）、予測知識作成部１１４０は新規知識として予測知識内に登録する（図１１のステップＤ６）。図１５は、図１４に予測知識として新規知識が登録された状態を示す図である。 After the parameter variable processing, the prediction knowledge creating unit 1140 dequeues the combination of the previous event “image registration (patch_image_2)” and the detected event “image delivery (machine_1, patch_image_2)” from the queue in the working area 1110 (FIG. 11 step D3). Since the dequeued combination does not match the predicted knowledge (N in step D4 in FIG. 11), the predicted knowledge creating unit 1140 registers the new knowledge in the predicted knowledge (step D6 in FIG. 11). FIG. 15 is a diagram illustrating a state in which new knowledge is registered as prediction knowledge in FIG. 14.
　キュー内にまだエントリが存在するため（図１１のステップＤ７のＹ）、予測知識作成部１１４０は、キューから「イメージ登録（変数）」「イメージ配信（ｍａｃｈｉｎｅ＿１，変数）」の組み合わせをデキューする（図１１のステップＤ３）。デキューした組み合わせが予測知識と一致しないため（図１１のステップＤ４のＮ）、予測知識作成部１１４０は、新規知識として予測知識内に登録する（図１１のステップＤ６）。図１６は、図１５に予測知識として新規知識が登録された状態を示す図である。 Since there are still entries in the queue (Y in step D7 in FIG. 11), the prediction knowledge creating unit 1140 dequeues the combination of “image registration (variable)” and “image distribution (machine_1, variable)” from the queue ( Step D3 in FIG. Since the dequeued combination does not match the predicted knowledge (N in step D4 in FIG. 11), the predicted knowledge creating unit 1140 registers the new knowledge in the predicted knowledge (step D6 in FIG. 11). FIG. 16 is a diagram illustrating a state in which new knowledge is registered as prediction knowledge in FIG. 15.
　予測知識作成処理終了後、プロセス予測実行装置１１００は、イベント処理を終了して次のイベントを待つ。 After completion of the prediction knowledge creation process, the process prediction execution apparatus 1100 ends the event process and waits for the next event.
　（５．のイベントが発生）
　次のシステム管理操作として、一定時間ＴＩ以上経過後、イメージ「ｐａｔｃｈ＿ｉｍａｇｅ＿３」が登録されると、イベント処理部１１２０は、イベント「イメージ登録（ｐａｔｃｈ＿ｉｍａｇｅ＿３）」を検出する（図８のステップＡ１）。 (Event 5 occurs)
As the next system management operation, when the image “patch_image — 3” is registered after elapse of a certain time TI or more, the event processing unit 1120 detects the event “image registration (patch_image — 3)” (step A1 in FIG. 8).
　イベントを検出すると、イベント処理部１１２０は、そのイベントを入力として予測実行部１１３０を呼び出す（図８のステップＡ２、図９）。予測実行部１１３０は、入力されたイベントをカレントイベントとしてワーキングエリア１１１０に格納する（図９のステップＢ１）。次に、予測実行部１１３０は、カレントイベントが予測知識格納部１１５２に格納されている予測知識のカレントイベントとマッチするか判断する（図９のステップＢ２）。予測知識内のカレントイベント「イメージ登録（変数）」とマッチするため、予測実行部１１３０は、同一エントリ内の予測イベント「イメージ配信（ｍａｃｈｉｎｅ＿１，（変数））」を予測実行する（図９のステップＢ２のＹ、Ｂ３）。 When an event is detected, the event processing unit 1120 calls the prediction execution unit 1130 with the event as an input (step A2 in FIG. 8, FIG. 9). The prediction execution unit 1130 stores the input event as a current event in the working area 1110 (step B1 in FIG. 9). Next, the prediction execution unit 1130 determines whether or not the current event matches the current event of the prediction knowledge stored in the prediction knowledge storage unit 1152 (step B2 in FIG. 9). In order to match the current event “image registration (variable)” in the prediction knowledge, the prediction execution unit 1130 predictively executes the prediction event “image distribution (machine_1, (variable))” in the same entry (step in FIG. 9). Y of B2, B3).
　この場合、図５、図６に定義されているように、イメージ配信オペレーション内の予測実行対象プロセスは「イメージ送信処理」であるため、「イメージ送信処理」のみ実行される。 In this case, as defined in FIGS. 5 and 6, the prediction execution target process in the image distribution operation is “image transmission processing”, and therefore only “image transmission processing” is executed.
　（６．のイベントが発生）
　次のシステム管理操作として、一定時間ＴＩ以内に、マシン「ｍａｃｈｉｎｅ＿１」に対してイメージ「ｐａｔｃｈ＿ｉｍａｇｅ＿３」が配信されると、図６のイメージ配信オペレーション内の「イメージ送信処理」は既に予測実行済みのため実行されずに「イメージ配信処理」のみ実行される。そのため、本管理操作が実行開始されてから実行完了までの実行時間が短縮される。 (Event 6 occurs)
As the next system management operation, when the image “patch_image — 3” is distributed to the machine “machine — 1” within a predetermined time TI, the “image transmission process” in the image distribution operation of FIG. Only “image distribution processing” is executed without being executed. Therefore, the execution time from the start of execution of this management operation to the completion of execution is shortened.
　以上説明したように、第１実施形態におけるプロセス予測実行装置１１００によれば、予めプロセスが定義されていない未知のフローに対しても、精度良く予測実行を行うことが可能となる。その理由は、第１実施形態におけるプロセス予測実行装置１１００は、予めプロセスが定義されていない未知のフローであっても、個々のシステム毎の特別な設計コストをかけずに、予測実行のためのデータを自動的に作成することができるからである。 As described above, according to the process prediction execution apparatus 1100 in the first embodiment, it is possible to perform prediction execution with high accuracy even for an unknown flow in which no process is defined in advance. The reason is that the process prediction execution apparatus 1100 according to the first embodiment is capable of performing prediction execution without incurring a special design cost for each individual system even for an unknown flow in which no process is defined in advance. This is because data can be automatically created.
　具体的には、以下の構成をそなえるからである。第一に、イベント処理部１１２０が、オペレーションをイベントとして検出する。第二に、予測実行部１１３０をが、そのイベントを入力として、予測実行処理を実行する。第三に、イベント連続性判断部１１２１が、検出したイベントが過去のイベントと連続性があるかどうかを判断する。第四に、予測知識作成部１１４０が、予測知識作成処理を実行する。 Specifically, it has the following configuration. First, the event processing unit 1120 detects an operation as an event. Secondly, the prediction execution unit 1130 executes the prediction execution process with the event as an input. Third, the event continuity determination unit 1121 determines whether the detected event is continuous with past events. Fourth, the predicted knowledge creating unit 1140 executes a predicted knowledge creating process.
　また、第１実施形態におけるプロセス予測実行装置１１００は、自動的に精度の良い予測実行が可能な状態となる。その理由は、プロセス予測実行装置１１００が、システム管理者により連続して実行された一連のオペレーションを予測実行候補として自動抽出するとともに、繰り返し実行されたオペレーションの統計データを管理して予測精度を向上させているためである。 In addition, the process prediction execution apparatus 1100 according to the first embodiment automatically enters a state in which accurate prediction execution is possible. The reason is that the process prediction execution apparatus 1100 automatically extracts a series of operations continuously executed by the system administrator as prediction execution candidates, and manages statistical data of the repeatedly executed operations to improve prediction accuracy. It is because it is letting.
　また、第１実施形態におけるプロセス予測実行装置１１００によれば、未知の予測実行候補を作成することができる。その理由は、予測実行のための契機（トリガ）となるイベントとして、パラメータの値も一致したイベントだけでなく、パラメータを変数化して、一致したイベントもその候補としているためである。 Also, according to the process prediction execution apparatus 1100 in the first embodiment, an unknown prediction execution candidate can be created. This is because, as an event serving as a trigger (trigger) for predictive execution, not only an event whose parameter value also matches, but also a parameter that is converted into a variable and a matching event is used as the candidate.
　＜第２実施形態＞
　次に、本発明の第２実施形態におけるプロセス予測実行装置２０００について説明する。 <Second Embodiment>
Next, the process prediction execution apparatus 2000 according to the second embodiment of the present invention will be described.
　図１７は、第２実施形態におけるプロセス予測実行装置２０００の構成を示す図である。図１７に示すように、プロセス予測実行装置２０００は、予測知識格納部２０１０及び予測実行部２０２０を含む。 FIG. 17 is a diagram illustrating a configuration of the process prediction execution apparatus 2000 according to the second embodiment. As illustrated in FIG. 17, the process prediction execution device 2000 includes a prediction knowledge storage unit 2010 and a prediction execution unit 2020.
　予測知識格納部２０１０は、連続性があるイベントの組合せを、現在実行中のイベントであるカレントイベント及び当該カレントイベントから予測されるイベントである予測イベントの組合せとして格納する。 The prediction knowledge storage unit 2010 stores a combination of events having continuity as a combination of a current event that is an event currently being executed and a prediction event that is an event predicted from the current event.
　予測実行部２０２０は、ユーザの操作により発生したイベントが、前記予測知識格納手段が格納するカレントイベントと一致する場合、当該カレントイベントと組み合わされている予測イベントに対応するオペレーションを実行する。 The prediction execution unit 2020 executes an operation corresponding to the prediction event combined with the current event when the event generated by the user operation matches the current event stored in the prediction knowledge storage unit.
　以上説明したように、第２実施形態におけるプロセス予測実行装置２０００によれば、予めプロセスが定義されていない未知のフローに対しても、精度良く予測実行を行うことが可能となる。 As described above, according to the process prediction execution apparatus 2000 in the second embodiment, it is possible to perform prediction execution with high accuracy even for an unknown flow in which no process is defined in advance.
　以上、各実施形態を参照して本発明を説明したが、本発明は以上の実施形態に限定されるものではない。本発明の構成や詳細には、本発明のスコープ内で同業者が理解し得る様々な変更をすることができる。 As mentioned above, although this invention was demonstrated with reference to each embodiment, this invention is not limited to the above embodiment. Various changes that can be understood by those skilled in the art can be made to the configuration and details of the present invention within the scope of the present invention.
　なお、本発明のプログラムは、これまでに説明した各動作を、コンピュータに実行させるプログラムであれば良い。 The program of the present invention may be a program that causes a computer to execute each operation described so far.
　以上、実施形態を参照して本願発明を説明したが、本願発明は上記実施形態に限定されるものではない。本願発明の構成や詳細には、本願発明のスコープ内で当業者が理解し得る様々な変更をすることができる。 The present invention has been described above with reference to the embodiments, but the present invention is not limited to the above embodiments. Various changes that can be understood by those skilled in the art can be made to the configuration and details of the present invention within the scope of the present invention.
　この出願は、２０１２年２月２日に出願された日本出願特願２０１２－０２１１３９を基礎とする優先権を主張し、その開示の全てをここに取り込む。 This application claims priority based on Japanese Patent Application No. 2012-021139 filed on February 2, 2012, the entire disclosure of which is incorporated herein.
　本発明によれば、ある程度運用方法が決まっていて、プロセスの予測実行にメリット（運用、パフォーマンスを向上させる等）があるシステムにおいて、事前の設計コストをかけることなく自動的にプロセスを予測実行したい用途に適用できる。 According to the present invention, in a system in which an operation method is determined to some extent and there is a merit (operation, improvement in performance, etc.) in process prediction execution, it is desired to automatically execute process prediction without incurring a prior design cost. Applicable to usage.
　１１００　　親配信サーバ
　１１００　　プロセス予測実行装置
　１２００、１３００　子配信サーバ
　２１００、２３００、２４００、２５００　配信対象マシン
　３１００、３２００　ネットワーク機器
　４１００、４２００、４３００　管理用ネットワーク
　１１１０　　ワーキングエリア
　１１２０　　イベント処理部
　１１２１　　イベント連続性判断部
　１１３０　　予測実行部
　１１４０　　予測知識作成部
　１１４１　　パラメータ変数化部
　１１５０　　データ格納領域
　１１５１　　オペレーション定義格納部
　１１５２　　予測知識格納部
　１１７０　　不揮発性記録媒体
　２０００　　プロセス予測実行装置
　２０１０　　予測知識格納部
　２０２０　　予測実行部 DESCRIPTION OF SYMBOLS 1100 Parent delivery server 1100 Process prediction execution apparatus 1200, 1300 Child delivery server 2100, 2300, 2400, 2500 Delivery target machine 3100, 3200 Network equipment 4100, 4200, 4300 Management network 1110 Working area 1120 Event processing unit 1121 Event continuity judgment Unit 1130 Prediction execution unit 1140 Prediction knowledge creation unit 1141 Parameter variable conversion unit 1150 Data storage area 1151 Operation definition storage unit 1152 Prediction knowledge storage unit 1170 Non-volatile recording medium 2000 Process prediction execution device 2010 Prediction knowledge storage unit 2020 Prediction execution unit

Claims (5)

  1. 　連続性があるイベントの組合せを、現在実行中のイベントであるカレントイベント及び当該カレントイベントから予測されるイベントである予測イベントの組合せとして格納する予測知識格納手段と、
    　ユーザの操作により発生したイベントが、前記予測知識格納手段が格納する前記カレントイベントと一致する場合、当該カレントイベントと組み合わせにされている前記予測イベントに対応するオペレーションを実行する予測実行手段と、
    　を含むシステム予測実行装置。     Predictive knowledge storage means for storing a combination of events having continuity as a combination of a current event that is an event that is currently being executed and a predictive event that is an event predicted from the current event;
    When an event generated by a user operation matches the current event stored in the prediction knowledge storage unit, a prediction execution unit that executes an operation corresponding to the prediction event combined with the current event;
    A system prediction execution device including:
  2. 　前記ユーザの操作により発生したイベントと、当該イベントの発生の一つ前の前記ユーザの操作により発生したイベントとの発生時間の間隔が、所定の時間以内の場合、２つの前記イベントに連続生があると判定し、連続性があると判定された前記２つのイベントの組合せを新たに前記予測知識格納手段に格納する予測知識作成手段と、
    　をさらに含むシステム予測実行装置。     If the interval between the occurrence time of the event generated by the user operation and the event generated by the user operation immediately before the occurrence of the event is within a predetermined time, the two events are continuously generated. Predictive knowledge creating means for newly storing a combination of the two events determined to be continuous and determined to have continuity in the predictive knowledge storage means;
    A system prediction execution apparatus further including:
  3. 　前記予測知識作成手段は、前記連続性があると判定された前記２つのイベントとクラスが同一のイベントの組合せが既に前記予測知識格納手段に格納されている場合であって、前記２つのイベントのパラメータの値が、前記既に予測知識格納手段に格納されているイベントの組合せのイベントのパラメータと異なる場合に、前記パラメータを変数化して前記予測知識格納手段に格納する、
    　請求項２に記載のシステム予測実行装置。     The predictive knowledge creating means is a case where a combination of events having the same class as the two events determined to have continuity is already stored in the predictive knowledge storage means, and When the parameter value is different from the event parameter of the combination of events already stored in the prediction knowledge storage means, the parameter is converted into a variable and stored in the prediction knowledge storage means.
    The system prediction execution apparatus according to claim 2.
  4. 　連続性があるイベントの組合せを、現在実行中のイベントであるカレントイベント及び当該カレントイベントから予測されるイベントである予測イベントの組合せとして格納し、
    　ユーザの操作により発生したイベントが、前記予測知識格納手段が格納する前記カレントイベントと一致する場合、当該カレントイベントと組み合わせにされている前記予測イベントに対応するオペレーションを実行する、
    　システム予測実行方法。     A combination of events having continuity is stored as a combination of a current event that is a currently executing event and a predicted event that is an event predicted from the current event,
    When an event generated by a user operation matches the current event stored in the prediction knowledge storage unit, an operation corresponding to the prediction event combined with the current event is executed.
    System prediction execution method.
  5. 　連続性があるイベントの組合せを、現在実行中のイベントであるカレントイベント及び当該カレントイベントから予測されるイベントである予測イベントの組合せとして格納し、
    　ユーザの操作により発生したイベントが、前記予測知識格納手段が格納する前記カレントイベントと一致する場合、当該カレントイベントと組み合わせにされている前記予測イベントに対応するオペレーションを実行する、
    　処理をコンピュータに実行させるプログラムを記録した不揮発性記録媒体。     A combination of events having continuity is stored as a combination of a current event that is a currently executing event and a predicted event that is an event predicted from the current event,
    When an event generated by a user operation matches the current event stored in the prediction knowledge storage unit, an operation corresponding to the prediction event combined with the current event is executed.
    A non-volatile recording medium that records a program that causes a computer to execute processing.
PCT/JP2013/000349 2012-02-02 2013-01-24 Process prediction execution device and process prediction execution method WO2013114830A1 (en)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
JP2012021139 2012-02-02
JP2012-021139 2012-02-02

Publications (1)

Publication Number Publication Date
WO2013114830A1 true WO2013114830A1 (en) 2013-08-08

Family

ID=48904880

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/JP2013/000349 WO2013114830A1 (en) 2012-02-02 2013-01-24 Process prediction execution device and process prediction execution method

Country Status (1)

Country Link
WO (1) WO2013114830A1 (en)

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US9916445B2 (en) 2014-02-26 2018-03-13 Mitsubishi Electric Corporation Attack detection device, attack detection method, and non-transitory computer readable recording medium recorded with attack detection program
JP2019133419A (en) * 2018-01-31 2019-08-08 シビラ株式会社 Data transmission/reception method, data transmission/reception system, processing device, computer program, and construction method for system
JP2019133650A (en) * 2018-01-31 2019-08-08 シビラ株式会社 Data transmission/reception method

Citations (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2004152276A (en) * 2002-10-09 2004-05-27 Matsushita Electric Ind Co Ltd Information terminal device, operation support method, and operation support program

Patent Citations (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2004152276A (en) * 2002-10-09 2004-05-27 Matsushita Electric Ind Co Ltd Information terminal device, operation support method, and operation support program

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US9916445B2 (en) 2014-02-26 2018-03-13 Mitsubishi Electric Corporation Attack detection device, attack detection method, and non-transitory computer readable recording medium recorded with attack detection program
JP2019133419A (en) * 2018-01-31 2019-08-08 シビラ株式会社 Data transmission/reception method, data transmission/reception system, processing device, computer program, and construction method for system
JP2019133650A (en) * 2018-01-31 2019-08-08 シビラ株式会社 Data transmission/reception method

Similar Documents

Publication Publication Date Title
US20220075643A1 (en) Unified resource management for containers and virtual machines
US10831633B2 (en) Methods, apparatuses, and systems for workflow run-time prediction in a distributed computing system
Li et al. Performance modeling and predictive scheduling for distributed stream data processing
US10956230B2 (en) Workload placement with forecast
Machida et al. Modeling and analysis of software rejuvenation in a server virtualized system with live VM migration
US9760420B1 (en) Fleet host rebuild service implementing vetting, diagnostics, and provisioning pools
US8978032B2 (en) Host naming application programming interface
Elmroth et al. A grid resource broker supporting advance reservations and benchmark-based resource selection
EP2859441B1 (en) Cloud application deployment portability
Larsson et al. Impact of etcd deployment on Kubernetes, Istio, and application performance
US20150095597A1 (en) High performance intelligent virtual desktop infrastructure using volatile memory arrays
US9110695B1 (en) Request queues for interactive clients in a shared file system of a parallel computing system
US11372871B1 (en) Programmable framework for distributed computation of statistical functions over time-based data
US10331488B2 (en) Multilayered resource scheduling
US10498817B1 (en) Performance tuning in distributed computing systems
US11327794B2 (en) Periodic task execution in an automated context
Ying et al. Optimizing energy, locality and priority in a mapreduce cluster
Wang et al. Design and implementation of an analytical framework for interference aware job scheduling on apache spark platform
JP2016091555A (en) Data staging management system
US11848983B2 (en) Remotely discover and map a network resource model in a cloud environment
Thamsen et al. Ellis: Dynamically scaling distributed dataflows to meet runtime targets
US8977752B2 (en) Event-based dynamic resource provisioning
WO2013114830A1 (en) Process prediction execution device and process prediction execution method
US20200012520A1 (en) Methods for scheduling multiple batches of concurrent jobs
US20170228383A1 (en) Active archive bridge

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: 13744397

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 13744397

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: JP