JP4685053B2 - Job execution order control device, method and program - Google Patents

Description

  The present invention relates to a technology of a computer system in which a plurality of devices cooperate to execute each job.

  2. Description of the Related Art Conventionally, a technology of a computer system is known in which a plurality of devices execute respective jobs in cooperation and provide a predetermined function as a whole of the plurality of devices (for example, Patent Document 1). For example, in a business system, a device that operates a business application and a device that operates a DBMS (database management system) that manages data used by the business application are provided. To provide various functions such as data editing and reference. When the business application uses the table 1 that is data managed in the DBMS, for example, after the DBMS executes a job or the like for generating the table 1, the business application displays the table 1 after the job ends. Then, a job or the like for editing is executed. In this case, information indicating the context such that a job for editing is executed after the job for generating the table 1 is registered.

JP 2005-173893 A

  Here, the business system is changed. For example, the business application uses not only the table 1 but also the table 2, whereby the execution of the job for generating the table 2 is executed after the job for generating the table 1 is executed in the DBMS. Is also needed. In this case, information indicating the context, such as a job for generating table 2 after the job for generating table 1 and a job for editing after the job for generating table 2 are executed. Changes are required.

  When this change is seen between devices, the context of the job that generates the DBMS table 1 and the job that edits the business application is discarded, and the job that generates the DBMS table 2 and the editing of the business application. The context before and after the job to be executed is newly registered. In other words, the apparatus on which the business application operates is affected by the change even though the job is executed in the apparatus on which the DBMS operates.

  If there is a one-to-one relationship between the device where the DBMS operates and the device where the business application operates, this change is not so much of a problem, but there are multiple relationships between the device where the business application operates and the device where the DBMS operates. If it has it (if it is a one-to-many relationship), it will be necessary to change the context for all of them, making the change complicated. Furthermore, in the case of a many-to-many relationship, there are cases where cooperation between devices is complicated and it is even difficult to grasp the context, and the change in this case is very difficult.

  SUMMARY An advantage of some aspects of the invention is that in a computer system in which a plurality of devices cooperate to execute each job, the influence of a job change executed on one device on other devices is reduced.

  An execution order control apparatus for a job according to an embodiment of the present invention executes a plurality of first jobs and executes a second job when execution of the plurality of first jobs is completed. And detecting the connection information for connecting the plurality of first jobs and the second job and the completion of all of the plurality of first jobs, the state of the connection information is changed. Monitoring means for performing the processing, and the second job executes processing when the state of the connection information is changed.

  In a preferred embodiment, the monitoring means is a job scheduler that controls the execution order of jobs, the connection information is a third job started by the job scheduler, and the job scheduler includes the plurality of job schedulers. When it is detected that all the first jobs are completed, the third job is executed, and when the execution of the third job is completed, the second job is executed.

  In a preferred embodiment, the third job is a dummy job.

  Hereinafter, an embodiment of the present invention will be described with reference to the drawings.

  FIG. 1 is a diagram illustrating a configuration example of a computer system including a job execution order control apparatus according to the present embodiment.

  The computer system according to the present embodiment (hereinafter sometimes referred to as “the present system”) includes one or more job execution terminals 100, one or more DB (database) servers 200, and a management terminal 400. The job execution terminal 100, the DB server 200, and the management terminal 400 are connected via a communication network 300, respectively. In the present embodiment, it is assumed that two job execution terminals 100 (job execution terminal A 100A and job execution terminal B 100B) and one DB server 200 are provided.

  The job execution terminal 100 is, for example, a computer (such as a personal computer) that provides a predetermined service using the DB server 200. Job execution terminal A 100A provides service a, and job execution terminal B 100B provides service b. Both the service a and the service b are services using a DB managed by the DB server 200. The job execution terminal A 100A includes a service a execution unit 110A that executes a job necessary for providing the service a. Further, the job execution terminal B 100B includes a service b execution unit 110B that executes a job necessary for providing the service b.

  The DB server 200 is, for example, a computer that manages a DB used for services provided by the job execution terminal 100. The DB server 200 includes a DB 220 that is used for the service a and the service b, and a DBMS 210 that executes a job for managing the DB 220.

  As described above, the job execution terminal 100 and the DB server 200 (hereinafter, referred to as “job execution apparatus” including both of them) in this embodiment are a job execution unit (here, service a execution unit) that executes a predetermined job. 110A, service b execution unit 110B, and DBMS 210). Each job execution unit 110 and 210 can execute each job in cooperation with other job execution units 110 and 210 other than itself. Cooperation here means that a job of a certain job execution unit 110, 210 and a job of another job execution unit 110, 210 are executed in a context. For example, when there is a relationship in which a job of another job execution unit 110 or 210 is started after a job of a job execution unit 110 or 210 is completed, the job execution units 110 and 210 are , Will be linked. The cooperation between the job execution units 110 and 210 is managed by the cooperation management unit 410 (and the scheduler 420), which will be described later, as information indicating the front-rear relationship of each job.

  The job execution units 110 and 210 are not limited to the service a execution unit 110A (service b execution unit 110B) and the DBMS 210 as exemplified in this embodiment. The job execution units 110 and 210 may be predetermined applications that execute jobs linked to jobs of other job execution units 110 and 210. Further, the present system does not necessarily include both the job execution terminal 100 and the DB server 200. For example, when the job execution units 110 and 210 provided in the plurality of job execution terminals 100 are linked to each other, the system is configured by the job execution terminals 100 without including the DB server 200. You can also.

  The management terminal 400 is a computer (such as a personal computer) that manages the job execution apparatuses 100 and 200 and enables the job execution units 110 and 210 included therein to provide predetermined functions. The management terminal 400 includes a cooperation management unit 410, a scheduler 420, and a schedule 430.

  The cooperation management unit 410 manages information about points (hereinafter referred to as “cooperation points”) at which the different job execution units 110 and 210 cooperate to execute each job. Here, the cooperation point is a connection point of jobs when a job of a job execution unit 110 or 210 and a job of another job execution unit 110 or 210 are connected in a context. That means. The information related to the linkage point includes, for example, information indicating jobs before and after forming the linkage point, and information indicating whether or not the previously executed job is completed, that is, the execution of the job executed later. Including information indicating whether or not to start. The cooperation point and information related to the cooperation point will be specifically described with reference to FIG.

  FIG. 2 is a conceptual diagram illustrating an example of a cooperation point.

  The figure shows a “service a operation program (J12A)” job executed by the service a execution unit 110A, a “service b operation program (J12B)” job executed by the service b execution unit 110B, and the DBMS 210. A cooperation point 510 formed by the “DB backup (J24)” job to be executed is shown. The arrows in the figure indicate that “DB backup (J24)” after the execution of both the “service a operation program (J12A)” job and the “service b operation program (J12B)” job are completed. Indicates that the job will be executed.

  The information 520 related to the cooperation point includes, for example, connection information 521 and a state 522. The link information 521 includes a job of “service a operation program (J12A)” and a job of “service b operation program (J12B)” that are executed first, and a job “DB” that is executed later. The backup (J24) "job is recorded in association with each other. The state 522 is information indicating whether or not the execution of a job to be executed later may be started. For example, in the state 522, “executable” is set when execution of a job to be executed later may be started, and “executable” is set when the job cannot be started. Therefore, in this example, in the state 522, the execution of both the “service a operation program (J12A)” job and the “service b operation program (J12B)” job is “executable”. Is set, otherwise “unexecutable” is set.

  The correspondence between the job executed first and the job executed later is limited to the case where there are two jobs executed first and one job executed later as in this example. Absent. There may be a one-to-one relationship between one job executed first and one job executed later, one of which is plural and the other is one-to-one (or one-to-many). It may be related. Furthermore, there may be a plurality of jobs executed first and jobs executed later, and a many-to-many relationship may be established. When there are a plurality of jobs to be executed later, the “executable” information set in the state 522 indicates that the execution may be started for each of all jobs to be executed later.

  The cooperation management unit 410 includes, for example, a cooperation point together with a scheduler 420 described later by providing a pseudo job (an empty job that does not perform any processing, hereinafter referred to as “dummy job”) for each cooperation point 510. Information 520 related to this can be managed. When using a dummy job, the cooperation management unit 410 executes the dummy job in accordance with an instruction from the scheduler 420 (however, since the executed job does not perform any processing, it ends immediately). Information indicating that the execution of the dummy job has ended is information indicating that the execution of a job to be executed later may be started. In other words, if the execution of the dummy job is completed, it is treated that “executable” is set in the above state 522, and if the execution of the dummy job is not completed, “execution” It will be treated as if “impossible” is set. The scheduler 420 can detect whether or not the execution of the dummy job has ended, that is, whether the status 522 is “executable” or “not executable”.

  Also, the cooperation management unit 410 can output information 520 regarding the cooperation points. As a result, an administrator who manages this system can easily know the job execution status at the linkage point 510 and can easily understand the system operation status.

  Returning to FIG. 1, when this system provides a predetermined function, the scheduler 420 instructs the job execution units 110 and 210 and the cooperation management unit 410 to start a job according to the contents of the schedule 430. Further, the scheduler 420 detects whether or not the job currently being executed has been completed in order to execute the jobs in order. For example, when providing the service a and the service b, the scheduler 420 refers to a schedule 430 as shown in FIG. 2 to be described later so that the job is executed in the execution order of the jobs registered in the schedule 430. The service a execution unit 110A, the service b execution unit 110B, or the cooperation management unit 410 is instructed to start a job.

  The schedule 430 is information regarding the job execution order that the scheduler 420 refers to when instructing the job execution units 110 and 210 to start a job. For example, in the schedule 430, for a function provided by the system (service a, service b, etc.), a job executed to provide the function and an execution order thereof are recorded. The schedule 430 can be recorded in a table format as shown in FIG. 2, for example.

  FIG. 3 is a diagram illustrating an example of the schedule 430 according to the present embodiment.

  In the schedule 430, for example, a preceding job 431 and a succeeding job 432 are stored in association with each other. That is, in the drawing, the job indicated in the succeeding job 432 is executed after the job indicated in the preceding job 431.

  In the preceding job 431, a job execution part name 4311 and a job name 4312 are recorded. The job name 4311 is the name of the job corresponding to the preceding job 431. The job execution unit name 4311 is the name of the job execution unit 110 or 210 that executes a job corresponding to the preceding job 431. As with the preceding job 431, the job execution unit name 4321 and the job name 4322 are recorded for the subsequent job 432 as well. An identifier for uniquely specifying the job execution units 110 and 210 may be recorded instead of the job execution unit names 4311 and 4321, and an identifier for uniquely specifying the job instead of the job names 4312 and 4322 may be recorded.

  In the drawing, an example of the execution order of jobs executed to provide the service a and the service b is described. For example, referring to the schedule 430, it is understood that the job of “data input of table a (J22)” of the DBMS 210 is executed after the job of “create table (J21)” of the DBMS 210. In addition, the “service a operation program (J12A)” job of the service a execution unit 110A is the same as the “date update (J11A)” job of the service a execution unit 110A and the “use of table a” ( J41) "is executed after both of the jobs are completed.

  In addition, jobs whose job execution unit names 4311 and 4321 are linkage management units, that is, “use start of table a (J41)”, “start use of table b (J42)”, and “end use of table a (J43)”. "And" End of use of table b (J44) "are all dummy jobs. As shown in the figure, the dummy job is also registered in the schedule 430 in the same manner as a normal job (a job executed by the job execution units 110 and 210). Then, the context of the two jobs forming the cooperation point 510 is registered in the schedule 430 via the dummy job. For example, the “service a operation program (J12A)” job of the service a execution unit 110A needs to be executed after the “table a data input (J22)” job of the DBMS 210 is completed. The relationship is registered through a dummy job of “start use of table a (J41)” in the cooperation management unit 410. That is, when the job of “data input of table a (J22)” of the DBMS 210 is completed, a dummy job of “start use of table a (J41)” of the cooperation manager 410 can be started, and “table a” of the cooperation manager 410 can be started. When the dummy job “Start use (J41)” is completed, the “service a operation program (J12A)” job of the service a execution unit 110A can be started (here, “date update (J11A)” job) Is also a condition).

  As described above, the cooperation point 510 is registered in the schedule 430 as a dummy job, so that the scheduler 420 does not need to be particularly aware of the cooperation point 510. Therefore, it becomes possible to realize management of job cooperation between apparatuses by the same mounting method as before.

  FIG. 4 is a diagram showing a processing flow when the system provides service a and service b.

  First, the scheduler 420 instructs the DBMS 210 to start a “table creation (J21)” job. Receiving the instruction, the DBMS 210 executes a job “Create table (J21)” (S01). When this job is executed, a table a used for the service a and a table b used for the service b are respectively created on the DB 220.

  After the “table creation (J21)” job is completed, the scheduler 420 instructs the DBMS 210 to start the “table a data input (J22)” job. Upon receiving the instruction, the DBMS 210 executes a job of “data input of table a (J22)” (S02). When this job is executed, predetermined data is written in the table a, and the table a becomes usable.

  After the “table a data input (J22)” job is completed, the scheduler 420 instructs the cooperation management unit 410 to start the “table a use start (J41)” job. On the other hand, the start of the job “data input of table b (J23)” is instructed. The job “start use of table a (J41)” executed by the cooperation management unit 410 is a dummy job. Therefore, this job is immediately terminated without actually performing anything (S03). As described above, the information that the execution of the dummy job has ended indicates that the execution of the job to be executed later (that is, the job of the “service a operation program (J12A)”) may be started. Information. On the other hand, when the “data input of table b (J23)” job is executed, predetermined data is written into the table b, and the table b becomes usable (S07).

  On the other hand, the scheduler 420 instructs the service a execution unit 110A to start a “date update (J11A)” job. Similarly, the scheduler 420 instructs the service b execution unit 110B to start a “date update (J11B)” job. These instructions may be given before instructing the DBMS 210 to start a “table creation (J21)” job, or after giving an instruction to start the “table creation (J21)” job. Also good. Upon receiving the instruction, the service a execution unit 110A and the service b execution unit 110B each execute a “date update (J11A, J11B)” job (S04, S09).

  After the “date update (J11A)” job in the service a execution unit 110A ends and the “table a use start (J41)” job ends, the scheduler 420 instructs the service a execution unit 110A to “ The start of the job of the service a operation program (J12A) is instructed. Upon receiving the instruction, the service a execution unit 110 executes a job of “service a operation program (J12A)” (S05). When this job is executed, provision of service a is started. The service a may be provided not only by this job ("service a operation program (J12A)" job) but also by a combination of execution of a plurality of jobs.

  When the provision of the service a is completed (here, the job of the “service a operation program (J12A)” is completed), the scheduler 420 sends a job of “use of table a (J43)” to the linkage management unit 410. Instruct the start of. Similar to the job “start use of table a (J41)”, the job “end use of table a (J43)” executed by the cooperation management unit 410 is a dummy job and immediately ends (S07). The information indicating that the job has ended is information indicating that execution of a job to be executed later (that is, a “DB backup (J24)” job) may be started.

  Processing similar to the processing for providing service a as described above is performed for service b.

  That is, after the job of “data input of table b (J23)” is completed, the scheduler 420 instructs the cooperation management unit 410 to start the job of “start use of table b (J42)”. As in the case of table a, the job “start use of table b (J42)” executed by the cooperation management unit 410 is a dummy job, and is immediately ended (S08). Then, after the “date update (J11B)” job in the service b execution unit 110B ends and the “table b use start (J42)” job ends, the scheduler 420 sends the service b execution unit 110B to the service b execution unit 110B. To instruct the start of the “service b operation program (J12B)” job. Upon receiving the instruction, the service b execution unit 110 executes a job of “service b operation program (J12B)” (S10). When this job is executed, provision of service b is started. Thereafter, when the job of “service b operation program (J12B)” is completed, the scheduler 420 instructs the cooperation management unit 410 to start the job of “use of table b (J44)”. As in the case of table a, the job “end use of table b (J44)” executed by the cooperation management unit 410 is a dummy job and immediately ends (S11).

  Finally, after the job “End use of table a (J43)” and the job “End use of table b (J44)” are finished, the scheduler 420 sends a “DB backup ( J24) "is started. The DBMS 210 that has received the instruction backs up the DB 220 (S12).

  The above is the description of the process flow when the system provides the service a and the service b.

  Here, for example, it is assumed that the “service a operation program (J12A)” job executed by the service a execution unit 110A is changed to use the table b in addition to the table a. In this case, before the “service a operation program (J12A)” job is executed, the “data input of table b (J23)” job needs to be completed. Therefore, by setting the preceding job of the job of “service a operation program (J12A)” to “start use of table b (J42)” in addition to the previous “start use of table a (J41)”, This change can be accommodated. That is, this change can be handled without affecting the DBMS 210.

  In the present embodiment, four linkage points (dummy jobs corresponding to linkage points) are created in consideration of the job context between job execution units. That is, in the present embodiment, the “service start program (J12A)” job is executed after the “table a data input (J22)” job and the “service start program (J12A)” job. The jobs are created so as to concatenate those jobs based on the context in which they are executed (the same applies to the table b). The context itself is obtained, for example, from the system requirements that the tables a and b need to be used by the other job execution units 110 and 210 after being initialized in the DBMS 210. Further, the dummy job “use of table a (J43)” is performed before and after the “DB backup (J24)” job is executed after the “service a operation program (J12A)” job is executed. It is created based on the relationship (the same applies to the table b). This context is obtained from the system requirement that the tables a and b need to be backed up by the DBMS 210 after the editing is completed. In other words, in the present embodiment, it can be considered that the linkage point is created in accordance with the requirements of the system that requests such a job context.

  On the other hand, the cooperation point can be created independently for each of the job execution apparatuses 100 and 200 without considering such system requirements. That is, for each job execution device 100, 200, the predetermined processing is completed in consideration of processing performed by itself (without considering a specific connection point with other job execution devices 100, 200). The point in time can be used as a linkage point. The time point when the predetermined process is finished can be, for example, the time point when the job is desired to be notified to the other job execution apparatuses 100 and 200. For example, when focusing on the DB server 200, since the table a and the table b are data disclosed to the other job execution apparatuses 100 and 200, status changes regarding the table a and the table b (for example, generation, deletion, It is desirable to notify other job execution apparatuses 100 and 200 of the access right change or the like. Accordingly, regarding the table a and the table b, the end point of a job for changing the status (for example, a job for creating a table, a job for deleting a table, a job for changing an access right, etc.) may be used as a linkage point. it can. By doing so, the meanings of the linkage points themselves (in this case, which status the table is in) is clarified, so that the other job execution apparatuses 100 and 200 are connected to the DB server 200. When considering the linkage, it becomes possible to easily grasp to which linkage point my job should be linked.

  Then, by monitoring the state 522 of the linkage point corresponding to the end point of such processing, it is possible to grasp how far the processing has progressed as a whole system, that is, the progress status of the whole system. For example, if the status 522 of the cooperation point corresponding to the job whose access right is changed is “executable”, it can be determined that the change of the access right is completed.

  Further, when a failure occurs in a certain job execution device 100, 200, and the processing is stopped or delayed, the job execution device 100 affected by the failure is examined by examining the linkage points related to the processing. , 200 (job execution apparatuses 100 and 200 that execute processing linked to the linkage point) can be easily specified.

  The above-described embodiments of the present invention are examples for explaining the present invention, and are not intended to limit the scope of the present invention only to those embodiments. The present invention can be implemented in various other modes without departing from the gist thereof.

It is the figure which showed the example of a structure of the computer system provided with the apparatus which manages the cooperation of the job between apparatuses which concern on this embodiment. It is the conceptual diagram which showed an example of the cooperation point. It is the figure which showed an example of the schedule which concerns on this embodiment. It is the figure which showed the flow of the process at the time of this system providing the service a and the service b.

Explanation of symbols

  DESCRIPTION OF SYMBOLS 100 ... Job execution terminal, 110 ... Service execution part, 200 ... DB server, 210 ... DBMS, 220 ... DB, 300 ... Communication network, 400 ... Management terminal, 410 ... Cooperation management part, 420 ... Scheduler, 430 ... Schedule

Claims (3)

A job execution order control device that executes a plurality of first jobs and executes a second job when execution of the plurality of first jobs is completed,
A job schedule registered via a third job, which is an empty dummy job that does not perform any processing, the context of the plurality of first jobs and the second job,
A job scheduler that controls the execution order of jobs according to the job schedule,
The job scheduler causes the third job to be executed when it is detected that all of the plurality of first jobs are completed, and causes the second job to be executed when the execution of the third job is completed. A job execution order control device. A job execution order control method for executing a plurality of first jobs and executing a second job upon completion of execution of the plurality of first jobs,
Control the job execution order according to the job schedule registered via the third job, which is an empty dummy job in which the first job and the second job do not perform any processing. The job scheduler executes the third job when it is detected that all the plurality of first jobs are completed, and executes the second job when the execution of the third job is completed. The job execution order control method. A job execution order control program for executing a plurality of first jobs and executing a second job upon completion of execution of the plurality of first jobs,
When the execution order control program for the job is executed on the computer,
Control the job execution order according to the job schedule registered via the third job, which is an empty dummy job in which the first job and the second job do not perform any processing. The job scheduler executes the third job when it is detected that all the plurality of first jobs are completed, and executes the second job when the execution of the third job is completed. A job execution order control program.

Images